SAPC Conference Programme 2014 by mark orriss

201 4 th Jul y 11 9t h-

y mar p r ir e: ca

g n i t e e l m a b glo es g n le l a ch E PR

A GR

y s it r ive

43 rd An nu al

ry a rim P c i em d a c rA o f iety c o S

urg b din E f

re a C

Co nfe ren ce - Ed inbur gh uk c. .a pc sa w. ww on: ati r: rm witte Info on T pcasm w us sa Follo pcacuk # @sa

Try out the SAPC APP from the app store

Save the dates in your diary for the

44th Annual Conference of the Societ y for Academic Primar y Care 8th to 10th July 2015 University of Oxford Abstract submission deadline: 10am Monday 23rd February 2015

Venue

Andrew Wiles Building, Mathematical Institute, Radcliffe Observatory Quarter, Oxford

Accommodation

Keble College and St Anneâ&#x20AC;&#x2122;s College - both just a few minutes walk from the venue

Social events

Drinks reception: Wednesday 8th July 2015, Andrew Wiles Building Conference dinner: Thursday 9th July 2015, Keble College Dining Hall

WELCOME

Dr Karen Fairhurst

Professor David Weller

We are delighted to be hosting the 43rd annual conference of the Society for Academic Primary Care at the University of Edinburgh in July 2014. Edinburgh is Scotland’s capital but also a truly international city with a rich academic and medical heritage. The theme for our 2014 meeting is ‘Meeting Global Challenges’ and the conference will be an opportunity to explore the many diverse challenges facing primary care practitioners, researchers and educators in the UK and around the world. To lead us in this task, we have three internationally renowned keynote speakers, backed up by an academic programme designed to stimulate discussion and lively debate. New to the SAPC conference this year: As part of the conference we are honoured to host the inaugural Helen Lester Memorial Lecture on Wednesday evening. The SAPC APP, we hope you will enjoy this innovation F Elevator pitches are a new concept for the SAPC annual conference and we hope they will provide a dynamic way for new ideas and work in progress to be discussed. F Also new this year we will welcome Special Interest Group meetings into the body of the programme on Thursday afternoon. F

We have a broad range of interesting submissions to the programme including workshops, oral presentations, ‘elevator pitches’ and dangerous ideas. Our social events will provide a great way to network with colleagues and make new acquaintances. On Wednesday evening there is the drinks reception at Old College and on Thursday, the conference dinner in the stunning Dynamic Earth. If you have not yet booked please enquire at the registration desk. All in all, we hope you spend a stimulating few days with your friends and colleagues in one of the world’s most beautiful cities. We look forward to welcoming you in Edinburgh in July. Karen Fairhurst & David Weller SAPC Conference Chairs, University of Edinburgh

Conference committee - University of Edinburgh Dr Christine Campbell - programme co-chair Dr Karen Fairhurst - conference co-chair Professor Brian McKinstry - programme co-chair Professor Scott Murray - dangerous ideas Professor David Weller - conference co-chair

Conference organisers Sue Stewart and Sharon Pidgeon, SAPC www.sapc.ac.uk Tel: 01865 331839

CONTENTS

PAGE

Programme

5-13

Keynote speakers 14 Social programme and optional activities 16 Peer review 18 Conference information 19 Exhibitors and sponsors 19 Special Interest Groups and Meetings

20-21

Workshops

22-25

Abstracts of oral presentations

26-75

Presenter index

76-81

Map

inside back cover

Appleton Tower venue plan

outside back cover

A NOTE ABOUT THE ABSTRACTS This book contains the abstracts for workshops and oral presentations in parallel sessions. Elevator pitch abstracts can be found on the SAPC APP and at www.sapc.ac.uk on the 2014 conference abstract and programme page. PROGRAMME AND ABSTRACT NUMBERING The abstracts are listed in programme order, for example

Parallel oral sessions: Programme number 1A.1 = 1= parallel session 1 (Wednesday), A = the first session within parallel 1 - Research methods .1 = the first presentation in parallel 1 session A

Wednesday 9th July 2014

Morning

HODs / HOTs / Former HODs - see programme pages

Appleton Tower

12.00-12.45

LUNCH

Appleton Tower

12.45-14.00

WELCOME AND OPENING SESSION

Opening address: Professor Sarah Cunningham-Burley, Head of the School of Molecular, Genetic and Population Health Sciences, University of Edinburgh

Plenary 1: Professor Sir Andy Haines

George Square Theatre

14.00-14.30

Presentations of distinction

George Square Theatre

14.30-15.00

TEA BREAK

Appleton Tower

15.00-16.15

Elevator pitch session 1 - see programme pages

Appleton Tower

16.30-17.30

Parallel session 1 Oral sessions and workshop - see programme pages

Appleton Tower

17.45-18.30

Inaugural Helen Lester Memorial Lecture Professor Debbie Sharp, University of Bristol

George Square Theatre

18.45-20.00 Drinks reception for conference delegates

The Playfair Library Old College

Thursday 10th July 2014

08.00-08.55

Breakfast meeting: Primary healthcare scientists (PHoCuS) group. Room 2.14, 2nd floor

Appleton Tower

09.00-09.45

Plenary 2: Professor Ruth McDonald

George Square Theatre

09.45-10.30

Yvonne Carter award and presentation Dr Helen Atherton Most Distinguished Paper from the Primary Health Care Research Conference 2013 Australia Dr Megan Elliott-Rudder

George Square Theatre

10.30-11.00

COFFEE BREAK - Primary healthcare scientists (PHoCuS) group gathering outside LT4, 1st floor

Appleton Tower

11.00-12.30

Parallel session 2 Oral sessions and workshop - see programme pages

Appleton Tower

12.30-14.00

LUNCH

Appleton Tower

12.55-13.55

SAPC AGM

Appleton Tower

14.00-15.30

Parallel session 3 Oral sessions and workshop - see programme pages

Appleton Tower

15.30-16.00

TEA BREAK

Appleton Tower

16.00-17.30

Special Interest Groups (SIGs) & workshops See programme pages

Appleton Tower

17.45

Optional activities - see programme pages

19.30-24.00 Conference Dinner

Dynamic Earth

Friday 11th July 2014

09.00-09.45

Plenary 3: Professor Frede Olesen

George Square Theatre

10.00-11.00

Elevator pitch session 2 - see programme pages

Appleton Tower

11.00-11.30

COFFEE BREAK

Appleton Tower

11.30-12.30

Parallel session 4 Oral sessions and workshop - see programme pages

Appleton Tower

12.35-13.00

CLOSING SESSION

Appleton Tower

13.00

Packed lunch to take away

Appleton Tower

13.30

Golf - taxi departs

Appleton Tower 5

LUNCH

12.00-12.45

3 minute presentations plus 2 minutes Q&A

EP-1B.04: Prescribing and investigations behaviour among

EP-1B.03: Patients experiences of patient safety in primary care in England. A systematic review and meta-synthesis. Jose Valderas

EP-1B.02: Patients’ perceptions of Chronic Disease Management within the Irish healthcare system Tatsiana Seraukina

EP-1A.02: Factors associated with later presentation and later stage at diagnosis of colorectal cancer: a prospective cohort study. Fiona Walter

EP-1A.03: Development of the ‘Aarhus Statement Tool for Researching Intervals in Diagnosis’ (ASTRID): Focus groups and interviews with patients, researchers and healthcare professionals Domenica Coxon

EP-1B.01: Ethnographic study of general practices’ engagement with clinical guidelines 2 Bruno Rushforth

EP-1A.01: Trial Of Personalised Care After Treatment - Prostate cancer (TOPCAT-P) Andrei Stanciu

EP-1C.03: Trust, cultural norms, networks and the economy: factors that affect cardiovascular risk prevention in the Pakistani community, UK Farina Kokab

EP-1C.02: Self-monitoring of blood pressure in pregnancy: The BuMP study Carole Crawford

EP-1C.01: Receptionist rECognition and rEferral of Patients with Stroke (RECEPTS): a prospective crosssectional observational study James Sheppard

EP-1D.03: Trends in the use of potentially inappropriate medicines in UK general practice prescribing for older people: A cross sectional analysis for 2003/4, 2007/8 and 2011/12. Suzanne Richards

EP-1D.02: Benzodiazpine prescribing in Irish children Kirsty O’Brien

EP-1D.01: Prevalence and clinical and educational associations of antibiotic prescribing for respiratory infections by Australian general practice trainees: a cross-sectional analysis from the ReCEnT study. Parker Magin

EP-1E.03: Diabetes prevention in the real world: Systematic review of the effectiveness of pragmatic lifestyle interventions for the prevention of type 2 diabetes and of the impact of guideline recommendations on effectiveness. Colin Greaves

EP-1E.02: Behaviour change initiation in adults at risk of developing diabetes Kelly Blockley

EP-1E.01: Physical and psychosocial outcomes after unintentional injury: short term outcomes from the Impact of Injuries Study Blerina Kellezi

EP-1F.03: Development and validation of a pain management instrument for population-based pain studies Christine Bond

EP-1F.02: What are the effects of an enhanced participant information sheet on recruitment in the Healthlines Study? A nested randomised controlled trial using the START model. Peter Bower

EP-1F.01: How representative are physical activity trial participants? results from the PACE-UP primary care trial Katy Morgan

ELEVATOR PITCH 1F Research methods, epidemiology and public health Room: 2.14, 2nd floor Chair: Brendan Delaney

ELEVATOR PITCH 1A Cancer and palliative care Room: LT1, ground floor Chair: Scott Murray

ELEVATOR PITCH 1 15.00-16.15

ELEVATOR PITCH 1E Diabetes, managing long term conditions and clinical consultation Room: LT5, 1st floor Chair: Joanne Reeve

TEA BREAK

14.30-15.00 ELEVATOR PITCH 1D Prescribing and patient safety Room: LT4, 1st floor Chair: Tony Avery

Multiple conditions and clinical trials: A systematic review of the impact of multimorbidity in inclusion, assessment and outcome of self-management studies Cassandra Kenning, University of Manchester

14.15 - PP2

ELEVATOR PITCH 1C Cardiovascular Room: LT3, ground floor Chair: Margaret Cupples

Presentations of distinction - highest scoring abstracts Teaching and training practices: in the wrong places and with the wrong patients? Eliot L Rees, Keele University

14.00 - PP1

ELEVATOR PITCH 1B Organisation and delivery of primary health care Room: LT2, ground floor Chair: John Campbell

WELCOME AND OPENING SESSION Chairs: David Weller and Karen Fairhurst Opening address: Professor Sarah Cunningham-Burley, Head of the School of Molecular, Genetic and Population Health Sciences, University of Edinburgh Plenary 1: Professor Sir Andy Haines, London School of Hygiene and Tropical Medicine Achieving Universal Health Coverage - the role of primary care research

12.45 -14.00

150m walk to George Square Theatre - See map on inside back cover

Appleton Tower Rooms on 2nd floor

HODs - Room 2.11 HOTs - Room 2.07 Former HODs - Room 2.06

09.00-11.45

EP-1G.04: Is the GP specialty programme clinical structured

EP-1G.03: Title: Role of providing core objectives and monitoring in Peer Assisted Learning Initiatives in Medical Education: a 2 year study Aya Musbahi

EP-1G.02: Evaluation of an on-line learning module to improve Prescribing in Primary Care Richard Knox

EP-1G.01: Correlation between medical student Feedback on General Practice placements and Patient Satisfaction Surveys Roxanne Keynejad

ELEVATOR PITCH 1G Pot pourri of topics Room: 2.11, 2nd floor Chair: Chris Burton

Appleton Tower

George Square Theatre

Appleton Tower

Location

Time

PROGRAMME: WEDNESDAY 9TH JULY 2014

EP-1A.12: Identifying vulnerability in grief: implications for primary care Roisin Bartlam

EP-1A.11: Hospice Enabled Dementia Care: where are we now and where do we need to be? Jacqueline Crowther

EP-1A.10: Informal Carer Experiences of End-of-Life Care and Bereavement: secondary analysis of qualitative interviews Grace Moore

EP-1A.09: Improving palliative care: an analysis of significant event reports in primary care Nicola Stirling

EP-1B.09: The Role of Reception Staff in the Triage of Patients Presenting to General Practice with Acute Stroke: Insights of Receptionists and Primary Care Staff Elizabeth Bates

EP-1C.08: Socio-cultural factors associated with cardiovascular risk in the Pakistani community: a narrative review Farina Kokab

EP-1B.08: A cross-sectional survey study of service users’ experiences of out-of-hours primary medical care in England Raff Calitri

EP-1A.08: Prolonged dwindling characterises the illness trajectory of nursing home residents Anne Finucane

EP-1C.07: Ethnic disparities in the consequences of the proposed reduction in the 10-year CVD risk treatment threshold Mark Ashworth

EP-1B.07: How doctors, politicians and managers work together in the new National Health Service: the example of health inequalities in north England Una Macleod

EP-1A.07: Comparison of Urgent Suspected Gynaecological Cancers with all Urgent Suspected Cancer Referrals: secondary analysis of a Scottish audit of compliance with referral guidelines and compliance. Pauline Williams

EP-1C.11: The prognostic utility of tests of platelet function for the detection of “aspirin resistance” in patients with established cardiovascular or cerebrovascular disease: A systematic review and economic evaluation David Fitzmaurice

EP-1C.10: The inter-arm difference in blood pressure: prevalence varies for different populations Christopher Clark

EP-1C.09: Accuracy of methods for diagnosing Atrial Fibrillation: A systematic review Jaspal Taggar

EP-1C.06: Does the predicted MVE (major vascular event rate) used by the CTTC show any correlation to the Framingham 10 year risk score when predicting baseline cardiovascular risk? Kamalpreet Cheema

EP-1B.06: Patients’ willingness to attend the NHS Cardiovascular Health Checks in primary care: A qualitative interview study Suzanne Richards

EP-1A.06: Incidence and Nature of Medical Litigation Claims for Delays or Failures in Cancer Diagnosis and Treatment Samantha Worrall

EP-1C.05: Incidence and Survival of Patients with a First Diagnostic Label of Heart Failure in General Practice Clare J Taylor

EP-1C.04: Accuracy of methods for detecting Atrial Fibrillation: A systematic review Jaspal Taggar

EP-1B.05: How do ‘poor performing’ general practices respond to QOF monitoring? Findings from five qualitative case-studies. Maria Kordowicz

General Practitioners consulting face-to-face and by telephone following an urgent request for care: insights from the ESTEEM trial Fiona Warren

EP-1A.05: The views and experiences of cancer patients towards the screening, diagnosis and management of emotional distress Christine Bond

EP-1A.04: Is the quality of cancer care in English Practices related to patient access to primary care? Kosala Perera

EP-1D.12: Antithrombotic treatment patterns for stroke prevention in relation to age: insights from the UK cohort of the international GARFIELD Registry Patricia Apenteng

EP-1D.11: Improving the quality of decision-making and safety in prehospital and urgent care transitions Maxine Johnson

EP-1D.10: An investigation if follow-up monitoring of cardiovascular medication therapy in UK primary care Vladislav Berdunov

EP-1D.09: GP practice views of clinical action performance measures: hyperlipidaemia control in patients with type 2 diabetes Rachel Foskett-Tharby

EP-1D.08: Feasibility and acceptability of a real-time feedback intervention to improve patient experience in general practice: preliminary results Christine Wright

EP-1D.07: Piloting a Patient Safety Toolkit for General Practice Kate Marsden

EP-1D.06: Describing the Pharmacoepidemiology of Antidepressant Prescribing in Canada using primary care data from The Canadian Primary Care Sentinel Surveillance Network (CPCSSN) Rachael Morkem

EP-1D.05: Acute topical - blocker exposure in asthma: a meta-analysis of controlled clinical trials Daniel Morales

EP-1D.04: A process evaluation of a cluster randomised controlled trial aiming to reduce potentially inappropriate prescribing in older patients in primary care (the OPTI-SCRIPT study) Barbara Clyne

EP-1E.09: Clinical trials for elderly patients with multiple diseases (CHROMED) pilot study Jo Middlemass

EP-1E.12: The role of primary care for Multiple Sclerosis: Experiences of patients and primary care professionals Abigail Methley

EP-1E.11: The “General Check Up” in the asymptomatic adult- a study of GPs in the North West of Ireland Mark Murphy

EP-1F.12: Trends of substance misuse and pharmaceutical treatment recorded in England and Wales General Practice (19942012). Hilary Davies

EP-1F.11: Risk of venous thromboembolism in individuals with Sickle Cell Trait: a cohort study using CPRD records Iain Little

EP-1F.10: Ethnicity and COPD prevalence in South-East London - is there less COPD in the Black Population? Alexander Gilkes

EP-1F.09: Risk prediction models to predict emergency hospital admission in community-dwelling adults: a systematic review Emma Wallace

EP-1E.08: Explaining variation in long term prescribing of opioids in primary care Ben Leaman

EP-1E.10: Commissioning for Long-term conditions. Hearing the voice of and engaging users: the role of clinical commissioing groups Patricia Wilson

EP-1F.08: Evaluation of the Alcohol Intensive Case Management Service Pilot in Nottingham City Jacqueline Mhizha-Murira

EP-1E.07: Validation of five noninvasive respiratory rate monitors in patients with COPD Noah Rubio

EP-1F.06: Confirming Whether a Model of a Clinical Process Built by Academics Matches the Experience of Patients and Clinicians Melanie Rimmer

EP-1F.05: Making continuous outcomes meaningful to clinicians Sally Kerry

EP-1F.04: Improving our understanding of treatment effectiveness by improving our use of therapeutic compliance data: making the case for CACE analysis Melanie Chalder

EP-1F.07: Regional differences in health inequalities associated with burn injury incidence in children under 5 years old: A UK populationbased cohort study Elizabeth Orton

EP-1E.06: “The health trainers are better at doing it than we are”: an evaluation of a lay-led long-term conditions pilot in County Durham, England Shelina Visram

EP-1E.05: ‘Light Touch’ telemonitoring service for people with chronic obstructive pulmonary disease (COPD): a pilot mixed method evaluation. Hilary Pinnock

EP-1E.04: The Accuracy and Reliability of Over-the-Counter Blood Pressure Monitors Daniel Wright

EP-1G.12: Gout - ‘not something you brag about’. A qualitative study of health related quality of life in gout. Jenny Liddle

EP-1G.11: Health Related Quality Of Life in gout: cross-sectional analysis from a prospective cohort study Jenny Liddle

EP-1G.10: A longitudinal evaluation of the patient public involvement research user group in the Greater Manchester Primary Care Patient Safety Translational Research Centre S Jill Stocks

EP-1G.09: Involving patients in a systematic review: how, why and what does it add? Catherine Hyde

EP-1G.08: Title: Development of the Gateway for Patient and Public Involvement Learning (G-PPIL): Engaging stakeholders in the process of developing an online learning resource for patients, carers and members of the public Kate Hill

EP-1G.07: ‘Tis the season to be… free from nocturnal leg cramps? The marked seasonality of quinine starts and cramp-related (Google) internet search volume Scott Garrison

EP-1G.06: An educational intervention for GPs to improve the management of musculoskeletal conditions with a focus on referral behaviour Victoria Tzortziou Brown

EP-1G.05: GP trainees’ seeking of in-consultation information and advice: findings from the Registrars Clinical Encounters in Training (ReCEnT) study. Parker Magin

reference form valid and reliable? One UK postgraduate deanery retrospective study Rodger Charlton

18.45-20.00

17.45-18.30

17.15 1C.4: Exercise referral schemes: GPs’ perceptions of barriers and facilitators to referral Neil Heron

Drinks reception for conference delegates

Professor Debbie Sharp, University of Bristol Motherhood and mental illness: thirty years of families in south London

INAUGURAL HELEN LESTER MEMORIAL LECTURE

MOVE TO GEORGE SQUARE THEATRE

1B.4: Long-term opioid prescribing for non-malignant pain in primary care: a qualitative study Carolyn Montana

1C.3: The use of champions to promote digital enablement in the community: a dallas evaluation study Claire Nugent

1B.3: Application of Prescribing Safety Indicators to the Clinical Practice Research Database: a retrospective cohort study S Jill Stocks

1A.3: The Primary Health Care Search Filter: Making a world of difference to literature searching for primary care professionals Lynsey Brown

1A.4: Establishing RAPPORT findings from a realist evaluation of patient and public involvement in NHS research Amanda Howe

1C.2: Addressing the lack of reliability and validity data from the General Practice Physical Activity Questionnaire (GPPAQ) using PACE-Lift physical activity trial data in 60-74 year old primary care patients Shaleen Ahmad

1B.2: Opioid prescribing across Scotland - rates and variations Nicola Torrance

1A.2: Advancing primary care through education and research: from translational research to translational scholarship Lucy Cooper

16.45

17.00

1C.1: Obesity in men: evidence synthesis findings and their implications for primary care Pat Hoddinott

1B.1: Effect of a multifaceted intervention on potentially inappropriate prescribing in older patients in primary care: a cluster randomised controlled trial (the OPTI-SCRIPT study) Barbara Clyne

1A.1: Understanding unscheduled use of urgent care services by adults with chronic obstructive pulmonary disease in Bristol: a feasibility study using system dynamics Melanie Chalder

PARALLEL SESSION 1C Health promotion Room: LT3, ground floor Chair: Pat Hoddinot

16.30

PARALLEL SESSION 1B Prescribing Room: LT2, ground floor Chair: Christine Bond

PARALLEL SESSION 1A Research methods Room: LT1, ground floor Chair: Sandra Eldridge

PARALLEL SESSIONS 1 16.30-17.30

1D.4: Measuring Socio-Economic Status (SES) in Medical School Applicants Kathryn Steven

1D.3: An observational study of patient health status and use of NHS services after unintentional injury Blerina Kellezi

1D.2: Are emergency hospital admission rates for ambulatory care sensitive conditions related to patients’ perceptions of access to their GP surgery? Cross sectional analysis Robert Fleetcroft

ID.1: Delivering responsive primary care: strategies, barriers and challenges. Carolyn Tarrant

PARALLEL SESSION 1D Organisation and delivery of primary health care Room: LT4, 1st floor Chair: John Howie

1E.4: Why is scabies such a challenge in residential care homes for the elderly? A report of outbreak investigations in the South East of England Ananth Nalabanda

1E.3: Does a complex intervention by primary care nurses increase walking in 60-75 year olds? Findings from the PACE-Lift (Pedometer Accelerometer Consultation Evaluation - Lift) cluster randomised controlled trial. Tess Harris

1E.2: Older people’s experiences of self-management strategies: a systematic review and thematic analysis Clare Clarke

1E.1: Exercise for reducing fear of falling in older people living in the community: A Cochrane systematic review Arun Kumar

PARALLEL SESSION 1E Older people Room: LT5, 1st floor Chair: Carolyn Chew-Graham

1F.4: Rapid appraisal of a community based element of a revised undergraduate curriculum - The UCL Medicine in the community course Melvyn Jones

1F.3: Meta-ethnography of student and patient perspectives of undergraduate medical education in the UK general practice setting Sophie Park

1F.2: Medical Undergraduate Primary Care teaching across the UK: what is being taught? Veronica Boon

1F.1: “When Learners Become Teachers” - Exploring Educational Learning Needs of General Practice Trainees Who Teach Undergraduate Medical Students Kerry Boardman

PARALLEL SESSION 1F Education Room: 2.14, 2nd floor Chair: Anne Stephenson

Map on reverse of ticket

The Playfair Library Old College

George Square Theatre

Palliative Care: a missing global development goal Convenors: Liz Grant and Scott Murray

WORKSHOP 1 Room 2.11, 2nd floor 16.30-17.30

PROGRAMME: WEDNESDAY 9TH JULY 2014 (CONTINUED)

2C.1: What works for whom? Improving patient care by implementing NICE Osteoarthritis guidelines in General Practice. Andrew Morden

2C.2: Raising work concerns in primary care consultations for musculoskeletal conditions: a qualitative study Elizabeth L Alcock

2C.3: Effectiveness and cost-effectiveness of a group pain self-management intervention for patients with chronic musculoskeletal pain identified in primary care: COPERS randomised controlled trial. Steph Taylor

2B.1: The SIMPle study: Supporting the Improvement and Management of Prescribing for UTIs Akke Vellinga

2B.2: Developing A Taxonomy of Techniques to Harness Placebo Effects in UK Primary Care Felicity Bishop

2B.3: Tangled between confidentiality and safety: Domestic violence and child safeguarding in general practice Jessica Drinkwater

2A.1: Development of a measure of Patient Reported Experiences and Outcomes of Patient Safety in Primary Care: the PREOS-PC instrument Jose Valderas

2A.2: Examining the associations between quality of care, patient satisfaction and health outcomes oin primary care in England: A structural equation model approach Jose Valderas

2A.3: Does Participation in a Quality Improvement (QI) Programs Improve Primary Healthcare Team (PHC) Functioning? A Qualitative Study. Jyoti Kotecha

11.00

11.15

11.30

2D.3: The Impact of OPTIMAL, an Occupation-based Selfmanagement Programme, on Occupational Participation for Individuals with Multimorbidity: a Randomised Controlled Trial Susan Smith

2D.2: Resolving conflicts in the multimorbid consultation: How do GPs balance diseases, drugs and the views of other doctors? Carol Sinnott

2D.1: The 3D Study: Improving the management of patients with multimorbidity in general practice Mei-See Man

2E.3: A randomised controlled trial of a community based, group guided, self-help intervention for anxiety and depression. Jill Morrison

2E.2: Selective Serotonin Reuptake Inhibitors and congenital heart anomalies: Comparative cohort studies of women treated before and during pregnancy and their children Irene Petersen

2E.1: Antidepressant use and the risk of cardiovascular outcomes: a cohort study using a primary care database Carol Coupland

PARALLEL SESSION 2E Mental health Room: LT4, 1st floor Chair: Andrew Wilson

PARALLEL SESSION 2C Musculoskeletal Room: 2.14, 2nd floor Chair: Dan Lasserson

PARALLEL SESSION 2B Generic issues Room: LT2, ground floor Chair: Miriam Santer

PARALLEL SESSION 2A Quality and patient safety Room: LT1, ground floor Chair: Joe Kai

PARALLEL SESSIONS 2 11.00-12.30

PARALLEL SESSION 2D Multimorbidity Room: LT3, ground floor Chair: Stewart Mercer

COFFEE BREAK - meeting point for PHOCUS group: first floor outside LT4 (please pick up a coffee on the ground floor)

10.30-11.00

2F.3: Systematic review of factors that predict the emotional impact of cancer screening Laura Bedford

2F.2: The risk of cancer in primary care patients with hypercalcaemia: a retrospective cohort study using electronic records. Fergus Hamilton

2F.1: Are cancer survival differences embedded in primary care? Brian Nicholson

PARALLEL SESSION 2F Cancer Room: LT5, 1st floor Chair: Christine Campbell

‘Stratified care’: what is it and how might it help patients? Convenors: Jo Protheroe, Tom Sanders, Nadine Foster

WORKSHOP 2 Room: 2.11, 2nd floor 11.00-12.30

Appleton Tower

George Square Theatre

Yvonne Carter award and presentation: Dr Helen Atherton, University of Oxford Evidence, experiences and the future. Email for consulting with patients in general practice

09.45-10.30

Australian Association for Academic Primary Care (AAAPC) Most Distinguished Paper from the PHCRIS Primary Health Care Research Conference 2013 Dr Megan Elliott-Rudder, University of New South Wales Faculty of Medicine, Rural Clinical School Continuation of breastfeeding improves with collaborative motivational support in a cluster randomised controlled trial

George Square Theatre

Plenary 2: Chair: Brian McKinstry Professor Ruth McDonald, University of Warwick Resisting ‘Leadership’

Appleton Tower Room: 2.14, 2nd floor

09.00-09.45

The meeting this year will include an update on the successful mentoring scheme for non-medical members of SAPC, thinking about issues for early career researchers and input from participants to three developments we want to progress in the coming year: Social Media, telephone conferences and regional meetings.

Primary healthcare scientists group (PHoCuS) Convenors: Christine Bond and Sandra Eldridge

Location

08.00-08.55

Time

PROGRAMME: THURSDAY 10TH JULY 2014

2B.6: Myths, markets and models - career dissatisfaction amongst newly qualified GPs Vanessa Nash

2A.6: Withdrawing Performance Indicators: Retrospective Analysis of General Practice Performance Under the UKs Quality and Outcomes Framework. Jose Valderas

LUNCH

SAPC AGM - all members welcome

PARALLEL SESSION 3A Managing long term conditions Room: LT5, 1st floor Chair: Christian Mallen

12.15

12.30-14.00

12.55-13.55

PARALLEL SESSIONS 3 14.00-15.30

3A.2: Introducing a predictive risk tool in primary care: expectations of practitioners Mark-Rhys Kingston

3A.3: Managing patients with advanced chronic kidney disease in primary care: a qualitative study with GPs. Sarah Tonkin-Crine

14.30

3A.1: People with CFS/ME should be managed in primary care: is the NICE guideline wrong? Carolyn Chew-Graham

14.15

14.00

2B.5: Health, Medication and Use of Health Services in a Homeless Population Kirsty O’Brien

2A.5: Health needs and services for homeless people in Glasgow: a qualitative study of support and accommodation providers’ views Jamie Stewart

12.00

2C.5: Disconnects between clinical guidelines and real world practices: a qualitative exploration of the role of NICE guidelines in primary care management of low back pain Felicity Bishop

2C.4: Defining latent phenotypes of patients with hand Osteoarthritis; a fresh approach to understanding musculoskeletal conditions Dan Green

2D.6: Multimorbidity and polypharmacy in stroke. Katie Gallacher

2D.5: Exploring the clustering of chronic pain, depression and heart disease in a general populationbased cohort Blair Smith

2D.4: Common patterns of morbidity and multi-morbidity and their impact on health related quality of life: evidence from a national survey John Campbell

2E.6: What self-harm means in general practice and how it relates to suicide prevention: a qualitative study Amy Chandler

2E.5: A Systematic Review of Population-Based Risk Prediction Modelling for Dementia Eugene Tang

2E.4: Minimum clinically important differences on the Beck Depression Inventory (II): Implications for outcome assessment in depression trials Katherine Button

3B.3: Development and validation of a clinical prediction rule to identify suspected breast cancer: a prospective cohort study Kirsty O’Brien

3B.2: A Description of the EarlyCDT-Lung test in Scotland (ECLS) Trial : a case-finding method in lung cancer. Frances Mair

3B.1: Age and gender variations in cancer diagnostic intervals: analysis of data from the UK General Practice Research Database Nafees Din

PARALLEL SESSION 3B Cancer Room: LT2, ground floor Chair: David Weller

3C.3: Interleukin-6 concentrations in the urine and dipstick analyses related to bacteriuria but not symptoms in the elderly: a cross sectional study of 421 nursing home residents Pär-Daniel Sundvall

3C.2: Antimicrobial resistance in urinary pathogens among Swedish nursing home residents remains low: a cross-sectional study comparing antimicrobial resistance from 2003 to 2012 Pär-Daniel Sundvall

3C.1: Derivation and validation of an algorithm to predict hospitalisation in children presenting to primary care with acute cough and RTI: the ‘TARGET’ prospective cohort study Alastair D. Hay

PARALLEL SESSION 3C Infection Room: LT3, ground floor Chair: XXX

3D.3: How do caregivers in England describe their health-related quality of life and experiences in primary care? Evidence from 195,364 carers responding to the 2012 General Practice Patient Survey. Gwilym Thomas

3D.2: Patterns of statin prescribing for the primary prevention of cardiovascular disease in people with severe mental illness Ruth Blackburn

3D.1: The Canadian Primary Care Sentinel Surveillance Network How Canada is Leveraging Electronic Medical Record (EMR) Data for National Surveillance and Research Rachael Morkem

PARALLEL SESSION 3D Population-based practice research Room: 2.14, 2nd floor Chair: XXX

3E.3: The REACH Heart Failure Manual: Development of a complex, home-based intervention to support self care for people with heart failure. Colin Greaves

3E.2:Patients anticipated actions in the event of symptoms of a transient ischaemic attack: a qualitative study. Parker Magin

3E.1: Meta-analysis of the estimated placebo effect in primary prevention of cardiovascular mortality Kamalpreet Cheema

PARALLEL SESSION 3E Cardiovascular Room: LT4, 1st floor Chair: Richard McManus

Join us to hear how SAPC is leading work to advance primary care through education and research and to find out how you can be a part of future activities.

2B.4: Explaining variation in emergency admissions: a mixed methods study of emergency and urgent care systems Alicia O’Cathain

2A.4: Never Event in UK general Practice: A cross-sectional survey to determine the incidence and likelihood Sudeh Cheraghi-Sohi

11.45

4.V-GP (Virtual GP): This novel, innovative mobile app will revolutionise the medical school curriculum. Many of the basic skills (communication and examination) acquired by medical student’s at GP placements can be replaced by this

3.Improve the quality of Significant Event Audit’ (SEA) by involving the patient fully in the process, including inviting them to the meeting and sending them a copy of the final report. Bob Fleetcroft and Amanda Howe

2.“No worse than diabetes”: GP-led universal opt-out HIV testing to reduce stigma and prevent transmission. Bryan Jackson

1. Any quality outcome measure that requires a financial incentive to be accepted has a price but no value and should be banned. Les Toop

PARALLEL SESSION 3F Dangerous ideas soapbox Room: LT1, ground floor Chairs: Bob McKinley and Scott Murray

2F.6: GPs, patients and cancer screening Trish Green

2F.5: Building the evidence base for the early symptomatic diagnosis of cancer: systematic review of the effect of time to diagnosis on cancer outcomes Richard Neal

2F.4: Understanding symptom appraisal and help-seeking in people with symptoms suggestive of colorectal cancer: a qualitative interview study Nicola Hall

Integrating social prescribing, exercise prescribing and broaching obesity conversations into GP teaching sessions with senior undergraduate medical students - sharing challenges and opportunities Convenors: Anny Wylie and Kathleen Leedham-Green

WORKSHOP 3 Room: 2.11, 2nd floor 14.00-15.30

Room: LT1, ground floor

Appleton Tower

Location

PROGRAMME: THURSDAY 10TH JULY 2014 (CONTINUED)

Appleton Tower Appleton Tower

TEA BREAK

Special Interest Groups (SIGs) and workshops - please see pages 20-21 for SIG descriptions and pages 22-25 for workshop abstracts. Clinical excellence awards - LT1, ground floor Palliative care SIG - Room 2.12, 2nd floor Education research SIG - Room G.05, ground floor Personal care SIG - M3, 1st floor Learning disabilities SIG - Room 2.05, 2nd floor Resource-poor countries SIG - LT3, ground floor

16.00-17.30

See page 16 for meeting point and times Dynamic Earth

Optional activities - Walking with scientists - Rebus Tour Company - Rounders on the Meadows - Tour of the Scottish Parliament Building

Conference Dinner Pre-dinner drinks, dinner and ceilidh with band Modhan

19.30 to midnight

Incorporating Genomics into General Practice: is it time for primary care researchers to take the lead? Convenors: Nadeem Qureshi, Blair Smith, Imran Rafi, Joe Kai, Heather Skirton

17.45-18.45

EVENING

Trials in the real world - what can go wrong and what can be done Convenors: Ceire Costelloe, Brennan Kahan

Exploring practical approaches to maximising data quality in electronic healthcare records in the primary care setting and associated benefits Convenors: Sheena Dungey, Rosemary Tate, Willie Hamilton, Tim Williams, Chris Bates

Supporting communication in cross-cultural primary care consultations: how to implement guidelines and training initiatives Convenors: Anne MacFarlane, Chris Dowrick, Frances Mair, Kate O’Donnell, Christos Lionis

WORKSHOP 7 16.00-17.30 Room: 2.14, 2nd floor

8. Self-referral cancer assessment clinics - dangerously bypassing GP gatekeeping. Richard Neal

7. Resistant super-bugs are not a threat in primary care (they really have been around for millions of years). The healthy microbiome will protect you. Beware the damage that antibiotics may inflict on your microbiome; restoring a weakened microbiome may require a faecal transplant (with a Krapsule?). Mark Ashworth

WORKSHOP 6 16.00-17.30 Room: 2.04, 2nd floor

3E.6: Atrial Fibrillation: is screening effective in identifying patients at risk of stroke? Deborah McCahon

WORKSHOP 5 16.00-17.30 Room: 2.07, 2nd floor

3D.6: Screening for atrial fibrillation in general practice: A national, cross-sectional study of an innovative technology. Gerard Bury

WORKSHOP 4 16.00-17.30 Room: 2.11, 2nd floor

3C.6: Point-of-care rapid HIV testing in primary care is a cost-effective intervention in high endemic areas: The results of RHIVA2 cluster randomised controlled trial Andreia C Santos

6. Survival of the sickest: An innovative pilot in primary care, using health coaching by nursing and medical students to reduce the burden on acute services. Sarvesh Saini

5. A concept could be: The Asymptomatic General Check Up should be abandoned. Mark Murphy

Location

15.30-16.00

3B.6: Explaining cancer screening attendance decisions in the UK: A systematic search and meta-synthesis of qualitative evidence

3E.5: A multicentre ethnographic study of unplanned hospital admission for heart failure. Rosemary Simmonds

£1.99 app which simulates patient contact via interactive group work with clinical ‘signs and sounds’. Are GP placements redundant? Mica Skilton

3A.6: Modification of cardiovascular risk scores for use with home blood pressure measurements Sarah Stevens

3C.5: Can point-of-care rapid HIV testing in primary care increase early identification of HIV? The RHIVA2 cluster randomised controlled trial Werner Leber

3D.5: Exposure to oral hormonal contraceptives and risk of venous thromboembolism: a nested case-control study using the CPRD database. Yana Vinogradova

3E.4: Treating uncomplicated stage 1 hypertension in primary care: an expensive mistake? James Sheppard

15.15

3B.5: “The tests always came back negative you know” The influence of a negative bowel screening result among patients who subsequently develop colorectal cancer. Karen Barnett

3D.4: Does faster nicotine metabolism affect chances of stopping smoking in pregnancy? Evidence from the SNAP trial cohort Luis Vaz

3A.5: Feasibility Study: Blood pressure (BP) telemonitoring for people with previous stroke/ transient ischaemic attack (TIA) Janet Hanley

3C.4: Antibiotic treatment for acute sore throat in primary care: antibiotic choice duration and effect Michael Moore

15.00

3B.4: Symptom interpretation and decision-making processes in patients with lung or colorectal cancer Sarah McLachlan

3A.4: Prioritising the physical: a qualitative study into case finding for depression and anxiety in osteoarthritis in primary care Bernadette Bartlam

14.45

EP-2A.01: The Other Side of the Story-Maternal perceptions of safety advice and information; a qualitative approach Joanne Ablewhite

3 minute presentations plus 2 minutes Q&A

EP-2A.07: Is Proteinuria Self-testing in Pregnancy Feasible? Eleanor Brunt

EP-2A.06: Prevalence of, and socio-economic variation in ADHD in English General Practice Vibhore Prasad

EP-2A.05: What is the role of primary care in supporting the reduction of attendances to emergency departments by families experiencing minor respiratory illness in children under five? A qualitative study exploring the views of clinicians working in a UK Paediatri Leah Bowen

EP-2A.04: Home risk and protective factors for falls on one level in young children: a multicentre case-control study. Penny Benford

EP-2A.03: The challenges of meeting the health needs of unaccompanied asylum seeking children: the importance of primary care Amelia Cook

EP-2A.02: Missed opportunities to keep children safe? National survey of injury prevention activities of children’s centres Clare Timblin

ELEVATOR PITCH 2A Children, young people and women’s health Room: LT1, ground floor Chair: Jane Roberts

ELEVATOR PITCH 2 10.00-11.00

14.30-15.00

EP-2B.08: Development of an online resource to support GP appraisers in the interpretation of multisource feedback and explore the potential for standard-setting Christine Wright

EP-2B.07: Primary Care development of Summative Situational Judgement Questions Richard Knox

EP-2B.06: Undergraduate teaching in general practice: stagnation and pending decline? Alexander Harding

EP-2B.05: Compassion and Resilience in Community Based Medicine Kirsty Shires

EP-2B.04: The Educational Benefit of Mass Casualty Simulation Training Harriet O’Nions

EP-2B.03: What is the effect of brief experience of community patient care on foundation year one doctors understanding of and orientation to primary care? Sarah Eccles

EP-2B.02: Retrospective Review of Prescriptions issued by GPs in Training - the initial stages of a pilot study Richard Knox

EP-2B.01: Pilot GP training posts in a West Midlands’ prison: A positive learning experience? Jane Coomber

ELEVATOR PITCH 2B Education, ethics and governance Room: LT2, ground floor Chair: Richard Knox

EP-2C.08: Preventing loneliness and social isolation: interventions

EP-2C.07: Towards a generic PROM for primary care: Development of a model of outcomes Mairead Murphy

EP-2C.06: Modified Early Warning Scores (MEWS) to support paramedics’ decisions to transport or treat at home: a time series study Nadya Essam

EP-2C.05: Enhancing patients’ trust in the GP - Shared decision making for the contemporary older patient Joanne Butterworth

EP-2C.04: Does the availability of a South Asian language in practices improve reports of doctor-patient communication from South Asian patients? Cross sectional analysis of a national patient survey in English general practices. Faraz Ahmed

EP-2C.03: The role of GPs in the rationing of treatment for morbid obesity Amanda Owen-Smith

EP-2C.02: Effects of non-clinical factors on the estimated probability of bacterial infection and antibiotic prescribing in primary care Thomas Round

EP-2C.01: The Effectiveness of Differential Diagnosis Generators: a Systematic Review of their Efficacy and Utility Nicholas Riches

ELEVATOR PITCH 2C Patient-GP relationship, person centred care and older people’s health Room: LT3, ground floor Chair: George Freeman

ELEVATOR PITCH 2E Multimorbidity, user experience Room: LT5, 1st floor Chair: Simon Griffin

EP-2D.07: Do viral diagnoses in primary care precede and predict

EP-2E.06: Falling through the cracks: patients’ experiences of care after an injury Sarah Earthy

EP-2D.06: Clinical symptoms and signs are diagnostic of pathogen detection in children presenting to primary care with respiratory tract infection Hannah Thornton

EP-2E.07: Exploring the relationship between morbidity and patient satisfaction Charlotte Weston

EP-2E.05: “I don’t know what the Allopurinol does, but it works ...”: patient experience and understanding of treatment for gout Jane Richardson

EP-2E.04: Systematic review: Primary care clinicians’ perceived barriers and enablers to the management of type 2 diabetes mellitus Bruno Rushforth

EP-2E.03: Does patient experience of multimorbidity predict selfmanagement and health outcomes? A prospective study in primary care Peter Bower

EP-2E.02: Who sets the goal posts? Shifting the care paradigm for patients with multimorbidity: a literature review Alice Shiner

EP-2E.01: How does gender influence multimorbidity? Secondary analysis of a large nationally representative crosssectional dataset Karolina Agur

EP-2D.05: Exploring limitations and solutions when using medical databases in primary care research: a literature review Natalia Calanzani

EP-2D.04: Supporting self-care for families of children with eczema: pilot RCT of web-based intervention with health care professional support Miriam Santer

EP-2D.03: Linking electronic medical records (EMR) with administrative data: The experience of the Canadian Primary Care Sentinel Surveillance Network Richard Birtwhistle

EP-2D.02: Help4Mood: an innovative system to support the treatment of depression. Christopher Burton

EP-2D.01: Findings from The Qualitative Process Evaluation Accompanying the Asthma Internet Self Management Intervention RCT (The RAISIN Study). Kathryn Saunderson

ELEVATOR PITCH 2D Information technology, infection and sexual health Room: LT4, 1st floor Chair: Frances Mair

EP-2F.07: Questions about Efficacy and Effectiveness of Anti-depressants in those with Depression and Chronic Illness: Results from a Large Primary Care Dataset Bhautesh Jani

EP-2F.06: Patients’ perspectives of taking two antidepressants for treatment-resistant depression: a qualitative study. Carolyn Chew-Graham

EP-2F.05: Recruiting to trials of depression in the community Jill Morrison

EP-2F.04: Systematic review of the evidence for the effectiveness of statins for the primary prevention of cardiovascular disease in people with severe mental illness. Ruth Blackburn

EP-2F.03: Antidepressants and risk of epilepsy or seizures Trevor Hill

EP-2F.02: Revisiting the J shaped curve, exploring the association between cardiovascular risk factors and concurrent depressive symptoms in patients with chronic disease: Findings from a large cross-sectional study. Bhautesh Jani

EP-2G.02: More than a ‘mini’-stroke: a systematic review of fatigue, cognitive and psychological impairment following transient ischaemic attack. Grace Moran

EP-2F.01: The effectiveness of mindfulness based cognitive therapy (MBCT) in reducing anxiety among Chinese people with generalized anxiety disorder: A three-armed randomized controlled trial Samuel Wong

EP-2G.08: The moderating effect of deprivation and quality of care upon healthy ageing in older people with musculoskeletal pain Gwydion Rhys

EP-2G.07: Experience of an online self-report registry for patients self-managing/monitoring their oral anticoagulation therapy. Deborah McCahon

EP-2G.06: An Evaluation of the Effectiveness of the New Medicine Service in Community Pharmacy in England Matthew J Boyd

EP-2G.05: A feasibility randomised controlled trial looking at the effect on lung cancer diagnosis of giving a Chest X-Ray to smokers aged over 60 with new chest symptoms - the ELCID trial Richard Neal

EP-2G.04: An Ethnographic Account of the Benefits and Challenges of Providing NHS Health Checks at Community Outreach Events. Jeremy Horwood

EP-2G.03: COPD admissions in an urban environment: factors influencing length of stay Timothy Harries

ELEVATOR PITCH 2G Pot pourri 2 Room: 2.11, 2nd floor Chair: Blair Smith

ELEVATOR PITCH 2F Mental health Room: 2.14, 2nd floor Chair: Tony Kendrick

After plenary move to Appleton Tower

George Square Theatre

09.00-09.45

Plenary 3: Chair: Scott Murray Professor Frede Olesen, Aarhus University, Denmark The doctor drug. New research results and new challenges for research and clinical work in general practice TEA BREAK

Location

Time

PROGRAMME: FRIDAY 11TH JULY 2014

4C.4: Facilitation of GP Trainer revalidation in the West Midlands Rodger Charlton

4B.4: Healthy behaviours reduce 10-year incidence of diabetes: a population cohort study Simon Griffin

4A.4: Clinically-led commissioning: what ‘added value’ do GPs bring to Clinical Commissioning Groups? Kath Checkland

CLOSING SESSION Chairs: Karen Fairhurst and Christine Campbell Prize giving SAPC ASM 2015 presentation - University of Oxford Packed lunch to take away

GOLF - Sign up at the registration desk. £21 for 9 holes including club hire. Meet by registration desk at 13.30

12.15

12.35-13.00

13.30

13.00

Session sponsored by BioMed Central Ltd

4D.3: Implementing selfmanagement for long-term conditions: a systematic review of implementation studies Hilary Pinnock

4C.3: Enabling senior medical students to engage with social and exercise prescribing, and address the lifestyle risk factors associated with obesity: what needs to be in place? Kathleen Leedham-Green

4B.3: Do Patient’s Characteristics Predict Failure to Receive Care for Type 2 Diabetes? Luke Mounce

4A.3: Non-alcoholic Fatty Liver Disease: What are GPs doing about it? A review of current practice in two London boroughs ahead of implementation of a novel referral pathway. Ruth E Gailer

12.00

4D.4: “The best thing since sliced bread”: patients’ perceptions of self monitoring of warfarin Patricia Apenteng

4D.2: Does self-management support increase the risk of mortality amongst patients with COPD? A systematic review of the evidence Kate Jolly

4C.2: Longitudinal Integrated Clerkships in the UK? Results of a four-year pilot in Exeter. Alexander Harding

11.45

4B.2: Diagnosis and monitoring of chronic kidney disease in primary care - systematic review and meta-analysis of bias of glomerular filtration rate estimating equations. Daniel Lasserson

4D.1: Understanding support for patient self-management: a proposed new taxonomy to help identify and design interventions - The SelfManagement Support Taxonomy. Steph Taylor

PARALLEL SESSION 4D Self-management Room: LT3, ground floor Chair: Amanda Howe

those obtained in secondary care: a temporal trend comparison study Michelle S. Toleman

4A.2: Can telephone triage systems help to improve satisfaction among patient groups who tend to report lower satisfaction? Raff Calitri

PARALLEL SESSION 4C Education Room: 2.14, 2nd floor Chair: Jill Morrison

EP-2C.09: Developing a multidisciplinary rehabilitation intervention following hip fracture (FEMuR): Integration of findings from a realist review, health professional survey and patient/ carer interviews Nefyn Williams

and outcomes Karen Windle

4C.1: The use of a Simulated General Practice placement for undergraduate teaching: Student perceptions of the benefits and drawbacks. Russell Hearn

Room: LT2, ground floor Chair: Brian McKinstry

Session sponsored by BioMed Central

PARALLEL SESSION 4B Managing long term conditions

EP-2B.10: Completion of training in Immediate Care by GPs in Ireland 2002-2013 Gerard Bury

EP-2B.09: ‘It’s my job now rather than my vocation’; What threatens the health of medical professionalism in the NHS? Sharon Spooner

4B.1: A Mindfulness-based programme for patients with chronic pain: Exploring the barriers and facilitators to engagement. Fathima Marikar Bawa

PARALLEL SESSION 4A Organisation and delivery of primary health care Room: LT1, ground floor Chair:Chris Salisbury

COFFEE BREAK

EP-2A.09: Screening for chlamydia related bacteria to prevent miscarriage or preterm birth: proof of principle study Pippa Oakeshott

EP-2A.08: Blood pressure monitoring in pregnancy: patient and staff perspectives: a qualitative study Lisa Hinton

4A.1: Community participation in primary care - an analysis of levers and barriers to implementation using Normalisation Process Theory Anne MacFarlane

11.30

PARALLEL SESSIONS 4 11.30-12.30

11.00-11.30

Time

4E.4: Why telling parents “antibiotics don’t work on viruses” doesn’t reduce expectation for antibiotic treatment for children with RTIs: a qualitative study Christie Cabral

4E.3: Parental home safety practices and risks for childhood poisonings: findings from a multi-centre case-control study. Gosia Majsak-Newman

4E.2: Parent’s perceptions of barriers to and facilitators for keeping children safe at home; a qualitative study Joanne Ablewhite

4E.1: Financial incentives to providers to improve smoking cessation in pregnancy and breastfeeding outcomes: acceptability and perceived consequences. Pat Hoddinott

PARALLEL SESSION 4E Children and young people Room: LT4, 1st floor Chair: Denise Kendrick

EP-2E.08: A qualitative study of service users’ experiences of alcohol treatment and support services: an Evaluation of the Alcohol Intensive Case Management Service in Nottingham City Jacqueline Mhizha-Murira

4F.4: Different dyings: living and dying with cancer, organ failure and physical frailty Marilyn Kendall

4F.3: Integrating Palliative Care into national health systems in four African Countries: developing models of best practice Liz Grant

4F.2: Anticipatory prescribing at the end-of-life in South Edinburgh care homes Anne Finucane

4F.1: Advanced Care Planning decisions amongst Irish Care Home Residents. Does it help to ‘Think Ahead’? Joseph Marry

PARALLEL SESSION 4F Palliative care Room: LT5, 1st floor Chair: Marilyn Kendall

EP-2F.09: Case management for older people with depression - a qualitative study: ‘Reducing the blind spots’ Carolyn Chew-Graham

EP-2F.08: What can we do for patients we don’t know how to help? Could social prescribing be the answer? Alexis Kilgarriff-Foster

Appleton Tower

Appleton Tower Room: LT4, 1st floor

Realising the potential of the Clinical Practice Research Datalink (CPRD) Convenors: Roger Jones, BJGP, John Parkinson and Antonis Kousoulis, CPRD

WORKSHOP 8 Room: 2.11, 2nd floor 11.30-12.30

Appleton Tower

EP-2G.09: Engaging GPs in clinical Location trials: Barriers and facilitators encoun- tered in the Thyroid Hormone Replacement for Subclinical Hypothyroidism (TRUST) study Carol Sinnott

KEYNOTE SPEAKERS Wednesday 9th July 12.45 - 14.00 Opening Session, George Square Theatre

Achieving Universal Health Coverage - the role of primary care research Professor Sir Andy Haines, London School of Hygiene and Tropical Medicine Professor Sir Andy Haines is Professor of Public Health and Primary Care at the London School of Hygiene and Tropical Medicine. He was Director of the London School of Hygiene & Tropical Medicine from 2001October 2010. In that role he was responsible for the management of over 1000 staff and 3700 postgraduate students. He was previously Professor of Primary Health Care and Head of the Department of Primary Care and Population Sciences at University College London, and worked part-time as a general practitioner in North London for many years. Before that he was a consultant in epidemiology at the Medical Research Council Epidemiology and Medical Care Unit. He was also formerly Director of Research & Development at the National Health Service Executive, North Thames and a member of the Council of the Medical Research Council. He has worked internationally, including in Nepal and the USA. He chaired a Task Force on Health Systems Research for WHO which reported in 2005 and the Scientific Advisory Panel for the 2013 WHO report - ‘Research for Universal Health Coverage’. He sits on many national and international committees. He was a member of the UN Intergovernmental Panel on Climate Change for the 2nd and 3rd assessment exercises and is review editor for the health chapter in the 5th assessment. He has published many papers in high impact journals on topics such as primary care, health systems research and the relationship between environmental change and health.

17.45 - 18.30 The Inaugural Helen Lester Memorial Lecture, George Square Theatre

Motherhood and Mental Illness - thirty years of families in South London Professor Debbie Sharp, University of Bristol Professor Helen Lester was an inspirational GP and academic who made a difference to the lives of many people. Helen challenged us all to think more creatively about many things - but especially about mental health. In this first of a new series of annual public lectures, Professor Debbie Sharp will pick up the baton Helen has left for us and consider how important it is for primary care to identify mental illness in women of childbearing age. Professor Debbie Sharp is Professor of Primary Health Care in the Centre for Academic Primary Care at the University of Bristol. She was previously Lecturer and then Senior Lecturer at the United Medical and Dental Schools of Guy’s and St Thomas’ in the Department of General Practice, and Honorary Senior Lecturer at the Institute of Psychiatry. In 1985 she obtained one of the first Mental Health Foundation GP Research Training Fellowships through which she completed a PhD on emotional disorders associated with childbirth in a cohort of women in South London, supervised by Professors Michael Shepherd and David Morrell. It was during her PhD that she met the late Channi Kumar who encouraged her to follow up the cohort of women from her PhD. Thus the South London Child Development Study was born She took up the foundation chair in Primary Health Care in Bristol in 1994, the first woman to be appointed to a substantive chair in medicine in Bristol and built up a world-class department over the next sixteen years. Her major research interest is in primary care mental health and she has been involved in numerous studies in this area over the last three decades. She has continued her interest in childbirth related emotional disorders and has recently completed a trial comparing antidepressants and health visitor listening visits for women with postnatal depression. Other research interests include early diagnosis of cancer, childhood obesity and complementary and alternative medicine. Since standing down as head of department she has become head of the School of Clinical Academic Training at HESW Severn.

Thursday 10th July 09.00 - 09.45 George Square Theatre

Resisting ‘Leadership’ Professor Ruth McDonald, Warwick Business School Ruth McDonald is Professor of Governance and Public Management at Warwick Business School. She is a political science graduate and a former UK National Health Service hospital finance director. In 2007/8 she spent a year at UC Berkeley as a Harkness Fellow in Health Care Policy and Practice, comparing pay for performance programmes in the UK and California. Ruth is a member of the executive committee of the European Sociological Association and co-chair of its Sociology of Professions Research Committee. She has also held posts at the Universities of Nottingham, Manchester, Liverpool and Leeds. Much of her research concerns simple ‘solutions’ to complex problems. In recent years much of this has concerned financial incentives for quality in the UK and beyond and she has published widely on the subject. She has also recently published an analysis of the doctors appearing on Radio 4’s Desert Island Discs. None of these are GPs! She has provided advice to a range of bodies including the Department of Health and The Health Foundation in the UK, the Indian Ministry of Health and Family Welfare and the Research Council of Norway.

Friday 11th July 2014 09.00 - 09.45 George Square Theatre

The doctor drug. New research results and new challenges for research and clinical work in general practice Professor Frede Olesen, Aarhus University, Denmark Frede Olesen graduated as a MD in 1973 and gained his GP certification in 1976 and his Doctor of Medical Science in 1991. Between 1992 and 2011 he was administrative and scientific director of the Research Unit for General Practice at Aarhus University. Since 1996 he has been professor at Aarhus University. He has participated in many CME-courses and courses in research methodology and several short-term visits to universities and general practices in Europe, Canada and the US. In 1999 he was visiting professor at the Chinese University of Hong Kong. He has had more than 250 scientific publications in national and international medical journals and many scientific presentations, reviews and chapters in CME-journals, books and national medical journals. He has been a referee on national and international journals, an assessor for several PhD and doctoral theses and member of appointments committees for professorship positions. His former and present memberships include: The Danish College of General Practitioners (1988-92 chairman); several Danish committees for health care planning and research including the Danish Medical Research Council; the UEMOdelegation for European unions of GPs, the European Working Party on Quality in Family Practice (EQuiP). 1995 - 1998 founding president of The European Society of General Practice/Family Medicine and vice president of the World Organization (WONCA). Since 2006 chairman of the Danish Cancer Society. He has received several research awards including the most prestigious award of the Danish Medical Association. Frede is Honorary Fellow of the Royal College of General Practitioners and honorary life member of the World Organisation of General Practitioners/Family Physicians.

SOCIAL PROGRAMME

Wednesday 9th July 2014 Conference drinks reception 18.45 to 20.00 Playfair Library, Old College, University of Edinburgh Drinks and light bites will be served at Old College. Piper: Iain Murray

Thursday 10th July 2014 Optional activities: Meeting point and time

Activity time

Activitiy

17.30 Appleton Tower foyer 17.30 Appleton Tower foyer 18.10 Scottish Parliament by Dynamic Earth (see map on inside back cover)

17.45-18.45 17.45-18.45 18.15-19.15

Walking with Scientists - a Rebus Tour Rounders on the Meadows A guided tour of the Scottish Parliament Building

Conference dinner 19.30 for 20.00 until midnight 19.30 to 20.00 Pre-dinner drinks and time to visit the Rainforest, 4Dventure and the Polar Regions galleries downstairs at Dynamic Earth 20.00 Dinner in the stratosphere followed by ceilidh in the Ozone, a sound proofed area with wonderful views up to Arthurâ&#x20AC;&#x2122;s Seat. Tickets available at the registration desk ÂŁ55 per person.

Friday 11th July 2014 Golf: Why not finish the conference with a swing! We have arranged Tee off times for up to 16 golfers at Braid Hills Golf Course, a very scenic course overlooking Edinburgh. It is only 15 minutes from the conference venue and on the road south and to the airport. The 9 holes will take from 2-4.30pm and the round and golf club hire (men, women and left handers) will cost ÂŁ21. Sign up at the registration desk.

PEER REVIEW Programme Committee chairs: Christine Campbell and Brian McKinstry, Edinburgh Independent Peer Reviewer Panel chairs: Denise Kendrick, Nottingham and Andrew Wilson, Leicester This year 332 abstracts including 12 workshops and 9 dangerous ideas were submitted. All abstracts were peer reviewed by at least two reviewers. Reviewers scored each abstract on six areas using a five-point scale. The areas considered when scoring were: the problem, the approach, the findings, consequences and overall impressions.

to mean score, and this ranking was then used as a starting point for discussion about acceptance by the Programme Committee, but it was not the sole criterion. The programme consists of 8 workshops, 150 elevator pitches and 113 oral presentations including 2 distinguished abstracts selected for presentation in the prize plenary session on Wednesday. The programme was compiled by the Programme Committee and any queries or appeals were considered by the Programme Committee chairs and Peer Review Panel chairs.

A mean score was obtained for each abstract using the scores assigned to it by each peer reviewer. If two scores varied by more than 8 points the abstract was assessed by a third reviewer. The abstracts were ranked according

Peer reviewers were drawn from the SAPC membership following nominations by Heads of Department and a request for volunteers with experience of reviewing. The Programme Committee thanks the peer reviewers for their valuable contribution:

Asmaa Abdelhamid

David Fitzmaurice

Sara Macdonald

Joanne Reeve

Sue Ashby

Susannah Fleming

Maggie Maggie Bartlett

Ignacio Ricci Cabello

Opeyemi O Babatunde

Katie Gallacher

Christian Mallen

Matthew Ridd

David Blane

Simon Gay

Michelle Marshall

Louise Robinson

Jenni Burt

Daniela Goncalves

Anna Matthews

Chris Burton

Gus Hamilton

Lorraine McFarland

Raff Calitri

Willie Hamilton

Robert McKinley

Christine Campbell

Andrew Hardcastle

Brian McKinstry

Paul Campbell

Richard Hayward

Liz Mitchell

Melanie Chalder

Alastair Hayward

Michael Moore

Carolyn Chew-Graham

Emma Healey

Andrew Morden

Will Coppola

Alyson Huntley

Helen Morris

Carol Coupland

Catherine Hyde

Sara Muller

Sam Creavin

Judith Ibison

Richard Neal

Jacquelline Crowther

Bautesh Jani

Barbara Nicholl

Margaret Cupples

Robert Jones

Pauline Ong

Fiona Walter

Christopher Dowrick

Kelvin Jordan

Diane Owen

David Weller

Kate Dunn

Nada Khan

Rupert Payne

Nefyn Williams

Hazel Everitt

Lily Lai

Sarah Purdy

Andrea Williamson

Karen Fairhurst

Rosa Lau

Cath Quinn

Fiona Wood

Bruno Rushforth Miriam Santer Niroshan Siriwardena Sarah Smithson Tim Stokes Ann Sunderland Athina Tatsioni

Sarah Tonkin-Crine Danielle van der Windt Oliver van Hecke Akke Vellinga Nicola Walker

CONFERENCE INFORMATION Conference APP

Publication of Abstracts

This is new to the SAPC annual conference this year and we hope you will enjoy using it. Download it from the APP store on to Apple and android devices. The full programme and all the abstracts (including the Elevator Pitches) can be found on the APP along with most of the information in this book, you can also access the evaluation form and your own attendance certificate.

At present there are no plans to publish the abstracts as a supplement to a journal.

In addition you may be interested in the Edinburgh Rewards Card and APP for conference attendees (see details at the end of this section).

Attendance Certificates If you wish to receive an attendance certificate you can either create one on the Conference APP (see first item above) or complete the Survey Monkey evaluation form and request one after the conference. The link will be emailed to conference delegates on Monday 14th July: www.surveymonkey.com/s/MKJGB7N

Pre-book Workshops, Special Interest Groups (SIGs) and Thursday evening optional activities

Edinburgh Rewards Card and App

Places may be pre-booked at the registration desk for Workshops, Special Interest Groups (SIGs) and Thursday evening optional activities: walking with scientists, rounders and a tour of the Scottish Parliament. See the programme pages and social event pages.

Working in co-operation with the Edinburgh Convention Bureau, we are delighted to offer you a rewards card which can be used in the various bars, restaurants, shops, and with the cityâ&#x20AC;&#x2122;s tour operators. In addition to the Rewards Card and map, there is a free app that can be downloaded via www.yogobogo.com - select Scotland, then Edinburgh, then Edinburgh Rewards.

Speaker Preview Room Room M2 sections B&C on the first floor. Presentations should be taken to the Speaker Preview room at least one hour before the start of the session in which you are presenting so that they can be remotely transferred into the appropriate session room. If you have already submitted your presentation by email you may review and update it in this room.

EXHIBITORS AND SPONSORS We are grateful to the following organisations for their financial support:

Parallel Oral and Elevator Pitch Sessions Session chairs and speakers are requested to go to the room where the session will take place at least 10 minutes before the start time to meet each other and familiarise themselves with the audio-visual equipment. A member of the conference staff team will be present during the session to assist. Time slots: Oral presentations in parallel sessions: 10 minutes presentation and 5 minutes Q&A Elevator pitch presentations: 3 minutes presentation and 2 minutes Q&A, the next speaker should get ready to present during the Q&A of the previous speaker. Please keep to time!

BioMed Central Ltd - www.biomedcentral.com Informatics in Primary Care Journal www.bcs.org

And to our exhibitors

Clinical Practice Research Datalink - www.cprd.com

Internet and Email Free WiFi is available at the University, please request a password at the registration desk. If you have access to Eduroam you should be able to use this.

The Medical and Dental Defence Union of Scotland www.mddus.com

Press Attendance - Notice to Presenters and Authors

Oxford University Press - www.oup.com Browse a selection of books at our unmanned display table near the registration desk.

There may be journalists attending the conference. Journalists will be asked by the organisers to identify themselves to presenters and presenters should notify the journalist if they do not wish their work to be reported. Presenters may also like to let the chair of their session know if they do not wish their work to be reported and ask journalists to identify themselves in the session.

SHARE Register - NHS Research Scotland www.registerforshare.org Scottish Primary Care Research Network www.sspc.ac.uk 19

SPECIAL INTEREST GROUPS AND MEETINGS Thursday 10th July, 08.00 to 09.55 Breakfast meeting: Primary Health Care Scientists (PHoCuS) Group

The current situation regarding applying for local (employer based) clinical excellence award in England

Room: 2.14, 2nd floor

This group is primarily for those who are non-medical. The meeting this year will include an update on the successful mentoring scheme for non-medical members of SAPC, thinking about issues for early career researchers and input from participants to three developments we want to progress in the coming year: Social Media, telephone conferences and regional meetings. We are also trying to create more opportunities for PHoCuS members to meet each other at other times during the conference, for example over coffee on Thursday morning and at the conference dinner. Come to the meeting to hear more!

F The current situation regarding ‘Distinction Awards’ in Scotland

Thursday 10th July, 16.00 to 17.30 Clinical Excellence Awards Meeting/Workshop Convenors: Tony Avery, Nottingham, John Campbell, Exeter and Graham Watt, Glasgow Room: LT1, ground floor In 2004, after many years of negotiation, senior academic GPs in England, Wales and Scotland became eligible to apply for clinical excellence awards. Since this time, senior academic GPs have done relatively well in the national clinical excellence award scheme in England and Wales, reflecting the tremendous work that colleagues have done (over and above expectations) in terms of the development and delivery of high quality clinical services, and research in educational activities that are of direct benefit to the NHS. Access to local (employer based) awards has been more patchy and since the NHS changes, introduced in April 2013, there are now only one or two places in the country where senior academic GPs are able to apply for these awards. Added to this, the whole clinical excellence award scheme is likely to change in coming years as a result of recommendations from the Doctors and Dentist Review Body, and on-going discussions between the Department of Health (in England), equivalent organisations in the devolved nations, and the British Medical Association. The SAPC Heads of Departments group have been working with a wide range of organisations over recent years to try to ensure continuing access to local and national clinical excellence awards for senior academic GPs, and favourable arrangements in any new system. In this CEA workshop Tony Avery, John Campbell and Graham Watt will bring colleagues up to date with the latest developments regarding clinical excellence awards. Depending on the interests of colleagues attending the workshop will cover the following topics:

The current clinical excellence award scheme an eligibility for, and advice on, applying for national awards in England and Wales

Our negotiations around supporting senior academic GPs who currently hold local clinical excellence awards in England

F Our negotiations around setting up a new ‘local’ clinical excellence award scheme for senior academic GPs in England F

Our latest thoughts on what the new clinical excellence award scheme might look like in the UK, including eligibility for senior academic GPs Planning for the future: in particular, we need ‘early’ career academic GPs to help shape our negotiating positions, and to contribute to helping to ensure that any new clinical academic award scheme is favourable for senior academic GPs

This meeting/workshop is particularly aimed at clinical academic GPs.

Education Research Convenor - Sophie Park, University College London Room: G.05, ground floor We would be delighted if you are able to join us for this year’s SAPC Education Research SIG. This meeting is open to all delegates who have an interest in primary care medical education research - including experiential-user and/or methodological expertise. There are a number of exciting education research projects in primary care, either in progress or near completion this year. We will use the SIG to a) strengthen collaborative links between colleagues interested in primary care education research b) share examples of on-going education research, including informal discussion and feedback c) discuss successful / completed research projects from 2013-14 and ways in which their findings might be developed further and d) consider ways to maximise the website utility for SIG members throughout the year.

Learning Disabilities Convenor: Umesh Chauhan, Manchester Room: 2.05, 2nd floor This month (July 2014) the Department of Health, NHS England and Public Health England will be publishing an update on the Action Plan to take forward the recommendations of the Confidential Inquiry into premature deaths of people with learning (intellectual) disabilities. This workshop aims to share and discuss the findings of the Confidential Inquiry, particularly those of relevance to Primary Care, and address a key issue identified by the Inquiry - the lack of any internationally comparable mortality data about people with learning disabilities. The workshop will be a mix of the presentation of key information, followed by small and large group discussions. Group discussions will focus on developing academic and/or practical strategies to tackle the situation in which we, like many other countries internationally, are not able to identify from national data the age and cause of death of people with learning disabilities.

This year we have 2 projects to spark our thinking on what patient-centred care looks like. Chris Salisbury will outline the proposed research from Bristol, Oxford, Edinburgh and Exeter on email consultations. Maxine Jones will introduce the relationship-based Nuka System of Care and discuss implications for primary care continuity and interprofessional boundaries.

Primary care in resource poor countries Convenor: Brian Nicholson Room: LT3, ground floor This SIG aims to support those involved in primary care in low-income countries. We welcome everyone with an interest to participate but invite those who would like to make a brief presentation of their work (research, teaching or other initiatives) at the meeting to contact Brian Nicholson brian.nicholson@phc.ox.ac.uk.

Palliative Care - Calling new palliative care researchers Convenor: Scott Murray, Edinburgh Room: 2.12, 2nd floor This is an opportunity to learn about exciting opportunities to take forward palliative care research internationally. We will also present the â&#x20AC;&#x153;toolkitâ&#x20AC;? we have recently published to help the development of palliative care in the community internationally. It is being translated into various European countries to promote palliative care there. Opportunities to promote in country developments and collaborative research will be identified. So welcome to novices and those more experienced in researching the problems that people have towards the end of life. See this link for the toolkit http://www.eapcnet.eu/Portals/0/ Organization/Primary%20care/ToolkitPrimaryCare_2014.pdf

Personal Care Convenors: Joanne Reeve, Liverpool and Chris Salisbury, Bristol Room: M3, 1st floor The Personal Care SIG meets to discuss topics related to patient-clinician relationships, communication and values. With more time and less formality than the regular parallel sessions, our sessions aim to stimulate new ideas and deeper thought in areas which are complex and often hard to fund.

WORKSHOP ABSTRACTS Workshop 1

Workshop 2

Wednesday 9th July, 16.30 - 17.30, Room 2.11, 2nd floor

Thursday 10th Jul, 11.00 - 12.30, Room 2.11, 2nd floor

Palliative Care: a missing global development goal

‘Stratified care’: what is it and how might it help patients?

Convenors: Dr Liz Grant and Professor Scott A Murray, Primary Palliative Care Research Group, The University of Edinburgh, Medical School, Teviot Place, Edinburgh, Makerere University, Kampala, Uganda and the African Palliative Care Association, Kampala, Uganda

Convenors: Jo Protheroe, Tom Sanders and Nadine Foster, Keele University

Aim and intended outcome/educational objectives To share the unmet need for global palliative care in the face of double and triple burdens of non-communicable and communicable diseases especially in low and middle income countries. To explore ways in which Family Medicine can contribute to integrating palliative care into national health systems. Format Short case studies of success in overcoming barriers to delivering palliative care globally followed by an interactive audience discussion and small group work on the two questions below. Content Through two questions we will look at some of the critical issues 1) Why has palliative care sat outside the system and how can it be better integrated into the health system? Closing the inequitable gaps in global, regional and national patient outcomes - such as those embodied in the health-related Millennium Development Goals is dependent on the strategic strengthening of the delivery capacity of health systems. The UN review analysis of the MDGs in 2010 identified silo working as a major obstacles to achieving the goals. The silo activity in prevention and management of diseases has been mirrored in the silo nature of the delivery of care at different stages of illness creating dysfunctional and disjointed services disadvantaging both patients and service providers. Audience will be asked to work in small groups to put forward five key reasons for the lack of comprehensive palliative care, and five ways in which integration might across in five continents. 2) What new and old innovations are viable for making a difference to care? The new public health approach to palliative care includes the paradigm of palliative care, ( the way that death, dying and quality living toads death are conceptualised in societies); the architecture of palliative care within a country (policies, workforce education; information; medical products, vaccines and technologies; financing; and leadership and governance) and the system delivery of such care at local level - the service delivery ( pain and other symptom management, patient referral and pathways to care, staffing and skills, resources and community engagement). A blue skies thinking time - Audience will be asked in groups to identify and share ideas on innovations Audience Up to forty people interested in palliative care globally and the way in which family medicine/general practice can support integration into the health system in different continents.

Aims 1. To discuss the acceptability of stratified care using the example of musculoskeletal pain conditions 2. To identify perceived barriers and practical solutions to the use of stratified care 3. To discuss the appropriateness of matching treatment options specifically to subgroups of patients based on their risk profiles Intended outcomes Participants will gain understanding of stratified care using the example of musculoskeletal pain conditions, and the added value of a ‘prognostic’ alongside a ‘diagnostic’ healthcare model for primary care clinical practice. Format The workshop will include an initial short presentation by the facilitators, followed by an audio recorded group discussion. Content Stratified care involves targeting treatment to subgroups of patients based on their key characteristics such as prognostic factors, likely response to treatment or underlying mechanisms. It aims to tailor therapeutic decisions in ways that maximise treatment benefit, reduce harm and increase healthcare efficiency by offering the right treatment to the right patient at the right time. This workshop will start with a short presentation (15 minutes) by the facilitators of a stratified care approach that has been developed, validated, tested and implemented using the example of low back pain. We will then have a facilitator-led group discussion with all workshop participants about the potential for stratified care to be used for other groups of patients. We will review how patients might be classified into low, medium and high risk subgroups. Each case scenario will provide an example of what a patient at ‘low’, ‘medium’ and ‘high’ risk might look like and what treatment options might be most appropriate for their particular condition and risk profile. Participants will be invited to ask questions, and discuss each case scenario. Participants will be divided into two or three groups (of 8-12 depending on numbers) who will discuss three questions in relation to the three risk case scenarios (eg. low, medium and high risk): 1) the acceptability of stratified care for the management of musculoskeletal patients, 2) the practical barriers and solutions to using stratified care in practice, and 3) the matching of appropriate treatment options on the basis of patients’ risk profiles. The group discussion will last approximately 45 minutes. The groups will reconvene to identify the key learning points and summarise the messages derived from both discussions (30 minutes). The group discussions will be audio recorded. Intended audience The workshop is aimed at primary care researchers and clinicians; GPs and other healthcare professionals such as rheumatologists, physiotherapists, and nurses working in primary or secondary care. Musculoskeletal pain is used as an example only, but the principle of stratified care has broad applicability to many clinical areas.

Workshop 3

Workshop 4

Thursday 10th July, 14.00 - 15.30, Room 2.11, 2nd floor

Thursday 10th July, 16.00 - 17.30, Room 2.11, 2nd floor

Integrating social prescribing, exercise prescribing and broaching obesity conversations into GP teaching sessions with senior undergraduate medical students sharing challenges and opportunities

Supporting communication in cross-cultural primary care consultations: how to implement guidelines and training initiatives

Convenors: Dr Ann Wylie and Dr Kathleen Leedham-Green, King’s College London Aims and intended outcome/educational objectives The workshop aims to look at these newer requirements for clinical practice and teaching; how they are being integrated into education; and, drawing on evidence based protocols, progress, barriers and the potential proactive role of students. By the end of the workshop participants: F will be familiar with recent literature and intervention evidence of effective approaches to behaviour change F will be aware of relevant Tomorrow’s Doctors requirements for medical education F will have shared experiences and challenges F will have discussed and planned teaching sessions and student involvement Format 1. Expectations of this session 2. An opening introduction to both social and exercise prescribing within clinical practice and medical education - what are our collective experiences and concerns? 3. where is the evidence and what resources support both clinical practice and medical education? A critical review and feedback 4. Working on plans that involve the role of students eg developing protocols, doing audits, planning evaluation etc. 5. Summary and take home messages Content BJGP recent papers and experiences at KCL School of Medicine including approaches to teaching, resources we have developed and our research findings. Intended audience Healthcare professionals and educators who have responsibilities for senior medical students with or without a curriculum development role F Healthcare professionals, researchers and educators with an interest in social prescribing F

Convenors: Anne MacFarlane, University of Limerick, Christopher Dowrick, University of Liverpool, Frances Mair, University of Glasgow, Kate O’Donnell, University of Glasgow, Christos Lionis, University of Crete Background Unreliable communication due to language or cultural differences has a negative impact on cross-cultural primary care. Migrant communities are particularly sensitive to this problem, still neglected in some primary care settings in Europe. While many guidelines and training initiatives exist to address this issue their implementation tends to be ad hoc. This reflects a well-documented translational gap between evidence and practice and is part of the wider problem of implementing guidelines and educational and quality interventions in routine practice. There is an urgent need to find effective strategies to bridge this translational gap and implement best practice in cross-cultural primary care consultations. Aim and intended outcome The aim of the workshop is to enhance understanding of how to implement effective communication in cross-cultural primary care consultations, using the example of RESTORE - an EU FP7-funded study to improve implementation strategies across Europe. The intended outcomes are: to expand knowledge of cutting-edge qualitative research into implementation strategies for effective communication in cross-cultural primary care consultations; to consider and evaluate culture as an important determinant of an effective primary care consultation; and to enhance participants’ confidence to undertake similar implementation-oriented research journeys. Format Scene setting and case studies (45 minutes); facilitated small group discussion and plenary review (45 minutes). Content We will provide a brief explanation of our field work methods. The study took place in five primary care sites, in Austria, Crete, England, Ireland and Netherlands. We recruited stakeholders (migrant service users, general practitioners, primary care nurses, practice managers and administrative staff, interpreters, cultural mediators, service planners, and policy makers). We conducted a mapping exercise to identify relevant guidelines and training initiatives. Then we initiated dialogues with stakeholders, brokered by Participatory Learning and Action (PLA) methods, around Normalization Process Theory’s four key constructs (coherence, cognitive participation, collective action and reflexive monitoring). Using a PLA process, stakeholders selected a single guideline or training initiative for implementation in their local setting. We investigated and supported the implementation journeys for the five selected interventions, including local and cultural adaptation, policy impact, delivery and evaluation. We will use three case studies to show how we chose, adapted, delivered and evaluated one guideline (Crete), one training initiative (England) and one combined guideline and training initiative (Ireland): all designed to improve communication in cross-cultural general practice consultations. These will be followed by facilitated small groups (n=3-5) during which participants will have the opportunity to consider the implications of RESTORE methods and findings for their own primary care research contexts, with focus on the impact on general practice performance and quality of primary care services. The workshop will finish with plenary feedback and proposals for next steps. 23

Workshop 5

Workshop 6

Thursday 10th July, 16.00 - 17.30, Room 2.07, 2nd floor

Thursday 10th July, 16.00 - 17.30, Room 2.04, 2nd floor

Exploring practical approaches to maximising data quality in electronic healthcare records in the primary care setting and associated benefits

Trials in the real world - what can go wrong and what can be done

Convenors: Sheena Dungey, CPRD / University of Sussex; Rosemary Tate, University of Sussex; Willie Hamilton, University of Exeter; Tim Williams, CPRD; Chris Bates, TPP Aim and intended outcome/educational objectives The use of electronic healthcare systems for recoding patient treatment history is well established in GP practices across the UK and the potential benefits of using such systems within the primary care setting are numerous. Electronic healthcare records (EHR) provide a complete picture of patient care over time which not only affords the opportunity to improve patient care directly through effective monitoring and identification of care requirements but also offers a unique platform for both clinical and service -model research essential to the longer term development of the health service. Exploiting the information contained within EHR data will be key to addressing major challenges to public health both nationally and globally, ultimately offering a means of maximising efficiency and equality in care. There are, however, significant challenges in using EHRs effectively and particularly in ensuring the quality of data recorded. Incorrect or missing data could render records as useless or indeed misleading such that conclusions drawn from the data could have a negative impact. Amongst other difficulties, recording data can be time consuming to the extent of conflicting with the GP’s primary focus of patient consultation in an already time-constrained environment. Understanding the requirements of and the demands upon GPs must be central to addressing the issue of data quality within EHRs. The aim of this workshop will be to explore issues of data quality in primary care EHRs from the perspective of different users of GP data but with particular focus on how and why data is recorded in the first instance. The intended outcome is a furthered understanding of both the challenges and the direct benefits to GPs of ensuring high quality data with a view to establishing a workable approach to recording data and maximising benefits to all users of EHRs. Format The workshop will be held as a structured discussion focussed around questions addressing key areas of the topic at hand. The discussion will be initiated by a panel comprising four speakers expert in the field of EHRs though with varied backgrounds including general practice, epidemiology and IT. Panel members will put forward their opinions on each question before the discussion is opened to the floor. Content The discussion will be focussed around three questions. 1. What are the areas of focus for data quality in EHRs? 2. What are the potential benefits to GPs on improving quality? 3. What could be useful to help GPs get the most out of data/what’s missing? Intended audience This workshop is aimed at anyone using or interested in the use of EHRs for any purpose and particularly for GPs and those using the EHRs systems in the first instance.

Convenors: Ceire Costelloe, Lecturer in Medical Statistics, Brennan Kahan, Lecturer in Medical Statistics, Pragmatic Clinical Trials Unit, Centre for Primary Care and Public Health, Queen Mary University of London Introduction In the researcher’s perfect world every patient considered for entry into a randomised controlled trial would satisfy all entry criteria, would provide informed consent, would complete their allocated treatment as described in the protocol, and would contribute data records which were complete in all respects. In practice this ideal is never achieved. Not all patients adhere to the therapy to which they were randomly allocated. This lack of compliance can take a number of forms and may be due to a decision made by the patient or the clinician. A patient may have been included in a trial when the inclusion criteria were not satisfied. The patient may fail to take any or part of their prescribed medicine. Either the patient or the clinician may choose to discontinue treatment, perhaps as a result of complete recovery, or the clinician may assign another treatment, or a therapy different from any prescribed in the protocol because of side-effects or inadequate response to the previous treatment. Aims By the end of this session participants should: 1. be aware of departures from random allocation that can occur within a trial 2. recognise appropriate methods to deal with each type of departure Format Participants will be randomised to a ‘real world’ multi-centre trial with individual outcomes provided to each participant, including details on if and how individuals have departed from allocation. A CONSORT flowchart will be populated by the group, reflecting where and what kind of departures have occurred. Using this ‘real world’ example the concepts of protocol deviation will be introduced and participants will work in small groups to assess how this will affect trial outcomes. Each group will feedback results to other groups and each section will be completed with a short summary presenting how each of these topics should be considered in trial design and analysis. Content The workshop will open with a brief introductory lecture. This will be followed by small group work and feedback to the wider group. There will be a summing up lecture at the end of the session and workshop materials and a reading list will be provided to each participant. Audience The intended audience will be both clinical and non clinical audiences (including PhD students and trainees) with limited trial experience.

Workshop 7

Workshop 8

Thursday 10th July, 16.00 - 17.30, Room 2.14, 2nd floor

Friday 11th 2014, 10.00 - 11.00, Room 2.11, 2nd floor

Incorporating Genomics into General Practice: is it time for primary care researchers to take the lead? Convenors: Professor Blair H. Smith, University of Dundee, Dr Imran Rafi, Chair of Clinical Innovation and Research, Royal College of General Practitioners, Professor Joe Kai, University of Nottingham, Professor Heather Skirton, University of Plymouth & Chair of the European Board of Medical Genetics Aim To inspire primary care academics to engage in translational genomics research, taking scientific advances byond the laboratory to the general practice surgery Educational Objectives 1. Describe the phases of translating genomics into clinical practice 2. Identify current gaps in the translation of genomics into clinical practice 3. Demonstrate exemplars of translational genomics in primary care practice across the research translation spectrum 4. Explore the potential of incorporating primary care expertise in health services research and population-based epidemiology in translational genomics research 5. Nurture research collaborations between workshop participants Format The 90 minute workshop will comprise: short interactive presentations on a proposed classification of translational genomic research (10 mins) and exemplars from academic departments of Primary Care (20mins) with 10 minute discussion; followed by small group discussions to explore participants’ transferable research experience and insight into areas for further discussion or collaboration (20mins x 2); and finally, small group work summarised in plenary session (10 mins). Content (i) Initial 10 minute interactive presentation on phases of the translational genomic research spectrum from initial clinical evaluation (phase T1), to clinical practice guideline development and application (T2 & T3), to assess health outcomes of application (T4) (ii) Series of 5 minute presentations on primary care research across translational spectrum: F Identifying genetic basis of pain and its management (T1) F Improving preconception genetic risk assessment in primary care (T2) F Improving familial hypercholesterolaemia identification in primary care (T3) F Improving breast cancer outcomes through familial breast cancer identification (T4) (iii) Chaired discussion (10 minutes) (iv) Two consecutive small group sessions (20mins x2) will: 1. explore participants’ experience of translating and implementing scientific advances into clinical practice and how this could be applied to T1 and T2 primary care genomic research 2. consider participants’ experience of implementing guidelines for technologies into primary care clinical practice and how this could be applied to T3 and T4 primary care genomic research (v) Plenary session (10 mins): summarise 2 key points for each small group and explore potential areas for collaborative work

Realising the potential of the Clinical Practice Research Datalink (CPRD) Convenor: Professor Roger Jones, BJGP, Dr John Parkinson and Dr Antonis Kousoulis, CPRD Aim and educational objectives The Clinical Practice Research Datalink (CPRD) is the new English NHS observational data and interventional research service, jointly funded by the NHS National Institute for Health Research (NIHR) and the Medicines and Healthcare products Regulatory Agency (MHRA). As well as providing a huge resource of high-quality and secure data for observational studies, the CPRD provides research support in fields such as pharmaco-epidemiology, pharmaco-economics and outcomes research, and is increasingly developing capacity in interventional research including patient recruitment, clinical trial management and long term follow up. The aim of this workshop is to provide an overview of the structure, function and capabilities of the CPRD, to describe and discuss its potential application to various forms of primary care research and to encourage primary care academics both to make use of it as a research resource and also to contribute data to it. The educational objective is to provide an understanding of what the CPRD is and how it can be incorporated into future planning for research projects Format We will use an interactive format, with an introductory description of the CPRD, including examples of the clinical impact of CPRD research, followed by an extended discussion, which may be structured around themes such as data security, validity, research designs, outcomes research and research governance and ethics .Participants will be encouraged to ask about how the CPRD might be best used in their own areas of research interest, and further discussion, after the workshop, of potential CPRD research will also be encouraged. Printed material about the CPRD and the clinical and research impact of selected published studies will be available. Content We will focus on the way in which the CPRD has developed from the GPRD and how it has now incorporated the multiple data sources mentioned above. We will concentrate on the new capabilities being developed at the CPRD and also describe the application, research governance and ethics procedures involved. We will emphasise the need to extend data capture for the CPRD to improve its capacity to study areas such as rare disease and rare drug effects, diseases in sub-groups of the population, and infectious disease epidemiology. We anticipate a discussion of the data security and ethical issues arising from the recent events surrounding the NHS England care.data project. Intended audience The workshop should be of interest both to researchers who have already worked with the GPRD and also those coming anew to large database research. There is no upper limit on the number of participants.

Intended audience Primary care academics with an interest in evaluating and implementing genomic advances in General Practice. This includes both health services researchers and those with expertise in analysing primary care or other population-based datasets. Practitioners involved in guideline development, implementation and evaluation may also be interested and are welcome.

ABSTRACTS OF ORAL PRESENTATIONS Prize plenaries followed by oral presentations in parallel sessions Wednesday 14.00 - 14.30 PP1 Teaching and training practices: in the wrong places and with the wrong patients? Eliot L Rees, Simon P Gay, Robert K McKinley School of Medicine, Keele University, North Staffordshire, UK The problem: Whether practices which teach undergraduate medical students (teaching practices) or train postgraduate GP registrars (training practices) are representative of practices across England is uncertain. The associations between practices teaching and training status and their practice level outcomes are also unclear. The approach: Practice level analyses (n=8,207) were conducted between English general practicesâ&#x20AC;&#x2122; undergraduate teaching and postgraduate training status and routinely collected and reported data on their socio-economic deprivation (Index of Multiple Deprivation), rurality (urban-rural indicator), workforce (number of GPs, full time equivalents, mean full time equivalence (FTE)), patient list sizes (including number of patients per FTE), patientsâ&#x20AC;&#x2122; health status (EQ-5D-5L), patient satisfaction (from the GP Patient Survey), and quality of care (from the Quality and Outcomes Framework). Chi-square, one-way analysis of variances, univariate logistic regression analyses and multivariate linear regression analyses were used on practice level measures. Findings: Of 8,207 English GP practices, 2,834 (34.5%) teach undergraduates and 2,301 (28.0%) train postgraduates. Practices which serve deprived populations are less likely to teach (Odds Ratio (OR): 0.78, 95%CI: 0.69, 0.88, p<0.001) and train (OR: 0.55, 95%CI: 0.48, 0.63, p<0.001). Rural and urban practices are equally likely to teach (OR: 1.02, 95%CI: 0.89, 1.15, p=0.82) but rural practices are more likely to train (OR: 1.55, 95%CI: 1.37, 1.77, p<0.001). Teaching and training practices have more GPs (F(3, 7772)=966.38, P<0.001) working less per week (p=0.002), with a smaller proportion of male GPs (F(3, 7772)=177.91, p<0.001). Teaching and training practices have more patients (F(3, 7700)=758.01, p<0.001), but fewer per FTE GP (F(3, 7700)=34.70, p<0.001). All five EQ-5D domain scores for patients of training practices reflect better health (all p<0.001) while scores for patients of teaching practices show no differences except for anxiety / depression which are lower than for non-teaching practices (F(3, 7764)=5.56, p=0.001). Teaching (B=.341, 95%CI: .233, .449, p<0.001) and training status (B=.601, 95%CI: .474, .729, p<0.001) were both associated with increased patient satisfaction. Teaching and training practices have higher scores on all QOF domains (p<0.001). Teaching status (B=.578, 95%CI: .282, .874, p<0.001) and training status (B=1.004, 95%CI: 0.656, 1.352, p<0.001) are both associated with total QOF score. Consequences: Teaching and training practices are not representative of English general practices: they are more rural, and care for patients who are less deprived with better health status than practices which donâ&#x20AC;&#x2122;t. Training GPs in leafy suburb and bucolic rural practices has implications for recruitment to deprived urban practices. Urgent investment is needed to increase numbers of deprived urban area training practices. Teaching and training status are associated with better quality of care and patient satisfaction: providing more training in deprived urban practices may have positive consequences for care.

PP2 Multiple conditions and clinical trials: A systematic review of the impact of multimorbidity in inclusion, assessment and outcome of self-management studies. Cassandra Kenning, Peter Coventry, Peter Bower University of Manchester, Manchester, UK The problem: Multimorbidity has many potential implications for health care delivery, but a particularly important impact of multimorbidity concerns the validity of trial evidence underpinning clinical guidelines for individual conditions. There has been no detailed analysis of the impact of multimorbidity in self-management trials. The current study explores how authors of published trials of self-management interventions have managed inclusion, description and impact of multimorbidity in exemplar disorders. The aim was to assess current approaches to the management of multimorbidity in the context of trials, and to consider implications for evidence synthesis and reporting. The approach: To achieve our aims, we restricted our analysis to a small number of exemplar conditions, where multimorbidity is common and where there is a known self-management literature: type II diabetes; coronary heart disease (CHD) and chronic obstructive pulmonary disease (COPD). We focussed our search on published Cochrane reviews. Data were extracted from the trials on inclusion/exclusion (proportion of studies excluding patients with multimorbidity at baseline); description (reporting of sample characteristics in terms of multimorbidity); and outcomes (analyses of the impact of multimorbidity on outcomes through secondary and moderating analyses). Findings: Eleven reviews identifying 164 unique trials were identified. Inclusion/exclusion: across conditions, 55% of trials reported excluding patients with forms of multimorbidity. Reasons for exclusion were poorly described or defined and the number of studies excluding based on multimorbidity varied across conditions. Description: reporting of multimorbidity within the trials was poor, with only 35% of trials reporting on multimorbidity in their patient samples. Those that did report other conditions only reported on a small number of comorbid conditions such as hypertension and heart disease. Outcomes: secondary analyses exploring the impact of multimorbidity was very rare. Three trials (1.8%) reported analysis of moderation by multimorbidity, 4 trials reported on impact of the intervention on comorbid depression/ anxiety, and 2 trials used multimorbidity in the analysis as a covariate. Consequences: The number of patients with multimorbidity is increasing and patients with one initial long-term condition are likely to develop further conditions. Regardless, intervention trials continue to be disease specific even though many conditions may benefit from similar interventions (ie education, diet and exercise). Trials often exclude patients with multimorbidity, and reporting of multimorbidity in trials including such patients is generally poor. This limits judgements of the external validity of the results for clinical populations. A consistent approach to the conduct and reporting of secondary analyses of the effects of multimorbidity on outcomes, using current best practice guidance, could lead to a rapid development of the evidence base.

ABSTRACTS OF ORAL PRESENTATIONS Oral presentations and parallel sessions

1A.2

1A.1

Advancing primary care through education and research: from translational research to translational scholarship

Understanding unscheduled use of urgent care services by adults with chronic obstructive pulmonary disease in Bristol: a feasibility study using system dynamics

Joanne Reeve1, Lucy Cooper1, Sean Harrington2, Jane Watkins2 1 University of Liverpool, Liverpool, UK, 2Advocacy in Wirral, Wirral, UK

Melanie Chalder1, Helen Baxter1, Sam Evans2, James Calvert3, Jonathan Benger4, Ali Heawood1, Nabil Jarad5, Sally Brailsford6, Sarah Purdy1 1 University of Bristol, Bristol, UK, 2University of Exeter, Exeter, UK, 3North Bristol NHS Trust, Bristol, UK, 4University of the West of England, Bristol, UK, 5University of Bristol NHS Foundation Trust, Bristol, UK, 6University of Southampton, Southampton, UK

The problem: We need to “Think Differently” about healthcare as “a necessary prerequisite for doing differently” (NHS Institute for Innovation). Knowledge translation - mobilising knowledge across the different settings of research and practice - is identified as one means to support innovation and change. Promoting and strengthening translational research is a priority for research funders. What we still lack is an adequate framework for understanding, evaluating and so supporting translational research.

The problem: Reducing unscheduled use of urgent care services is one of the biggest challenges currently facing the National Health Service. Recently published figures suggest that people living in Bristol are more likely to end up in hospital unexpectedly than others throughout the region or even nationally. Hospital admission for acute presentation of chronic respiratory disease is common, with emergency admissions for chronic obstructive pulmonary disease (COPD) and asthma being 25% higher in Bristol than the national average in 2010/11. Previous research suggests that demand for services by patients with COPD can be reduced through effective and pro-active management of the condition in the community setting. The approach: We conducted a mixed methods feasibility study using system dynamics as a means of exploring unscheduled use of urgent care services locally. This analytical methodology combines a number of qualitative and quantitative approaches - semi-structured interviews, observation, data extraction and linkage - in order to enhance understanding of complex networks such as patient flows and care pathways. The qualitative aspect of work constructs of a conceptual map of the overall system, describing its structure, component parts and their relative configuration whilst the quantitative aspect aims to develop this map into a computer model which can illustrate real and simulate hypothetical change. Our research considered how services were provided across the primary, secondary, community and social care settings in Bristol. Study participants were healthcare professionals, managers and adults with COPD. Findings: The study had three purposes - firstly, to establish ‘proof of concept’ in relation to the system dynamics methodology when applied to health and social care settings; secondly, to determine the feasibility of using the methodology to explore the use of unscheduled urgent care services by adults with COPD in Bristol; and thirdly, to model the effect of changing patient flows on local demand, capacity and provision. Our research is ongoing but interim feedback suggests that the first two of these aims have been met and the third is achievable. The qualitative stage of the systems dynamics approach involved the conduct of 17 interviews and the resulting conceptual map illustrates how 22 local care providers are configured. Availability of quantitative data was, however, more limited, particularly from community and social services. Consequences: System dynamics offers a novel and pragmatic approach to exploring patient flows and care pathways in a complicated, multi-agency environment. In this instance, additional work is required to ensure that the quantitative model is fully populated and reflective of the entire local health economy for COPD care. However, the technique has considerable utility and could be adapted to explore the potential impact of service reconfiguration across a diverse range of clinical populations and health conditions.

The approach: We use our NIHR funded Innovation project - Bounce Back - as a case study to explore these issues. Led by Advocacy in Wirral (AiW), a mental health charity, in partnership with Liverpool University (UoL), the project aims to tackle three barriers to accessing mental health care identified from previous research. And so to transform the way we provide care to people who are distressed. The goal is to integrate the AiW model of practice into the general practice setting. The researchers’ role is to support the evaluation of both the process of integration and outcomes of the intervention. Recognising the AiW approach as a complex intervention (CI), we are using Normalisation Process Theory (NPT) as a framework to support evaluation. Our presentation derives from a critical reflection on the process of developing, delivering and evaluating the project framed by the existing literature on evaluating translational research. Findings: Three themes emerge. 1) Both partners have had to modify expectations about the outputs of the project. AiW sought data on their service impact to support future bids for contracts. UoL sought understanding of the process of integrating a CI. With a sharing of expertise to achieve both goals. In practice, our output has been creating a new understanding of the complex intervention. 2) Actions for both sides have changed. AiW staff have had to develop new scholarship roles - in the critical practice of curiosity and creativity. UoL staff have taken on new roles in service development and delivery. 3) Much of our time has been taken up with developing a shared vision. Within the team, in order to support capacity to manage change. And with wider stakeholders, in order to support meaningful engagement and a shift in focus from transacting (contracting) to transforming services. Consequences: Our observations resonate with Evans’ (2014) description of translational research as moving from ‘bridging’ between practices to ‘blurring’ traditional boundaries. Like Evans, we note implications for the “depth of research done”. Although for us, we believe we have achieved greater depth of learning but at a ‘cost’ of time. Leaving us with a new question of how to evaluate this alternative approach to translational research. Drawing on Pozen, we will discuss how our experience and findings are supporting the development of a new framework of translational scholarship. 1A.3 The Primary Health Care Search Filter: Making a world of difference to literature searching for primary care professionals Lynsey Brown, Amanda Carne, Petra Bywood, Ellen McIntyre Primary Health Care Research & Information Service (PHCRIS), South Australia, Australia The problem: It is widely accepted that policy and practice should be based on evidence, but a challenge faced by time-poor researchers, policy makers and practitioners is how to access relevant evidence efficiently. Given the rapidly increasing volume of the primary health care (PHC) literature

ABSTRACTS OF ORAL PRESENTATIONS evidence base, Flinders Filters and the Primary Health Care Research & Information Service collaborated on a project to develop a search filter. Search filters are evidence-based, literature search strategies which offer a standardised, systematic, subject-based search with a known level of performance. The PHC Search Filter is designed to enable efficient and effective retrieval of relevant literature, facilitating access to PHC resources. The approach: The Filter was developed in Ovid Medline based on a validated methodology, comprising five phases: constructing a gold standard set of PHC-specific articles; identifying appropriate index terms and textwords; testing combinations of terms; assessing the most effective strategy; and translating the Filter for use in PubMed. The Filter retrieves citations from both unindexed and indexed sources, enabling searchers to attain the most up-to-date information. During development the Filter achieved a retrieval effectiveness of 88 per cent. In addition to a range of limits (e.g., free full text, specified time frame), it offers two search options: ‘one-click’ where searchers choose from 12 topics appropriate to their area of interest; and ‘build-your-own’, for customised searches. In the 18 months post-launch an evaluation was conducted to assess uptake and impact of the Filter. This mixed-methods study combined web statistics on access and search preferences; findings from an online survey exploring reasons for use; and responses to open-ended survey questions around perceptions of the tool. Findings: Google Analytics statistics suggest the Filter is a useful tool, with high numbers of individuals exploring its capabilities, and a frequent rate of return visitors. These statistics also illustrate the popularity of the ‘build-your-own’ option, and interest in ‘one-click’ topics around chronic disease management and mental illness. With 90 respondents, the survey provided details about the positive response to the tool in terms of ease of navigation, recommendations to others and usefulness; and informed future updates. Qualitative feedback was supportive with comments describing the breadth of the Filter and its ability to simplify PHC searches. Consequences: Ready access to evidence of best practice to inform care, guide decisions and assess value remains one of the key challenges for primary care professionals. Use of a search filter can facilitate access, increase the value of the results retrieved and potentially reduce the burden associated with literature searching. It is believed that by promoting access to quality, relevant literature, the freely accessible PHC Search Filter may encourage uptake of evidence into research, practice and policy. 1A.4 Establishing RAPPORT - findings from a realist evaluation of patient and public involvement in NHS research Amanda Howe2, Sally Kendall1, Patricia Wilson1, Fiona Poland2, Julia Keenan2, Elspeth Mathie1, Clare Goodman1, Marion Cowe1, Diane Munday1, Penny Vicary2, Amander Wellings2, Sophie Staniszewska3 1 University of Hertfordshire, Hatfield, Herts., UK, 2University of East Anglia, Norwich, Norfolk, UK, 3Warwick University, Warwick, Warks., UK The problem: Patient and public involvement (PPI) is a growing expectation of all health sectors, but little is known about how this is operationalised in NHS research practice. The RAPPORT study (Research with patient and public involvement: a realist evaluation) aimed to F Determine the variation in types and extent of PPI in a sample of six clinical areas with a broad range of settings and research methods F Describe key processes and mechanisms of PPI F Explore perceptions of their experiences with the researchers and members of public F Assess the mechanisms which contribute to public involvement being routinely incorporated in the research process F Evaluate the impacts on research processes and outcomes, and 28

Identify barriers and enablers to effective public involvement in research.

The approach: project sampling occurred via the NIHR database. 23 case studies were undertaken, collecting specified data on PPI backed up with 2 or 3 interviews per case study with researchers and lay contributors. The stages of normalisation process theory were used to analyse the level and extent of PPI over the timeline of each project. Data analyses were iterated between three researchers to ensure credibility of data interpretation, working with the wider research team and an advisory group to check interpretation and implications. The research team also monitored and reflected on the impacts and success of their own PPI inputs. Findings: projects had variable actual PPI input: continuous in-depth input over the whole project was uncommon, with most PPI in research design and dissemination phases. A classification of four models of PPI was developed, with emerging evidence of different impacts according to level of involvement. There was some evidence of PPI being poorly understood and undervalued. Authentic relationships between researchers and members of the public appeared a crucial dimension of the extent to which PPI representatives felt able to make an effective contribution. Clarity about expectations, resources to support input, mutual commitment to maximising the PPI contribution, and developing interpersonal relationships over time all appeared to be key to effective PPI. Consequences: There is evidence that researchers may need more training and guidance on how to make PPI work within their projects: that more could be done to ensure that researchers account for and evaluate the PPI within their funded projects: and that settings which have an ongoing PPI resource between individual research projects are more likely to use PPI in an effective and sustained way in individual projects - because both the process and the relationships have become normalised into routine practice. The models developed should enable researchers and PPI leads to clarify expectations, and may also assist resource allocation. The important finding about interpersonal relationships will act as a challenge to researchers to consider how they include their PPI colleagues as part of their team. 1B.1 Effect of a multifaceted intervention on potentially inappropriate prescribing in older patients in primary care: a cluster randomised controlled trial (the OPTI-SCRIPT study) Barbara Clyne1, Susan M Smith1, Carmel M Hughes2, Fiona Boland1, Marie C Bradley2, Janine A Cooper2, Tom Fahey1 1 HRB Centre for Primary Care Research, RCSI, Dublin, Ireland, 2Queen’s University Belfast, Belfast, UK The problem: Potentially inappropriate prescribing (PIP) (use of medicines that introduce a greater risk of adverse drug-related events where a safer, as effective alternative is available) is common in older people and can result in increased morbidity, adverse drug events and hospitalisations. The prevalence of PIP in Ireland in those aged ≥ 70 years is 36% with an associated expenditure of over €45 million. The aim of this study is to determine the effectiveness of a multifaceted intervention in reducing PIP in Irish primary care. The approach: A cluster randomised controlled trial (RCT) was conducted with 21 GP practices based in the greater Dublin area and 196 patients, all of whom were identified as having PIP. Practices were allocated using minimization to intervention or control arms after baseline data collection. Intervention participants received a complex multifaceted intervention incorporating academic detailing, medicines review with web-based pharmaceutical treatment algorithms that provided recommended alternative treatment options, and tailored patient information leaflets. Control practices delivered usual care and received simple, patient-level

ABSTRACTS OF ORAL PRESENTATIONS feedback on PIP. The primary outcomes were the proportion of patients with PIP and the mean number of potentially inappropriate prescriptions. Secondary outcomes included incidence rates of drug-specific outcomes, well-being and beliefs about medication. Routinely collected national Irish prescribing data were also analysed, acting as an additional contemporary national comparison. Findings: All 21 practices and 190 patients were followed up (6 patients were lost). At intervention completion, there was a significant reduction in the proportion of patients in the intervention group with PIP compared to control (OR 3.06 95% CI 1.43, 6.53). The average PIP in the intervention group was also significantly lower than control (0.70 v 1.19 p=0.004). In the control group, the incidence rate for PIP was 1.42 times the incidence rate for the intervention group. The intervention had a statistically significant effect in reducing proton pump inhibitor (PPI) prescribing but not on the other PIP drug-specific secondary outcomes. There was no difference in patient self-reported outcomes of well-being and beliefs about medication. Analysis of national prescribing data indicated that nationally PIP increased during the study period. A process evaluation highlighted that patient information leaflets were not used at all. Consequences: The study found that a multifaceted intervention incorporating academic detailing with a pharmacist and medicines review with web-based pharmaceutical treatment algorithms that provide alternative treatment options is effective in reducing PIP in primary care. This reduction in PIP was attributable to reduction in PPI prescribing. Implementing a system of structured reviews for older patients with PIP in Irish primary care may be effective but formal resourcing of such services would need to be considered. 1B.2

Findings: Initial analysis finds a continuous and substantial increase in the rate of prescription of opioids over the 10-year period. The most commonly prescribed drug was co-codamol (over 2 million prescription items in 2012). Tramadol showed an almost threefold increase over the ten-year period; >1.01M items in 2012. A graduated rate of prescribing across the SIMD quintiles in 2012 was seen, with those in the most deprived SIMD receiving 3.5 times the number of prescriptions compared to those in the most affluent. In 2012, 949,108 patients were prescribed any opioid or combination drug, representing almost a fifth of the Scottish population. Co-codamol was the most commonly prescribed (n=615,520 patients) opioid, followed by Tramadol (n=205,514 patients) and with over forty thousand patients prescribed morphine. There were marked differences in regional prescribing rates by NHS health board. Further analysis by age & sex of patients receiving opioid prescriptions and GP practice prescribing patterns is ongoing. Consequences: This study demonstrates an increase in opioid prescribing throughout Scotland with apparent regional and socio-demographic variations. The reasons for these substantial increases are unclear. This study will help inform current practice and education, identify factors associated with good opioid prescribing and with actual or potential problems nationally. 1B.3 Application of Prescribing Safety Indicators to the Clinical Practice Research Database: a retrospective cohort study S Jill Stocks1, Evangelos Kontopantelis1, Sarah Rodgers2, Anthony Avery2, Darren Ashcroft1 1 University of Manchester, Manchester, UK, 2University of Nottingham, Nottingham, UK

Opioid prescribing across Scotland - rates and variations Nicola Torrance1, Bradley Kirby6, Lesley Colvin2, Steve Gilbert3, Gary Macfarlane4, Michael Serpell5, Alex Baldacchino1, Timothy Hales1, Peter Donnan1 1 University of Dundee, Dundee, UK, 2University of Edinburgh, Edinburgh, UK, 3NHS Healthcare Improvement Scotland, Fife, UK, 4University of Aberdeen, Aberdeen, UK, 5University of Glasgow, Glasgow, UK, 6 Information Services Division NHS Scotland, Edinburgh, UK The problem: Opioids are well-established as effective agents for acute pain and cancer-related pain, with systematic reviews also confirming their effectiveness in some chronic non-malignant pain conditions. Opioids are associated with many side effects and longer term adverse effects but despite these concerns, prescription of opioids, misuse rates, overdose and opioid related deaths have all increased steadily, with most published data coming from the USA. Detailed UK data on opioid prescribing are currently lacking although there is limited evidence of overall increased prescribing. In Scotland, in the year to April 2012, there were 2,263,233 prescriptions for opioids dispensed (excl. compound preparations) at a gross ingredient cost of >£29.6M The approach: We aimed to describe opioid prescription rates over a 10-year period across Scotland and explore the relationship between socio-demographic factors and local/regional opioid prescribing rates. Data were obtained from Information Services Division (ISD) of NHS Scotland, including the number of paid (prescription) items and number of defined daily doses of all opioids prescribed and dispensed between April 2003 and March 2012, by individual drug. For 2012 only, data included classification by: i) Health Board area; ii) GP practice; iii) sociodemographic factors (Scottish Index of Multiple Deprivation (SIMD) quintiles); v) individual exposure to drugs. Classification of drugs is based on the British National Formulary and subdivided into “weak”, “strong” and “combination” opioids.

The problem: Prescribing errors in primary care cause significant harm; adverse drug events account for around 7% of hospital admissions and half of these are considered preventable. Over half of these admissions are associated with the following groups of drugs: anti-platelets, diuretics, NSAIDs and anticoagulants. Eight prescribing safety indicators (PSI, listed below) to electronically identify patients at increased risk of a prescribing error have been applied in 72 English general practices. In Scotland the prevalence of a set of indicators of high risk prescribing has also been measured in 315 practices. However there is a need to systematically identify which PSI are the most useful for identifying patients at increased risk of a prescribing error, and how frequently these events occur in a large nationally representative dataset. The approach: The prevalence for each PSI listed below is being measured within the Clinical Practice Research Database (over 600 practices) F Diagnosis of peptic ulcer and prescribed a NSAID without PPI protection F Diagnosis of asthma and prescribed a betablocker F Aged ≥75 or over prescribed ACEI or loop diuretics and have not had urea and electrolyte monitoring F Diagnosis of thrombosis and prescribed a combined oral contraceptive F Prescribed methotrexate and not had a recent full blood count or liver function test F Prescribed warfarin and not had an INR F Prescribed lithium and not had a lithium level check F Prescribed amiodarone and not had a thyroid function test For each PSI the prevalence is measured relative to an audit date using a rolling time window. For long term conditions, such as asthma or peptic ulcer, the cumulative number of patients at risk since their first recorded diagnosis is included in the denominator. If being at risk is reversible such as requiring monitoring when prescribed a medication the denominator is the patients prescribed the medication within a fixed time prior to the audit date. Multilevel regression models will be used to explore the associations 29

ABSTRACTS OF ORAL PRESENTATIONS between prevalence and trends and practice attributes such as list size and patient attributes such as age, gender and co-morbidities. Findings will be reported as outlined below: F the mean annual prevalence of patients with a PSI recorded from 2000 - 2010 F the variation between practices in recorded PSI and associations with practice attributes e.g. list size F the associations between recorded PSI and patient demographics e.g. age, gender, multimorbidity F changes over time in the prevalence of PSI Consequences: This is the first report of the prevalence of PSI in the Clinical Practice Research Database. It will show whether these PSI are reliable in terms of identifying groups at increased risk of a prescribing error. 1B.4 Long-term opioid prescribing for non-malignant pain in primary care: a qualitative study Carolyn Montana, Robbie Foy, Jose Closs, Duncan Petty, Lucy Ziegler, Liz Glidewell, Allan House University of Leeds, Leeds, West Yorkshire, UK The problem: The place of opioids in management of chronic nonmalignant pain is contested. Even so their use is escalating, leading to concerns that patients are prescribed strong opioids inappropriately and alternatives to medication are under-utilised. We know little about factors that trigger escalation from weaker to more potent opioids or that maintains long-term use. It seems particularly important to bring together patient and GP perspectives to understand such trajectories. The approach: We sought to understand the mechanisms that maintained or escalated opioid prescribing in general practice. We used qualitative methods to explore the experiences, beliefs and expectations of patients and general practitioners (GPs). Individual in-depth interviews were undertaken with 23 patients currently prescribed long-term opioids and two focus groups were conducted with 15 GPs from Leeds and Bradford, UK. We analysed audio-recorded transcripts using constant comparative analysis, delineating shared and contrasting perspectives on the management of chronic pain and opioid prescribing. Findings: Long-term opioid prescribing could be understood through patient and GP characteristics that influenced clinical transactions. Patients were driven by the need for explanation, pain relief and to improve or maintain quality of life. Responses to these needs appeared to be influenced by patient levels of distress and their thoughts about the pain, about their GP, about what would happen in the future and about where responsibility for change resided. Clinicians varied in how they understood, or were interested in, chronic pain especially when the medical diagnosis or expectations of treatment were uncertain. They had varying negotiating styles, psychological skills and interest in dealing with distress or in managing conditions where their own approach is at odds with the patient’s wishes. These factors influenced four characteristics of clinical transactions, and thereby prescribing decisions: where perceived control resided and who had the ability to make change; the sense of continuity in the relationship, which influenced the drive for stability or change; clarity of strategy, including the degree to which long term planning was subverted by reactions to the immediate content of consultations and; the sense of trust and mutuality, the degree to which the question of pain management became a vehicle for negotiating and resolving wider aspects of the relationship. Consequences: Both patients and doctors struggled with chronic pain and 30

its treatment. As well as alleviation of pain and distress, patients seek explanations and continuity within trusting relationships. Doctors recognised their own discomfort in feeling unable to treat problems underpinning pain, identify a sufficient range of options to treat pain, and negotiate clear management plans. Findings suggest that reversal of the continuing rise in opioid prescribing needs to include three main targets: patient expectations; doctors’ beliefs and abilities and; practice systems for managing complex conditions and prescribing. 1C.1 Obesity in men: evidence synthesis findings and their implications for primary care Pat Hoddinott1 ,2, Daryll Archibald2, Clare Robertson2, Fiona Stewart2, Dwayne Boyers2, Flora Douglas2, Edwin van Teijlingen3, Alison Avenell2 1 University of Stirling, Stirling, UK, 2University of Aberdeen, Aberdeen, UK, 3 University of Bournemouth, Bournemouth, UK The problem: More men than women are overweight or obese in the UK. Men appear more likely to misperceive their weight, less likely to consider their weight a health risk, and to attempt weight loss. Little is known about effective ways to engage obese men in interventions. The approach: We undertook a realist synthesis of qualitative and quantitative evidence to investigate: what weight management interventions work for men, with which men, and under what circumstances? Concurrent quantitative (n=4), economic and qualitative systematic reviews were undertaken, using Cochrane methodology. Studies reporting weight management in men aged over 16 years were included. Findings were integrated thematically using a realist deductive and inductive approach which conceptualised interventions as a Context, Mechanisms, Outcome framework. Findings: Data were included for 1238 men from 11 trials in a review of men only RCTs; 12,934 men and women from 20 RCTs of men and women; 11,426 men and 63,990 women from 26 reports of UK non-randomised interventions; 13 qualitative studies linked to interventions and nine UK qualitative studies not linked to interventions. Included men were mostly white, middle class and not particularly obese. Few data on health outcomes or risk factors were reported, yet health concerns were an important motivator for men, along with image and family/peer triggers. No studies of interventions to increase engagement of men were identified and men were rarely consulted on programme design. Programmes with a reducing diet, physical activity and behaviour change training together are most effective. Interventions in trials were intensive. Few long term follow-up or economic data were available. Participating men prefer physical activity to reducing diets (although reducing diets considerably enhance effectiveness), resist rigid or extreme diets, and like facts and using pedometers. Fewer men engage in group interventions, but men do well in groups. Men are much less likely to drop out than women. Some individual tailoring is important. For some men, but not all, attending men-only groups may enhance effectiveness. The comparative effectiveness of NHS and commercial programmes for long-term weight loss was unclear for men. Men were less likely to choose commercial programmes. The setting (e.g. workplaces, sports’ clubs), humour and word-of-mouth from peers are important to increase engagement. Some expressed preference for longer follow-up, feared relapse, but also feared losing too much weight. Men are less likely than women to lose weight with orlistat. Consequences: Men’s preferences for reducing diet, physical activity and behaviour change programme components differ from women. Social and workplace settings may be preferred to health settings. Involvement of

ABSTRACTS OF ORAL PRESENTATIONS obese/overweight men in programme design is recommended to find innovative means of delivering services to increase the engagement of hard to reach groups. Funding: National Institute of Health Research, Health Technology Assessment Programme 1C.2 Addressing the lack of reliability and validity data from the General Practice Physical Activity Questionnaire (GPPAQ) using PACE-Lift physical activity trial data in 60-74 year old primary care patients Shaleen Ahmad1, Tess Harris1, Derek Cook1, Sally Kerry1, Elizabeth Limb1, Ulf Ekeland2 1 St George’s University of London, London, UK, 2University of Cambridge, Cambridge, UK The problem: The health and economic benefits of increasing physical activity (PA) in older adults has been widely documented. GPPAQ is a widely used primary care self-assessment tool to measure PA up to age 74 years, with PA interventions indicated for those who are less than active. It is included in the current hypertension QOF and forms part of the NHS Health Checks. Despite GPPAQ’s ubiquitous use, published reliability and validity data are extremely limited, particularly for 60-74 year olds. Additionally, occupation is used in scoring, so GPPAQ may be less accurate in retired people. Walking is the main PA in older adults and GPPAQ includes walking questions but they are not used in the scoring. The approach: We will use baseline and 3 month data from the PACE-Lift PA trial in 60-74 year old primary care patients to assess the reliability and validity of GPPAQ. Reliability will be assessed by comparing control group baseline and 3 month GPPAQ scores. Validity will be assessed by comparing the sensitivity and specificity of GPPAQ to a gold standard of objective PA assessment by accelerometry with both measures assessing the same 7 day period at baseline in both control and intervention groups. We will also examine the effects of occupational status (working versus retired) and the effect of including walking data in the GPPAQ score on the validity of GPPAQ. Findings: Of 298 PACE-lift participants (150 intervention, 148 control), 270 had GPPAQ and accelerometer data recorded at baseline and 236 had both recorded at 3 months. 15.9% (43/270) were classified as active on GPPAQ at baseline and 191/288 (66%) were retired. The GPPAQ had reasonable reliability (58% (61/105) scored the same at baseline and 3 months) but poor validity in this age group, with a sensitivity (ability to identify those who were truly active using the gold standard accelerometry) of 20% (13/64) and specificity (ability to identify those who were truly inactive) of 84% (173/207). Data on the effects of occupational status and walking questions on GPPAQ validity will be presented. Consequences: Our findings suggest that GPPAQ has poor validity in older adults, which questions the appropriateness of its use within NHS health checks and in the current hypertension QOF in this age group. Whether the validity can be improved by including data on walking and limiting its use to those who are working will be discussed. 1C.3 The use of champions to promote digital enablement in the community: a dallas evaluation study Claire Nugent, Marilyn McGee-Lennon, Matt Mouley-Bouamrane, Frances Mair University of Glasgow, Glasgow, UK

The problem: Innovative technology can improve the health and wellbeing of an aging population in which multimorbidity is increasing. However, the rate of technology innovation often exceeds the ability of health and social care systems to adapt delivery of care, and of the public to engage in its use. Empowering champions to promote digital enablement in their communities may allow more people to use technology to live well and independently. However research into the use of champions has been limited to academic and professional champions. In 2011, a large-scale telehealth and telecare project (dallas) was launched, aiming to recruit 169,000 users. One of the four dallas communities (More Independent (Mi) Liverpool) developed a champions programme. This study aims to describe the champions programme and evaluate its implementation. The approach: Qualitative methods. Thirteen champions and seven staff members from Mi and two of its partner organisations - PSS and TUC were interviewed or took part in a focus group. Staff members included: project/team leaders; a service manager; project workers; an informatics co-ordinator; and the regional co-ordinator of the group training digital champions. Participants were asked about the structure and function of the champions programme and the barriers and facilitators to its implementation. Interviews and a focus group were audio-taped and transcribed to provide the data for analysis. The data was analysed using a coding framework conceptually underpinned by Normalisation Process Theory. Findings: Two types of champions were recruited, trained and hosted by different organisations in the city: community and digital champions. Both groups volunteered varying amounts of time to their role. A key role of community champions was to signpost members of their community to products and services that could improve their health and wellbeing; the main role of digital champions was to promote digital inclusion through education. Emerging themes on the process of implementation include: the finding that champions themselves benefited from their involvement in the programmes; champions could shape training to suit their interests and learned valuable information about the public’s perception of using technology to improve health and wellbeing. Sustainability of the programme was a priority for staff and partnerships with primary care services and health trainers were being developed. While champions managed to reach a wide range of people, measuring their impact was perceived as a challenge. Consequences: This study provides novel insights into the potential role of community champions to promote digital inclusion in a large scale service deployment. It illustrates the evolving definition of a champion and how members of the community can be empowered to promote change in the way people live, particularly in promoting lifestyle changes, such as smoking cessation but highlights the difficulties of ascertaining the full benefits of this approach. 1C.4 Exercise referral schemes: GPs’ perceptions of barriers and facilitators to referral Christine Rea, Andrew Gilliland, Margaret Cupples, Neil Heron, Mark Tully Queen’s University, Belfast, Belfast, Northern Ireland, UK The problem: Approximately two thirds of the UK population report physical activity levels which are insufficient for health benefits. Primary care practitioners have a responsibility to promote physical activity. Exercise referral schemes, providing tailored physical activity programmes, may encourage people to incorporate regular activity into their daily lives but there is a lack of robust evidence regarding the effectiveness of such schemes and a suggestion that GPs are reluctant to refer patients. A recent systematic review called for more in-depth exploration of barriers to exercise referral schemes. Thus, we aimed to explore GPs’ perceptions of exercise referral schemes and of barriers or facilitators to referral. 31

ABSTRACTS OF ORAL PRESENTATIONS The approach: We invited a purposeful sample of GPs, of differing age and gender, working in communities with different levels of socio-economic deprivation, urban and rural locations and differing proximity to the nearest exercise scheme, to participate in one-to-one semi-structured interviews. This approach aimed to maximise variation within the sample and the potential variability of experiences. A topic guide for interviews was based on findings of a literature review and discussion with local exercise referral scheme co-ordinators. A self-completed questionnaire, administered at the start of the interview, recorded interviewees’ age, gender, physical activity (International Physical Activity Questionnaire) and frequency of referral to exercise schemes. The interviews were audiotaped, transcribed and analysed by two researchers independently, using a thematic framework and NVIVO software. Recruitment continued until data saturation was reached. Findings: Analysis of 10 interviews identified three main themes which contributed to GPs’ decisions to refer: (1) experience of referral schemes, (2) GP’s own physical activity and (3) perceived locus of responsibility. Experience that schemes were inconvenient for patients, referral processes were not straightforward, staff were not enthusiastic or feedback was poor was a barrier to referral: the converse of these experiences facilitated referral. GPs who were physically inactive or did not believe that exercise had positive health benefits were less likely to refer than those who were active and perceived value in physical activity. GPs who perceived that they had responsibility for patients’ physical activity levels and health were less likely to refer; those who thought this responsibility belonged to the government or patient were more likely to refer but their readiness to refer was tempered by their experience of referral and patients’ compliance. Perceptions of the value of incentives for referral or patient attendance were mixed. Consequences: The findings suggest that GPS’ referral to exercise schemes may be increased by ensuring that schemes are easily accessible, the process of referral is straightforward and enthusiastic staff provide ready feedback. Increased referral and uptake should allow more robust evaluation of the effectiveness of these schemes in helping support patients to increase their physical activity and thereby improve their health.

service delivery: proactive population orientation, reactive population orientation and individual patient orientation. Although proactive approaches were key to aligning services with the needs of diverse groups, staff tended to describe the use of predominantly reactive strategies. There were few examples of PCOs systematically collecting and using data on the characteristics of their patient population, or systematically seeking feedback from patients from diverse groups. We identified a range of challenges and barriers to PCOs working proactively to improve responsiveness. Consequences: Proactive approaches such as mapping the characteristics of local populations and planning for the needs of vulnerable groups are essential to delivering a service that meets the needs of all patients. Improving responsiveness requires the development of a shared understanding of the notion of responsiveness and of the strategies that will help achieve it, and consideration of the drivers, incentives and barriers to improvement. 1D.2 Are emergency hospital admission rates for ambulatory care sensitive conditions related to patients’ perceptions of access to their GP surgery? Cross sectional analysis Robert Fleetcroft1, Nicholas Steel1, Sarah Purdy2, John Ford1, Amanda Howe1 Norwich Medical School, Norwich, UK, 2University of Bristol, Bristol, UK

The problem: Emergency admissions are costly and rising. ‘Ambulatory Care Sensitive Conditions’ (ACSCs) are those which are considered to be potentially avoidable. Some commentators have suggested that reduced access to primary care services may lead to an increase in avoidable hospital admissions. There has been little research into the causes of variations in emergency admissions between general practices, and the effects of access to general practice and other factors have not yet been empirically tested. We aimed to examine the relationship between ACSCs and patient access to general practices, adjusted for other potential explanatory factors.

1D.1 Delivering responsive primary care: strategies, barriers and challenges. Carolyn Tarrant1, Emma Angell1, Richard Baker1, George Freeman3, Mary Boulton2 1 University of Leicester, Leicester, UK, 2Oxford Brookes University, Oxford, UK, 3Imperial College, London, UK The problem: Primary care organisations (PCOs) do not always respond to the needs of diverse groups of patients so certain groups are underserved and disadvantaged. GP practices are encouraged to be more responsive to address these inequalities. Although responsiveness may be an aspiration, there is a need for understanding about how to make it a reality. The approach: As part of a larger study to develop a patient-report measure of responsiveness, we conducted interviews with 27 members of staff in 13 PCOs in England, purposively sampled for diversity in demographics and job roles (including GPs, practices managers, pharmacists, and administrative staff). Staff were asked about approaches to being responsive to the needs of their patients, and about the challenges and barriers to achieving this. We conducted additional interviews with seven staff in four PCOs to further explore issues relating to understanding the characteristics and needs of their local populations. Data were analysed using a combination of Framework analysis and the constant comparative method. Findings: Staff understood responsiveness as requiring alignment between service delivery and patient needs. These could be understood in terms of generic needs across the population, or as needs of diverse patient groups. From the staff interviews, we identified three components of responsive 32

The approach: We conducted a retrospective cross-sectional study in all English general practices with complete datasets for the year 2010/11, at the practice level. We obtained data on practice characteristics from the Health and Social Care Information Centre. We used the composite access indicator from the GP patient survey, with responses from 1,994,410 million patients (response rate 36%). We performed multiple linear regression modelling at the practice level, the outcome variable being the number of emergency admissions for an ACSC condition, weighted for a standardised population. The main explanatory variable was the composite access indicator from the GP patient survey. Other explanatory factors were list size; index of multiple deprivation; geographic location; nursing home patients; numbers of patients per GP; ethnicity; combined prevalence and performance in clinical care for ACSC conditions within QOF. Findings: 6618 practices (81%) had complete sets of data, with 670,800 admissions for ACSCs. The mean emergency admissions for an ACSC for a standardised practice population of 6790 patients was 105.6 patients per year, with substantial variation between practices (s.d. 47.2). The mean composite access score was 7.18 (scale 0-10, s.d.0.84). There was no significant association between admission rates and patient access scores. ACSC admissions were higher in populations that were more deprived, had a higher prevalence of ACSC conditions and a higher proportion of white persons. These associations were small. The model explained only 6.6% of the variation in ACSC admissions. Reanalysis of the data with inclusion of the Primary Care Trust of each practice as a categorical variable explained a further 7% of variation (R2 13.6). Consequences: There was wide variation in ACSC admission rates between practices which was not associated with patient access scores or clinical

ABSTRACTS OF ORAL PRESENTATIONS performance in ACSC conditions. 86.4% of the variation in ACSCs remained unexplained by practice factors. The lack of significant factors related to general practice amenable to change suggest that organisational approaches should be explored further at the commissioning and secondary care level, as hospital capacity has already been identified as a cause for variation in admissions. Limitations include ecological associations at the practice level which may not apply to individuals.

1 University of St Andrews, St Andrews, Fife, UK, 2University of Dundee, Dundee, Tayside, UK

The problem: Over 100 years ago, Flexner wrote ‘We have no right...to set up standards which will close the profession to “poor boys”(1). ‘. However medicine remains a profession dominated by those from a higher SES. Arguments for widening participation (WP) include: 1) improving social mobility 2) public health benefits and 3) educational benefits.

1D.3 An observational study of patient health status and use of NHS services after unintentional injury Blerina Kellezi, Darrin Baines, Denise Kendrick Nottingham University, Nottingham, UK The problem: Unintentional injuries account for more than 11,000 deaths, 660,000 hospital admissions, 5.5 million emergency department (ED) attendances and result in 1.8 million disability life years lost per year in the UK. 43% of deaths, 45% of admissions and 52% of ED attendances occur in working age adults. Little is known about how much injuries in working age adults cost the NHS, hence little priority is afforded to injury prevention and rehabilitation, despite the potential to improve the lives of injured people and release resources for use elsewhere. The approach: We undertook a multicentre longitudinal study measuring physical, psychological, occupational and economic impact of unintentionally injured adults. Participants were those admitted to hospital following an injury from acute NHS Trusts in 4 English study centres, followed up for 1 year post injury. Outcomes and resource use for care post discharge were collected from self-completed questionnaires. Costs were derived by assigning price weights to units of patient-reported resource use. Price weights were collected from published sources: BNF, NHS Reference Cost Schedule, PSSRU and ONS. Costs have been estimated over time (1, 2, 4 and 12 months); by secondary and primary care, injury and patient characteristics. Findings: Data were available for one year on resource use after discharge from the index admission on 328 participants. The most frequently used primary and community care services were GP consultations (used by 66%, range 0-20 visits, total 716), GP telephone consultations (43%, range 0-14, total 314), practice nurse consultations (35%, range 0-112, total 505) and physiotherapy (26%, range 0-36 total 420). The average cost of primary care services ranged from £138 in the first month to £165 between the fifth month and 1 year. The most frequently used secondary care services were ED (42%, range 0-5 visits, total 179), in-patient care (42%, range 0-38 nights, total 938 nights), physiotherapy (73%, range 0-9 visits, total 525) and fracture clinics (87%, range 0-8 visits, total 791). The average cost of secondary care services ranged from £1788 in the first month to £228 between the fifth month and 1 year. Injuries have important cost consequences for CCGs as commissioners of local patient services. This paper will present analysis that estimates the costs of injuries to CCGs. Consequences: Unintentional injuries in working age adults are common, with high NHS costs. 1/5 of costs are incurred in primary care and 4/5 in secondary care. Costs are highest in the first month but are still being incurred 12 months after injury in both settings. Further research is required to explore potential cost savings from targeting prevention and rehabilitation services at those likely to incur the greatest costs.

The potential public health benefits of WP are particularly pertinent for primary care. Significant health inequalities exist in the UK. A problem with underdoctoring of deprived areas is one part of the complex issues underlying the inverse care law (2). Increasing evidence suggests that individuals from minority backgrounds are more likely to practice medicine in an environment mirroring their ethnicity, culture or upbringing (3, 4). Increasing diversity in medical school has been suggested as one strategy to improve awareness and understanding of care of patients from deprived backgrounds. In the US students from lower income families have been shown to be more likely to pursue family medicine (5). Work (unpublished) by Dowell et al presented in the GMC report ‘Identifying best practice in selection’ demonstrated that doctors from deprived backgrounds were four times more likely to go on to work in one of the 100 most deprived practices (6). Understanding how we can widen participation in the UK may lead to future generations of GPs who are more ready, willing and able to practice in the most deprived areas. However there are difficulties around measuring SES (7, 8). Without a clear method for accurately measuring SES, it is challenging to know 1) how best medical schools should identify medical students from deprived backgrounds and 2) how primary care could monitor the make-up of its workforce to monitor the effectiveness of WP strategies in delivering more equitable GP provision. The approach: Statistical analysis of routinely collected data from the UKCAT (United Kingdom Clinical Aptitude test) Consortium and UCAS (Universities and Colleges Admissions Services) was undertaken to examine: 1) The distribution of four measures of SES in medical school applicants and medical school ‘accepted offers’. 2) How these measures compare with each other in measuring SES. Findings 1) There are more applicants from a higher SES regardless of the measure used. The number of accepted offers/ applicants falls in lower SES groups. 2) Analysis suggests only a weak association between SES markers. Consequences 1) WP remains a problem in the UK. Results suggest that applicants from a lower SES are less likely to obtain an accepted offer at medical school. On-going work is required both to attract a broader demographic of medical schools applicants and to review admissions processes. 2) We suggest that medical schools and governing bodies review carefully measures of SES used to monitor WP activities. 1E.1 Exercise for reducing fear of falling in older people living in the community: A Cochrane systematic review Arun Kumar, Denise Kendrick University of Nottingham, Nottingham, UK

On behalf of the Impact of Injuries Study Group. 1D.4

The problem: Fear of falling is widespread among older people and impedes quality of life. This systematic review examines the evidence to prescribe exercise to reduce fear of falling in older people.

Measuring Socio-Economic Status (SES) in Medical School Applicants Kathryn Steven1, Bruce Guthrie2, Jon Dowell2, Cathy Jackson1

The approach: Nine bibliographic databases were searched for published or current randomised control trials (RCT) and quasi-randomised trials 33

ABSTRACTS OF ORAL PRESENTATIONS (QRT) from 1887 to August 2013(Cochrane Bone, Joint and Muscle Trauma Group Specialised Register, the Cochrane Central Register of Controlled Trials, MEDLINE, EMBASE, CINAHL, PsycINFO, AMED, The WHO International Clinical Trials Registry Platform, and Current Controlled Trials). Trials were included where fear of falling was recorded as a primary or secondary outcome in community dwelling older people aged 65 years and older, not restricted to a specific medical condition, participants were provided with single component exercise interventions and control group included routine care which did not increase exercise. Reference lists of included studies were searched and experts in the field contacted to suggest relevant unidentified studies.

on-going. Two main themes emerged: 1) Self-management as a physical task A range of tools were used to self-manage pain and although all were found to be helpful by some participants in the studies reviewed, there was a preference for strategies that could be embedded into everyday life (e.g. household tasks) over more traditional strategies (e.g. exercise classes). 2) Self-management as an emotional task

Findings: Thirty studies met the inclusion criteria (2,578 participants, mean age 70 to 84 years, 70% women). Structured exercise programmes were associated with a significant reduction in fear of falling immediately post-intervention (SMD 0.24, 95% CI 0.14 to 0.34). There was no significant effect of exercise interventions on fear of falling beyond the end of the intervention period (three studies included data up to six months and two included data at six months and beyond). Structured exercise programmes were associated with a significant reduction in the rate of falls (pooled rate ratio 0.70, 95% CI 0.56 to 0.89); however, there was no significant effect on depressive symptoms (four studies), self-reported physical activity levels (four studies) or adherence rates (16 studies).

A number of sub-themes are emerging with on-going analysis.

Consequences: Exercise is an effective means to reduce fear of falling in older people living in the community whilst the exercise is being prescribed but there is a lack of evidence on the duration of the effect. 1E.2 Older people’s experiences of self-management strategies: a systematic review and thematic analysis Clare Clarke1, Derek Jones2, Denis Martin3, Pat Schofield4, Paul McNamee5, Amanda Clarke1, Blair Smith1 1 University of Dundee, Dundee, UK, 2Northumbria University, Newcastle, UK, 3Teeside University, Middlesbrough, UK, 4University of Greenwich, London, UK, 5Aberdeen University, Aberdeen, UK The problem: Chronic pain is a major health condition associated with ageing. Effective self-management is important, but presents particular challenges for older people; we know little about these. Research is required to identify which self-management strategies may be effective in this group, and how they may be implemented successfully. The approach: We aimed to characterise older people’s experiences of self-management in chronic non-malignant pain, to identify successful approaches, and factors that facilitate or prevent their implementation. The search was designed to find studies published in peer-reviewed journals between 1990 and 2012 limited to the English language. The review considered non-experimental, observational and qualitative designs which drew upon the self-management of chronic pain in older adults (=>65 years). Self-management was defined as, “A single approach or combination of approaches that can be initially taught by any health professional or learned by an individual to enable them to minimise the impact their chronic pain can have on everyday life.” Papers meeting the inclusion criteria were quality appraised using the Critical Appraisal Skills Programme (CASP) checklist for qualitative studies. Thematic analysis was undertaken independently by two researchers and then consensus on main themes and sub-themes agreed. The Findings: Iterative analysis of the findings of the papers reviewed is 34

The work of self-management is evidently experienced as a complex process involving cognitive adjustment and adaptation to living with chronic pain.

Consequences: Self-management for chronic pain is facilitated when it is embedded in daily life and acknowledged as an on-going task needing support. Further in-depth analysis will highlight specific factors to be addressed, to promote a customised approach to self-management that meets the needs of the older adult living with constant pain. 1E.3 Does a complex intervention by primary care nurses increase walking in 60-75 year olds? Findings from the PACE-Lift (Pedometer Accelerometer Consultation Evaluation - Lift) cluster randomised controlled trial. Tess Harris1, Sally Kerry2, Christina Victor3, Ulf Ekelund6, Alison Woodcock7, Steve Iliffe4, Peter Whincup1, Carole Beighton8, Michael Ussher1, Elizabeth Limb1, Lee David9, Debbie Brewin9, Annabelle Rogers1, Rika Adams1, Derek Cook1 1 St George’s University of London, London, UK, 2Queen mary’s University of London, London, UK, 3Brunel, University of London, London, UK, 4 University College, London, London, UK, 5Cambridge University, Cambridge, UK, 6Norwegian School of Sports Science, Oslo, UK, 7Royal Holloway, London, UK, 8South Bank University, London, UK, 910 minute CBT, Hertfordshire, UK The problem: Physical activity is vital for older peoples’ physical and mental health, yet only a small proportion achieve the guidelines recommended for health. Walking is older people’s commonest physical activity (PA), light intensity if strolling, moderate if brisker. We assessed whether a primary care nurse delivered complex intervention increased older peoples’ walking. The approach: 2-arm parallel design, cluster randomised controlled trial with household as the randomisation unit. 988 people aged 60-75 years were randomly selected from 3 Oxfordshire and Berkshire UK general practices and invited to participate, of whom 298 people from 250 households were allocated by internet randomisation to intervention (150 individuals) or usual care control group (148 individuals). Participants had no restrictions to increasing their walking. The intervention group received four practice nurse PA consultations incorporating behaviour change techniques, feedback on pedometer step-counts and accelerometer PA intensity, an individual PA diary and plan. Outcomes: Change in average daily step-counts over 7 days (primary outcome) and average time spent in at least moderate intensity PA weekly (MVPA) (secondary outcome) between baseline, 3 and 12 months, assessed by accelerometry. Other outcomes included adverse events, anthropometry, depression, anxiety and pain scores. Assessors were not blind to group allocation. Findings: Primary outcome data available for 280 (94%) individuals. Mean (s.d.) baseline average daily step-count was 7453 (2463) in control and 7388 (2218) in intervention groups. At 3 months both average daily step-counts and average weekly MVPA increased significantly in the intervention compared to control group: daily step-count difference 1037 (95% C.I. 513, 1560), weekly MVPA difference 66 minutes (95% C.I. 36,

ABSTRACTS OF ORAL PRESENTATIONS 96). At 12 months the differences were 609 steps per day (95% C.I. 104, 1115) and 40 minutes of weekly MVPA (95% C.I. 10, 70). There were no significant between group differences in other outcomes. Consequences: The PACE-Lift trial is the largest pedometer-based walking intervention in older people to date. It showed an increase in both step-counts and time spent in MVPA, nearly all within 10 minute bouts, in the intervention compared with the control group at 3 months, with no effect on adverse events. We also demonstrated over half of the 3 month increase was still present at 12 months, 9 months post-intervention. While the intervention itself, particularly the use of accelerometer feedback, is of novel interest, the key methodological development was the successful embedding of the intervention within primary care. 1E.4 Why is scabies such a challenge in residential care homes for the elderly? A report of outbreak investigations in the South East of England Jackie Cassell1, Ananth Nalabanda3, Kirsty Hewitt2 1 Division of Primary Care and Public Health, Brighton and Sussex Medical School, Brighton, UK, 2Kent, Surrey and Sussex Public Health England Centre, Sussex, UK, 3Centre for Primary Care and Public Health, Blizard Institute, Barts and the London School of Medicine and Dentistry, London, UK The problem: Scabies is an important public health problem in residential care homes for the elderly. Scabies in immunocompromised and elderly patients may present atypically and this can lead to late recognition of outbreaks in care homes. Early diagnosis of scabies can be challenging, and misdiagnosis lead to prolonged outbreaks which are difficult to control. Scabies is usually diagnosed on the basis of clinical signs, or less commonly by microscopic examination of skin scrapings, or dermatoscopy. The approach: We aimed to describe the initial recognition and management of scabies outbreaks in care homes, including barriers to early diagnosis and treatment, and identify areas for intervention. A semi structured survey of managers, affected residents and staff of care homes reporting suspected scabies outbreaks to Surrey and Sussex Health Protection Unit between November 2012 and April 2013 was undertaken.

1F.1 “When Learners Become Teachers” - Exploring Educational Learning Needs of General Practice Trainees Who Teach Undergraduate Medical Students Kerry Boardman King’s College London School of Medicine, London, UK The problem: There is an expectation by the UK General Medical Council (GMC) for doctors to teach, and it is one of the core areas in the Royal College of General Practitioners (RCGP) curriculum. With increasing amounts of undergraduate medical education in the UK and elsewhere taking place in the community, there is also a concurrent need for additional teaching capacity. Involvement of GP trainees in teaching is one way to help meet this need. This is increasingly used and evaluated in Australia, but there is relatively little UK based evidence in relation to the amount of undergraduate medical student teaching that GP trainees deliver, or the learning needs of GP trainees regarding teaching. What does exist, suggests that GP trainees want to be involved with teaching, but have variable opportunities to do so and feel the need for both educational and organisational support to facilitate this. What support is appropriate for this specific group of tutors, who are themselves trainees, has not been determined. The approach: Our faculty has started to work with General Practice Specialty Trainees (GPSTs) to enable them to teach our undergraduate medical students. As part of this change, we were keen to explore GPSTs’ conception of teaching and assess their educational learning needs in order to improve future training and support for any GPSTs who teach for our team, and to add to the current body of knowledge about GPSTs as tutors. Data have been collected via: F An initial questionnaire exploring GPSTs’ previous teaching and learning experience, beliefs about teaching, and perceived learning needs with regards to training for a teaching role F Focus groups to further expand these ideas F Observation of teaching exercises led by the GPSTs, with contemporaneous observer field notes to identify additional learning needs unknown to the GPSTs but recognised by medical school faculty as experienced educators

Findings: Seven outbreaks were reported, involving 39 resident cases and 29 staff. 26/39 (72%) resident cases were female. Three cases of crusted scabies were detected in two homes. Attack rates among residents ranged from 2-50%, while most cases (37/39, 95%) had dementia. The median time from symptom onset to diagnosis was 7 days (IQR 2-47), and diagnosis to first scabicidal treatment 4 days (IQR 2-6). Time from first to second treatment ranged from 6-18 days. All cases were diagnosed by GPs (59%) or care home staff (41%), none by a dermatologist. Longer delays in diagnosis were experienced by incontinent residents (median 8 vs. 4 days); mobile residents (20 vs. 6 days); and those without dementia (64 vs. 7 days). 2/29 staff reported symptoms in their household contacts. The coordination of mass treatment was time consuming and logistically difficult, requiring liaison with several prescribers and pharmacies. Prophylaxis prescribing was inconsistent, and insufficient treatment lotion was often prescribed.

Findings: Data analysis is currently underway, using thematic analysis of free text survey comments, focus group transcripts and faculty observer teaching field notes. Early findings suggest that GPSTs draw on their own recent and continuing education when constructing their conceptions of learning and teaching and that these conceptions appear to be sophisticated and in line with the learner-centred, constructivist approaches to learning we aim to encourage. Further findings from the remainder of the analysis will be presented and discussed.

Consequences: Avoidably late recognition of index cases occurred in most outbreaks. Several were not diagnosed with scabies until further cases had arisen, and symptoms confused with other skin conditions such as eczema. Our findings illustrate the difficulty of diagnosing scabies for non-specialists and the need to support care homes in in early and accurate diagnosis through specialist support and novel diagnostic tools for primary care. Improved diagnostic approaches and systemic treatment are research priorities.

Medical Undergraduate Primary Care teaching across the UK: what is being taught?

Consequences: Vertical integration of teaching has the potential to ease the problem of GP tutor recruitment for teaching and to provide valuable experience for trainees and students. Ascertaining trainees’ educational learning needs should help tailor appropriate faculty development for this group. 1F.2

Veronica Boon, Andrew Blythe, Matthew Ridd University of Bristol, Bristol, UK The problem: UK medical schools all use Primary Care settings to deliver their undergraduate courses. However, unlike some medical specialties, there is no national undergraduate Primary Care curriculum and it is 35

ABSTRACTS OF ORAL PRESENTATIONS thought that the nature and purpose of Primary Care teaching varies considerably between medical schools. The aim of this project is to establish what is being taught about Primary Care across the UK medical schools, identify common themes and compare this with the postgraduate RCGP Primary Care curriculum. The approach: We wrote to all 32 UK medical schools outlining the aims of the project and requesting Primary Care curricula handbooks. A reminder email was sent individually to each University that did not respond. Information was extracted, based on key headings from the RCGP postgraduate curricula, into an Access database. Initially, two researchers independently extracted the aims and objectives and relevant information from the core text from two university handbooks. Data were compared and any discrepancies discussed with a third researcher to achieve consensus. A further two sets of handbooks were mapped individually by two researchers to confirm agreement. The remainder were mapped by one researcher. Findings: 30 UK medical schools supplied relevant documents. The majority map their Primary Care curriculum to GMC’s Tomorrow’s Doctors with only 20% mapping to the RCGP postgraduate curriculum. However, 96% of medical schools cover over half of the RCGP postgraduate curriculum headings and 43% cover over 80% of the curricula. Topics taught at all medical schools include: Consulting and communication skills, Leading and working in teams, Holistic care, and Developing yourself and others. Other common topics include: Chronic disease, and Health promotion and prevention. 80% of medical schools teach about specific disease presentations with 60% having core problem lists. Primary Care is integrated across the whole medical curriculum at a third of universities. Novel topics include: Learning difficulties, Genetics, and Multi-morbidity. Ethics and Public Health are recurrent parallel themes. Consequences: This project provides valuable information about Primary Care teaching at an undergraduate level across the UK. Although it confirms a widespread variation in learning outcomes, it also highlights significant common ground. This initial work could be used to help develop guidance for a national undergraduate Primary Care curriculum that could create a ‘joined up’ spiral curriculum extending from early undergraduate years to postMRCGP general practice. The database developed could also be used by medical schools as a tool to develop their curriculum and could encourage collaboration between medical schools to share resources and develop new Primary Care teaching. Further research is needed to explore current undergraduate Primary Care teaching in more depth and develop a consensus for an undergraduate Primary Care curriculum. 1F.3 Meta-ethnography of student and patient perspectives of undergraduate medical education in the UK general practice setting Sophie Park1, Nada Khan1, Alice Malpass2 1 University College London, London, UK, 2University of Bristol, Bristol, UK The problem: Up to 10-15% of undergraduate medical education in the UK is conducted in a general practice setting. We are using meta-ethnography to synthesise empirical qualitative evidence of student and patient perspectives, identified during a systematic review of this field. The approach: We systematically searched bibliographic databases using general practice and medical education terms. From 169 included empirical studies, 74 qualitative studies were identified. We adapted Dixon 36

Woods’ categorisation of relevance, quality and theoretical utility to independently rate each paper. We then used Noblit and Hare’s 7 stage process to map and translate concepts across included papers. We independently developed empirical verbatim data (first order) and author interpretation (second order) constructs to develop reviewer (third order) interpretations. This process allows researchers to develop new insights and knowledge by making connections across included papers. This also identifies theoretical and conceptual gaps within the existing literature. Findings: We identified 10 key papers describing the student and patient perspective of undergraduate teaching in general practice. Two main thematic groups emerged from the paper: group 1 described the interpersonal interactions within the teaching consultation and group 2 described sociocultural spaces of learning which shape those interactions. Group 1 papers emphasised the importance of the GP as a broker who influenced the triadic/dyadic interactions between patients and students through setting the ‘stage’ of the teaching consultation; maintaining a scaffold for learning during the teaching encounter; and overseeing patient consent for participation in teaching. The papers highlight that triadic learning within the teaching consultation emphasise that ‘flat’ participatory learning is beneficial for the student and for the patient. Group 2 papers represent qualitative research exploring the socio-cultural spaces of learning that shape and provide boundaries for the interpersonal interactions which occur within the teaching consultation. These papers suggest that students act as mediators between the two polarised environments of hospital and general practice teaching, and act to integrate the knowledge and learning of these competing cultures. General practice was seen as a setting where students can make links between affective, cognitive and practical aspects of learning. However, students feel a tension between the ‘real world’ medical training offered in general practice, which prepares students for their work as doctors, versus the ‘textbook medicine’ offered in the hospital setting and which students felt dominated their exams and assessments. Consequences: GPs need to be aware of their role as a powerful broker of interactions between patients and students within the teaching consultation, facilitating legitimate and effective participation of students and patients in teaching interactions. This has implications for effective training of GP tutors. Students are managing competing cultures between general practice and hospital settings, and need support to address their expectations and learning within general practice placements. 1F.4 Rapid appraisal of a community based element of a revised undergraduate curriculum - The UCL Medicine in the community course Melvyn Jones UCL, London, UK The problem: UCL implemented a new curriculum in 2012. As part of the revision to the course the GP medicine in the community course was extended to run over the whole 1st clinical year across a range of specialities with all students constantly rotating out of hospital to GP placements. The approach: We used a series of measures to assess the delivery of the GP curriculum from various stakeholders. We used traditional online student evaluation (SEQs), but also used emails (students, GP tutors, hospital clinicians) and module management group minutes as a source more immediate feedback. Where problems were identified these were rapidly addressed with students or tutors. Findings: We placed 350 students rotating out to 40 General Practices. Initial oral feedback from students was negative, which was amplified by hospital clinicians who were hostile to the curriculum changes eg “my students are never on the ward”. Student emails also initially suggested

ABSTRACTS OF ORAL PRESENTATIONS logistical difficulties with the changes. However, once we were in receipt of formal feedback (SEQs) from the whole student body we identified that the GP course was very well received (89% satisfied or very satisfied) and this data helped counter balance more vocal detractors of the reforms.

instrument will be analysed through a two-stage pilot-test involving 10,000 patients from 50 practices across England.

Consequences: As regulators (GMC) push UK medical schools to increase students’ community exposure methods to rapidly assess the success or otherwise of GP placements are required. Use of student leaders, email and verbal interaction with stakeholders are required. However care must be used with such sources as it may only reflect a narrow view and broader more systematic approaches can reveal a different (an in our case) more favourable view.

Examining the associations between quality of care, patient satisfaction and health outcomes in primary care in England: a structural equation model approach

2A.1 Development of a measure of Patient Reported Experiences and Outcomes of Patient Safety in Primary Care: the PREOS-PC instrument Ignacio Ricci-Cabello1, Anthony J Avery3, Jose M Valderas1 ,3 Department of Primary Care Health Sciences, University of Oxford. Oxford, Oxford, UK, 2University of Exeter Medical School, Exeter, UK, 3Division of Primary Care, School of Community Health Sciences, University of Nottingham Medical School, Nottingham, UK

The problem: Despite the enormous potential for adverse events in primary care, the knowledge base about patient safety in this context is still sparse. The lack of appropriate measurement methods is one of the main factors limiting the research in this field. We aimed to develop a patient reported instrument to measure patient safety experiences and outcomes in primary care in England. The approach: First we conducted a Systematic Literature Review for all available instruments measuring patient reported experiences and/or outcomes of patient safety in Primary Care, alongside a systematic review and meta-synthesis of qualitative studies examining patients’ experiences of safety in Primary Care. We also conducted four focus groups to identify domains and themes related to patient safety in primary care as perceived by the patients. We developed a conceptual framework identifying relevant constructs. We then identified all the relevant items from the previously identified instruments and selected the best candidates for inclusion in the new tool. Through an iterative process, we reviewed the selected items and introduced new items, and/or modifications to the items and/or the response scales with the help of an Expert Panel and with subsequent rounds of cognitive testing with patients. Findings: The meta-synthesis of qualitative studies included fifty-five articles. Main themes identified were patients’ experiences or concerns about 1) adverse drugs events, 2) inadequate patient-provider communication, and 3) diagnosis delay or misdiagnosis. Fourteen themes emerged from the focus groups, most of them related to factors contributing to the occurrence of adverse events. The systematic review of instruments identified 23 patient reported instruments to measure patient safety in primary care. Four main domains emerged: 1) patient safety culture, 2) perceptions of patient safety, 3) experiences of patient safety, and 4) patient safety outcomes. No instrument covering these four aspects was identified. As a result of multiple iterations of expert review and cognitive testing we developed the “Patient Reported Experiences and Outcomes of Safety in Primary Care” (PREOS-PC). This instrument contains 82 items covering the four main domains above mentioned. Consequences: PREOS-PC is the first patient reported instrument to measure patient safety experiences and outcomes in primary care. It has been developed using state of the art methods for instrument development that support its face validity. Further psychometric evaluation is needed. Feasibility of administration methods and psychometric properties of the

2A.2

Ignacio Ricci-Cabello1, Sarah Stevens1, Andrew R Dalton1, Robert I Griffiths1, Jose M Valderas1 ,2 1 Department of Primary Care Health Sciences, University of Oxford. Oxford, Oxford, UK, 2University of Exeter Medical School, Exeter, UK The problem: Providing high-quality clinical care is a clear priority for most health care systems, including the National Health Service (NHS). However, the nature of the associations between the different components of quality of care, and their impact on patients’ satisfaction and health remains unclear. The aim of this study was to analyse the relationships between quality of care, patient satisfaction and health outcomes in general practices in England. The approach: We developed a conceptual framework to describe hypothesised relationships between quality of care, patient satisfaction and health outcomes. We subsequently conducted a cross-sectional study to empirically test the hypothesised associations. Practice level indicators for the period 2011/2012 were obtained from the Quality of Outcome Framework (QOF) and the GP Patient Survey for 7,607 practices in England. All the indicators were examined and classified into ten categories, which reflected areas of quality of care (Quality and Productivity, Education and Training, Medicine Management, Access to Additional Services or Access to Surgery, Clinical quality and Patient centred care), health outcomes (intermediate outcomes and patientreported health status) and patient satisfaction. Statistical analysis consisted in a hybrid structural equation model that combined confirmatory factor analysis and path analysis. Confirmatory factor analysis was used to create ten latent variables corresponding to the ten groups of indicators previously mentioned. Path analysis was used to examine the hypothesised associations between the latent variables. Findings: The Structural Equation Model suggested the existence of two distinct paths. In the first path, “Access to the Surgery” was associated to “Patient Centred Care” (ß=0.61), which in turn was strongly associated to “Patient Satisfaction” (ß=0.88). In the second path, “Education and Training” was associated to “Clinical Quality” (ß=0.34), which in turn was strongly associated to “Intermediate Outcomes” (ß=0.80). These two paths were substantially independent, with weak associations being observed between “Clinical Quality” and “Patient Centred Care” (ß=0.09), and between “Clinical Quality” and “Patient Satisfaction” (ß=0.05). Finally, “Patient-reported health status” was weakly associated to “Clinical Quality” (ß=0.01), “Intermediate Outcomes” (ß=0.02), and “Patient Centred Care” (ß=-0.05). The model showed good model fit (standardised root mean squared residual =0.077). Consequences Our findings suggest that evaluation of performance must take into account these three aspects of quality that seem to be rather independent from each other: clinical quality, patient experiences and patient reported outcomes. 2A.3 Does Participation in a Quality Improvement (QI) Programs Improve Primary Healthcare Team (PHC) Functioning?-A Qualitative Study. Jyoti Kotecha1, Judith Belle Brown2, Stewart B. Harris2, Grant Russell4, Michael Green1, Susan Webster-Bogaert2, Meghan Fournie2, Amardeep 37

ABSTRACTS OF ORAL PRESENTATIONS Thind2, Moira Stewart2, Sonja Sonja Reichert2, Sharon Roberts3, Richard V, Birtwhistle1 1 Queen’s University, Kingston, Ontario, Canada, 2Western University, London, Ontario, Canada, 3University of Waterloo-Renison University College, Waterloo, Ontario, Canada, 4Monash University, Melbourne, Australia The problem: Primary Healthcare team functioning can impact the quality of patient care. Quality improvement (QI) programs are frequently implemented to support primary healthcare (PHC) team development and to improve care outcomes. A common approach is the use of QI learning collaboratives (LC), in which members from different teams attend joint education sessions prior to implementing lessons into practice. In Ontario, Canada, the Quality Improvement and Innovation Partnership (QIIP), now amalgamated into Health Quality Ontario, offered a primary healthcare LC program. This program was offered to support the development of interdisciplinary team functions, and to improve chronic disease management, disease prevention and access to care. LC consisted of 3 learning sessions, 2 action periods and 1 congress. Teams were supported by QI coaches. The approach: The objective of our study was to explore the influence of the LC program on team functioning in primary healthcare teams. For this purpose a qualitative study was conducted as part of a mixed method third party evaluation. Data was collected through semi-structured telephone interviews that were audiotaped and transcribed verbatim. Transcripts were analyzed using thematic analysis through an iterative and interpretive approach. Participants: PHC participants (n=31, clinicians and administrators) QI Coaches (n=7), and the LC administrators (n=4). A stratified purposeful sampling strategy was used to identify PHC teams (n=10), from which these individuals were selected using a purposeful recruitment process, to maximize variation of team roles. Additionally, the QI coaches working with the interview participants and the LC administrators were interviewed. Findings: The shared experience of participating in the program contributed to improved team functioning. Participants reported increased trust and respect for each other’s clinical and administrative roles and became inspired through learning about different approaches to interprofessional care from other teams. This lead to enhanced collegial relationships, collapsing of professional silos, improved communication, and increased interprofessional collaboration. Consequences: Participation in LC appeared to enhance the team functioning of these PHC teams. 2A.4 Never Event in UK general Practice: A cross-sectional survey to determine the incidence and likelihood Sudeh Cheraghi-Sohi1, Jill Stocks1, Carl De Wet2, Paul Bowie2, Rahul Alam1, Stephen Campbell1 1 Greater Manchester Primary Care Patient Safety Translational Research Centre, University of Manchester, Manchester, UK, 2NHS Education for Scotland, Glasgow, UK The problem: Never Events are defined as serious, largely preventable patient safety incidents that should not occur if the available preventable measures were implemented by healthcare workers. Never Event lists and policies have been developed and implemented for acute hospital settings in many national healthcare systems, including the NHS in England & Wales. The concept of Never Events in Primary Medical Care however has thus far not been addressed. A preliminary list of 10 Never Events for the primary medical care context has been developed through a professional 38

consensus exercise and published by the authors. The identified events all have the potential for catastrophic consequences for some patients. For example, the event regarding the planned referral of a patient, prompted by clinical suspicion of cancer not being sent, would have implications for delaying a patient’s diagnosis. A systematic approach to reducing the occurrence of Never Events would therefore be valuable. However, the incidence of these specific Never Events and their likelihood of occurrence are unknown. Determining GP’s estimations of the frequency of these Never Events occurring is the next step in evaluating this approach as a safety improvement intervention. The approach: The study design comprises a cross-sectional survey of General Practitioners (GPs). The survey will be administered concurrently in two sites: in the Greater Manchester (GM) area of England and Scotland. All GPs in 527 practices in GM (approximately 2700 GPs, excluding locums) and all GP Educational Supervisors (GPESs) in Scotland (approximately 700 or a quarter of all Scottish GPs) have been invited to participate. Data collection is underway and will be completed by April 2014. The analysis will use SPSS. Findings: The response rate and frequencies for each never event per GP responding will be reported. Differences between GP responses will be compared in terms of location, level of experience and gender. For the GM practices incidence rates per patient will be estimated using the mean frequency of never events reported per practice as the numerator and consultation rates and GP practice list sizes as the denominator. In addition, the free text responses will be analysed to provide a sense of whether GPs recognise and agree with the proposed events as well as providing a space for other suggested events. Consequences: This work is the first to investigate the concept of Never Events amongst frontline UK GPs. It provides an indication of whether the proposed list of never events is recognisable to GPs in the UK as well as providing data on the occurrence of these events. The work could provide a platform for policymakers to take forward the development of a formal Never Events list akin to those found in acute hospital settings. 2A.5 Health needs and services for homeless people in Glasgow: a qualitative study of support and accommodation providers’ views Sara Macdonald, Jamie Stewart, Andrea Williamson Department of General Practice and Primary Care, University of Glasgow, Glasgow, UK The problem: In the UK, homelessness is acknowledged as a major social issue, evolving as society changes. There is accumulating evidence that homelessness tends to be experienced by people who have had a series of adverse life events and it is those that are deeply socially excluded that tend to ‘fall though the gaps in policy and services’ into more extreme forms of homelessness (Fitzpatrick, 2011). This understanding of homelessness as a complex form of exclusion has led to different responses from services, for example, hostel decommissioning and a specialist homeless health service in Glasgow. However, evidence remains sparse about whether these changes have brought improved access and utilisation of health services by people who are homeless. Furthermore, our theoretical understanding of what underlies these issues is limited. The approach: Using qualitative methods we are exploring the views of support and accommodation providers about the health needs and current health service provision for homeless people in Glasgow. Using the social theory of candidacy, which describes the process of access to health services, we are exploring the ways in which people’s eligibility for medical

ABSTRACTS OF ORAL PRESENTATIONS attention and intervention might be negotiated between individuals and health services, and may help explain access to, and utilisation of, those services. Exploring our participants’ views allows insight into this process as they negotiate what we term ‘candidacy by proxy’; acting as a conduit between homeless people and health services that facilitates access to health services. We are conducting semi-structured interviews with a purposive sample of 10 participants. They have been chosen because they work with a range of homeless groups in Glasgow including rough sleepers, young people, people with substance misuse problems and people who are in more stable homelessness accommodation. Our results are being analysed using Framework analysis. Findings: Preliminary findings suggest that support and accommodation providers have a crucial role in identifying homeless people’s candidacy and subsequent navigation of health services. They add weight to our idea that support and accommodation providers adopt a unique role in the context of homeless people apparently resisting engagement with health services. This sheds light upon the socially constructed process that leads to health service intervention. Furthermore, there is evidence of a complexity of multiple candidacies that homeless people have when accessing and utilising health services. Consequences: Extant research using the theory of candidacy has involved secondary data analysis. It has not been examined in the context of homelessness, with a vulnerable group of health service users who have multiple needs and identities. We hope this will build on the model of candidacy described by Dixon-Woods, deepen our understanding of health needs and current health service provision for homeless people in Glasgow and inform a future programme of work about homelessness health. 2A.6 Withdrawing Performance Indicators: Retrospective Analysis of General Practice Performance Under the UKs Quality and Outcomes Framework. Evangelos Kontopantelis1, David Springate1, David Reeves1, Darren Ashcroft1, Jose Valderas2, Tim Doran3 1 University of Manchester, Manchester, UK, 2University of Exeter, Exeter, UK, 3University of York, York, UK The problem: In the UK, the Quality and Outcomes Framework (QOF) incentive scheme provides family practices with financial rewards linked to performance on a range of over 100 quality of care indicators, mostly related to processes of care for common chronic conditions. The QOF is reviewed annually in a process currently overseen by the National Institute for Health and Clinical Excellence (NICE), which makes recommendations for individual indicators to be modified or removed. Final agreement on changes to indicators is reached in negotiations between the Department of Health and the British Medical Association. For the third year of the QOF (2006/7) three clinical indicators were removed; a further eight were removed for the eighth year (2011/12). Performance for removed indicators is not monitored in the central Quality and Management Analysis System database. The aim of this study was to assess the effect that removing the incentives for these indicators from the QOF scheme had on subsequent performance, both on performance as measured by the same indicator and on related indicators. The approach: We used a retrospective longitudinal design and data from 644 general practices, from 2004/5 to 2011/12, extracted from the Clinical Practice Research Datalink. Data were available for all patients registered with any of the practices over the study period, 13,772,992 in total. We monitored performance on eight clinical quality indicators withdrawn from a national incentive scheme: influenza immunisation (asthma) and lithium therapy monitoring (psychosis), removed in April 2006; blood pressure monitoring (coronary heart disease, diabetes, stroke), cholesterol level monitoring (coronary heart disease, diabetes) and blood glucose

monitoring (diabetes), removed in April 2011. Multilevel mixed effects multiple linear regression models were used to quantify the effect of incentive withdrawal. Findings: Mean levels of performance were generally stable after the removal of the incentives, both short- and long-term. For the two indicators removed in April 2006, levels in 2011/12 were very close to 2005/6 levels, although a small but statistically significant drop was estimated for influenza immunisation. For five of the six indicators withdrawn from April 2011, there was no significant impact on performance following removal and differences between predicted and observed scores were small. Performance on related outcome indicators retained in the scheme (e.g. blood pressure control) was generally unaffected. Consequences: Following the removal of incentives, levels of performance across a range of clinical activities generally remained stable. This indicates that health benefits from incentive schemes can potentially be increased by periodically replacing existing indicators with new indicators relating to alternative aspects of care. However, all aspects of care we investigated remained indirectly or partly incentivised in other indicators and further work is required to assess the generalisability of the findings when incentives are fully withdrawn. 2B.1 The SIMPle study: Supporting the Improvement and Management of Prescribing for UTIs Akke Vellinga1, Sinead Duane1, Sandra Galvin1, Aoife Callan1, Andrew W Murphy1, SIMPle Team1 ,2 1 NUI Galway, General Practice, Galway, Ireland, 2University Hospitals Galway, Galway, Ireland Problem: Antibiotic resistance poses a threat to our healthcare system. Improving antibiotic prescribing can contribute to addressing this problem. A multidisciplinary team from epidemiology, social marketing, health economy and microbiology developed an intervention in primary care to improve the quality and quantity of antibiotic prescribing for urinary tract infections (UTIs). UTIs are frequent consultations for General Practitioners (GPs) and typically antibiotics are empirically prescribed even though less than 40% are in accordance with prescribing guidelines. Approach: The SIMPle study offers GPs interactive workshops, audit and feedback reports and automated electronic prompts summarising recommended first line antibiotic treatment and, in a third intervention arm, provide GPs with tools to improve communication to promote delayed antibiotic prescribing. For patients, multimedia applications (game and infographic) are available. Novel data collection methods include text messaging, a smart phone app and remote data extraction from the GP practice software. Findings: 30 Practices were recruited into three groups: control, improved prescribing and improved, delayed prescribing. Primary outcome is improved quality of prescribing and secondary is reduced antibiotic consumption due to delayed prescribing. Interim analysis at 5 months showed more than 700 enrolled patients of whom a third provided additional information on antibiotic consumption and symptoms by text. Antibiotic prescribing during this period improved by at least 15% in the intervention group. Consequences: The SIMPle study is a complex intervention to which shows improved prescribing for UTIs in the community. Patient and GP participation is high. The novel, e-health data collection and multimedia applications support the delivery and evaluation of this complex intervention.

ABSTRACTS OF ORAL PRESENTATIONS 2B.2 Developing A Taxonomy of Techniques to Harness Placebo Effects in UK Primary Care Felicity Bishop, Beverly Coghlan, Adam Geraghty, Hazel Everitt, Paul Little, George Lewith University of Southampton, Southampton, Hampshire, UK The problem: There is growing interest in developing a translational research agenda to harness placebo effects in clinical practice. This interest has been driven by recent systematic reviews that confirm placebo effects can be as large in magnitude as treatment effects. However, the very notion of harnessing placebo effects in clinical practice is contentious and the use of placebo treatments is ethically dubious outside of specific situations in clinical trials. Furthermore, it is not at all clear what procedures could be used in clinical practice to harness placebo effects to benefit patients. We therefore aimed to develop a taxonomy to organise different ways of harnessing placebo effects and help progress research in this area. The approach: There were two phases. In phase 1, a systematic review of clinical trials and placebo studies was conducted to identify procedures that might elicit placebo effects. We conducted a secondary analysis of 124 primary studies from recent systematic reviews of the placebo effect focusing on conditions where pain is the primary outcome. All procedures that placebo recipients in these studies were subjected to were extracted, categorised and described in a table. For each procedure, we counted the total number and proportion of primary studies that reported using it, gave an illustrative example, and suggested translational adaptations that resulted in techniques for use in clinical practice. In phase 2, we obtained the views of patients, clinicians and researchers on six techniques to elicit placebo effects in primary care. We ran five groups with 20 primary care patients and four groups with 22 primary care clinicians, and surveyed 21 placebo researchers. Results: The taxonomy consists of 33 techniques that could be used to harness the placebo effect in UK primary care. These techniques are derived from procedures used in clinical trials and experiments on placebo effects. They are grouped under the domains: the patient (n=10 techniques); the practitioner (n=4); the healthcare setting (n=4); the treatment (n=8); the patient-practitioner relationship (n=7). GPs and patients expressed diverse views but there was some indication that techniques based on changing how GPs interact with patients were more acceptable than techniques which involved deception or using placebo treatments. Consequences: This taxonomy can be used to guide further research on the factors that influence the magnitude of placebo effects and the acceptability of translational efforts to harness placebo effects in primary care to benefit patients. Currently it is not possible to comment on the potency of each item in the taxonomy but we hope that it will help improve our ability to address such questions in future systematic reviews. Some methods for harnessing placebo effects in primary care may be acceptable to GPs and patients and should be prioritised in future research.

The problem: Domestic violence is a global challenge with significant health and social consequences for victims, their children, and families. A key characteristic of primary care is that it is family centred with equal caring responsibilities for all family members. In the UK, primary care professionals have been criticised for insufficient engagement in domestic violence and child safeguarding, but there is uncertainty about what constitutes best practice at the interface of these two issues. This project aims to explore the primary care response to domestic violence in families with children and the barriers to engagement The approach: We conducted qualitative semi-structured telephone interviews with a purposive sample of 42 GPs, 12 practice nurses, and 15 practice managers. Interviews were conducted across six sites in England, chosen to represent areas with different domestic violence services. We used vignettes to explore practitioners’ experiences of identifying and responding to domestic violence and child safeguarding and the relationship between them. We also elicited views and experiences of interagency collaboration and training were also elicited. With consent, telephone interviews were audio recorded and transcribed verbatim. We analysed the data from multiple disciplinary perspectives using framework analysis. The findings: Primary care professionals are ambivalent with regards to the appropriate management of domestic violence in families with children. There was considerable variation in professionals’ responses to the vignettes, both within and between professional groups. Professionals were conflicted about documenting domestic violence in the notes of multiple family members. Most professionals were aware this is good practice in cases of child safeguarding. However, some were concerned documenting domestic violence could lead to potential breaches of confidentiality making the situation worse for both the victim and their children. This resulted in a multitude of approaches to documenting that varied within practices and across the sample. Professionals also varied in response to sharing information. They were keen to discuss cases and share information within their own practice or with named local colleagues; they did not describe this as breaking confidentiality. Most professionals agreed it was acceptable to break confidentiality and refer to social services in cases where children were at risk. However, professionals had different perspectives on the level of risk for children in families experiencing domestic violence. Consequences: The lack of clear guidance and training on the interface between domestic violence and child safeguarding is resulting in variation in primary care practice and uncertainty in managing these families. This is likely to be the case internationally. The findings are informing the design and content of an interagency training intervention aiming to increase professionals’ confidence and skills in managing these complex issues, and improving interagency working. This report is independent research commissioned and funded by the Department of Health Policy Research Programme (Bridging the Knowledge and Practice Gap between Domestic Violence and Child Safeguarding: Developing Policy and Training for General Practice, 115/0003). The views expressed in this publication are those of the author(s) and not necessarily those of the Department of Health.

2B.3 2B.4 Tangled between confidentiality and safety: Domestic violence and child safeguarding in general practice Jessica Drinkwater1 ,2, Nicky Stanley3, Eszter Szilassy2, Cath Larkins3, Marianne Hester2, Kelsey Hegarty4, Gene Feder2 1 University of Leeds, Leeds, UK, 2University of Bristol, Bristol, UK, 3 University of Central Lancashire, Preston, UK, 4University of Melbourne, Melbourne, Australia

Explaining variation in emergency admissions: a mixed methods study of emergency and urgent care systems Alicia O’Cathain1, Emma Knowles1, Jon Nicholl1, Janette Turner1, Steve Goodacre1, Enid Hirst2, Ravi Maheswaran1 1 University of Sheffield, Sheffield, UK, 2Sheffield Emergency Care Forum, Sheffield, UK

ABSTRACTS OF ORAL PRESENTATIONS The problem: Recent increases in emergency admission rates have caused concern. Some emergency admissions may be avoidable if services in the emergency and urgent care system are available and accessible. We aimed to understand variation in avoidable emergency admissions between different emergency and urgent care systems in England to identify how to reduce avoidable admissions.

The approach: A 133-item questionnaire was developed to determine the health status, use of health services and healthcare needs of a homeless population. A number of standard questionnaires were used including the AUDIT-C, PHQ-9, GAD-7, and SF-12. Prescription data was extracted from participants electronic health record and participants were recruited from within the Safetynet service.

The approach: A set of 14 conditions, likely to be rich in avoidable emergency admissions, was identified by expert consensus. The study design was a sequential mixed methods study. In Phase I we calculated an age sex adjusted avoidable admission rate. We located routine data on characteristics of emergency and urgent care systems and used linear regression to explain variation in avoidable admissions rates in 150 systems. In Phase II we undertook in depth case studies in six systems to identify further factors. A key part of these case studies was interviews with commissioners, service providers and patient representatives, totalling 82 interviews.

Findings: The majority of participants rated their physical health (70%) and mental health (75%) as good to excellent. Despite this, 97% of the population reported currently having at least one physical or mental health condition. The average number of current physical conditions reported per person was 5.3 (range 1-11, n=105). Most of the participants were experiencing multimorbidities, with 84% (88/105) reporting currently having two or more mental or physical conditions. The most common physical health conditions were skin problems (52%, 55/105); dental problems (48%; 50/105); eye problems (31%; 32/105); joint problems (23%; 24/105) and asthma (21%; 22/105). Of the blood-borne diseases, the most commonly reported disease was hepatitis C (23%; 24/104). A large proportion of participants had at some point received a formal diagnosis of a mental health condition (70%; 73/105), including depression (50%; 52/105), anxiety (36%; 38/105), schizophrenia (13%; 14/105) and bipolar disorder (6%; 6/105). More than a third of the population had been formally diagnosed with an addiction disorder (39%; 41/105), that was predominantly related to drugs (72%, 28/39). On average, participants reported that they visited the GP 6.5 times (range 1-30) in the previous six months. The majority of participants reported that they would have gone nowhere else (35%; 33/93), to another GP (26%; 24/93) or to accident and emergency (A & E) department (23%; 21/93) on their last visit, if a Safetynet GP were not available.

Findings: The 14 conditions accounted for 3,273,395 admissions in 2008-11 (22% of all emergency admissions). The mean age sex adjusted admission rate was 2258 per year per 100,000 population, with a 3.4 fold variation between systems (1268 to 4359). Characteristics of the population explained the majority of variation: deprivation explained 72% of variation, with urban/rural status explaining 3% more. Systems serving populations with high levels of deprivation and in urban areas had high rates of potentially avoidable admissions. Interviewees described the complexity of deprivation, representing high levels of morbidity, low awareness of alternative services to emergency departments and high expressed need for immediate access to urgent care. Factors related to emergency departments, hospitals, emergency ambulance services and general practice explained a further 10% of variation in avoidable admissions. Systems with high potentially avoidable admission rates had high rates of acute beds, high rates of attendance at emergency departments (which has been associated with poor perceived access to general practice), high rates of conversion from emergency department attendances to admissions, and low rates of non-transport to emergency departments by emergency ambulances. The six case studies revealed further possible explanations of variation: hospitals coded admissions differently; some systems focused proactively on admission avoidance whereas others were more interested in hospital discharge e.g. use of multidisciplinary teams based at acute trusts; there were different levels of integration; and some systems faced more challenging problems around geographical boundaries operating for different services in the system. Interviewees often described admission as the easy or safe option.

The majority of participants were receiving prescribed medications (81%; 85/105).The most commonly prescribed medication was ATC N (nervous system) medications such as paracetamol, methadone, psycholeptics and antidepressants. The second most prescribed category was systemic anti-bacterials. Consequences: Compared to the general population, the participants in this study were experiencing high levels of morbidity. Safetynet has attempted to address these needs by breaking down some of the barriers to accessing healthcare services for people who are homeless and at risk of homelessness. 2B.6

Consequences: Deprivation explains most of the variation in avoidable admission rates. Research is needed to understand the complex relationship between deprivation and avoidable admission, and to develop interventions tailored to avoid admissions from deprived communities. Standardisation of coding of admissions would reduce variation.

Myths, markets and models - career dissatisfaction amongst newly qualified GPs

2B.5

The problem: The landscape in General Practice worldwide is more challenging than ever. We face increasing demand from a growing multi-morbid population, shifting responsibilities to primary care and finite financial resources. For the UK to cope with this it is vital to recruit and retain the best GPs, and ensure they are satisfied in their work. Little research has been done into newly qualified GPsâ&#x20AC;&#x2122; views and choices regarding their GP careers, and what aspects are currently provoking dissatisfaction. By understanding their concerns and the factors that are currently leaving them frustrated, we have a far better chance of attempting to rectify matters, resulting in greater numbers of more satisfied GPs, likely to remain within and contribute to the NHS for many years to come.

Health, Medication and Use of Health Services in a Homeless Population Kirsty Oâ&#x20AC;&#x2122;Brien1, Claire Keogh1, Anthony Hoban1, Austin Oâ&#x20AC;&#x2122;Carroll2, Tom Fahey1 1 RCSI, Dublin, Ireland, 2Safetynet, Dublin, Ireland The problem: Homelessness is associated with high rates of mortality, morbidity, alcohol and drug use and other risky health behaviours, making them a particularly vulnerable group. As a group, they experience higher rates of health problems relative to the general population but often face barriers when accessing services. This study aims to characterise the health and use of health services by homeless people using a recently established primary care service for homeless people in Dublin (Safetynet).

Vanessa Nash1, Amanda Howe1 ,2 1 University of East Anglia, Norwich, UK, 2RCGP, London, UK

The approach: An online survey was distributed via the RCGP to all First5s (GPs within 5 years of qualification). 886 responses were received (18% response rate). This research deals with the qualitative open text responses. If participants expressed dissatisfaction with the posts available 41

ABSTRACTS OF ORAL PRESENTATIONS or access to additional activities, they were asked to expand on this. All participants were asked for any additional comments regarding GP Partnership. Results were coded and analysed using NVIVO. Findings: Major concerns exist regarding the models of work available, particularly lack of ‘permanent’ jobs i.e. salaried/partnership. As noted in studies previously, there was a perceived clash with personal and family life and a lack of flexibility restricting career choices. Questions were also raised about the market - with poor flow of information about job availability and little transparency regarding packages offered, alongside insufficient remuneration and variable quality of jobs. Respondents repeatedly commented that they felt stressed and overloaded. The ‘myths’ about the future of general practice, with the current political climate and uncertain impact of NHS reforms made them nervous of commitment, particularly exacerbated by negative media portrayal and excessive ‘GP Bashing’. Despite this there was great interest shown in engaging in ‘additional activities’ such as teaching and GPwSI work. The perception of partnership as a positive opportunity with responsibility and autonomy was also acknowledged. Consequences: This research raises key areas for discussion regarding engagement of newly qualified GPs. It highlights areas of dissatisfaction and potential avenues to pursue to assist practices to recruit and retain GPs. Lateral thinking may be required to incorporate more portfolio careers providing flexibility, whilst still supplying new GPs with security of employment and support to progress. Further research would be useful to investigate views in more depth and consider plausibility of recruitment and retention strategies, as well as applicability to primary care worldwide. This work has profound policy implications for workforce planning, as if not acknowledged we risk losing many GPs from the system completely. 2C.1 What works for whom? Improving patient care by implementing NICE Osteoarthritis guidelines in General Practice. Andrew Morden, Clare Jinks, Bie Nio Ong, Mark Porcheret, Krysia Dziedic Keele University, Keele, UK The problem: Osteoarthritis (OA) related joint pain is a prevalent condition which can cause disability, reduced quality of life and detrimental costs to individuals and society. OA is one of the most frequent reasons for primary care consultations, yet patients with osteoarthritis (OA) receive suboptimal treatment. A complex intervention was designed to enhance care for OA and promote the uptake of the 2008 NICE OA guidelines. The intervention was designed to facilitate patient centered holistic care and enable provision of information and comprised a GP consultation, use of an OA guidebook, and referral to a nurse for a series of consultations. The intervention has been tested in a cluster trial within which we embedded a qualitative evaluation. The approach: In-depth face to face interviews were chosen because they offered the opportunity to understand whether the advice and support offered was relevant to patients and how it informed their understanding of OA. We recruited a total of twenty-nine patients who had consulted for OA and had received the intervention (GP and nurse consultations). Thematic analysis was undertaken using principles drawn from Grounded Theory. Findings: Patients’ pre-intervention consultation experiences had invalidated the legitimacy of their complaint, left them feeling that there were no answers, and that they were not taken seriously. During the study GPs were mostly considered as gatekeepers to the nurse intervention. The OA guidebook helped all patients understand their condition better, reassured their fears, anxieties, and uncertainty (especially about the future), and provided a sense that they were not alone. Perspectives about receiving holistic care from the nurse-led clinics were split by gender. For women the time, information and attention of an expert, being taken 42

seriously, and the ability to voice personal worries and concerns were considered important. Male participants focused on obtaining technical information about OA and ultimately wanted a cure. Notably they did not discuss pastoral elements of the consultation. Seemingly the information provided in the guidebook about pain and emotions was sufficient for men. Consequences: The model underpinning the intervention can be a useful way of providing care recommended in the 2008 NICE OA guidelines. However, there may be variability regarding what elements of an intervention are effective for different patient groups (e.g gender). When designing interventions attention needs to be paid to; a) differences in how men and women experience ill health; b) how they communicate; and c) who they may wish to communicate with. 2C.2 Raising work concerns in primary care consultations for musculoskeletal conditions: a qualitative study Elizabeth L Alcock, Tom Sanders, Gwenllian Wynne-Jones, Carolyn Chew-Graham, Bie Nio Ong, Zoe Paskins Keele University, Staffordshire, UK The problem: Musculoskeletal conditions affect a person’s ability to work, resulting in reduced productivity, sickness absence and long-term work incapacity. An ageing workforce is likely to exacerbate this. Consequently, there is a need to identify optimal ways to provide support to patients with musculoskeletal problems in primary care. Previous research has highlighted inefficient support systems for patients with work-related issues and a need to explore further the barriers to successful support. Moreover, insufficient research exists exploring the individually differing and complex workrelated issues of older adults (50 +) living with musculoskeletal conditions. The study aimed to explore barriers to the optimal management of musculoskeletal conditions and work-related concerns within the primary care consultation, as well as the patient’s experiences of the clinical encounter. The approach: 100 video recorded GP consultations, in which work related difficulties were discussed, were sampled from an existing dataset and analysed utilizing constructivist grounded theory. A purposive sample of 20 patients aged over 50 years were identified from a cohort study (having indicated experiences of both musculoskeletal conditions and work related issues and given consent to further contact) and subsequently 19 of the patients were interviewed utilizing a semi-structured approach to explore patient perspectives on consulting for musculoskeletal problems. A comparative analysis of the two data sets is underway. Findings: Preliminary analysis has highlighted key themes including; 1) the normalisation or dismissal of work-related concerns as a barrier to support, 2) the patient struggle to negotiate legitimacy and deservedness as a barrier to both consulting and discussing work concerns and 3) the socio-cultural co-construction of credibility and deservedness to consult and raise work. Work-related discussions are often tokenistic and initiated by GPs and patients as a tool to provide contextual examples of disruption to everyday life, or to assign causal explanation to a condition. Responses from GPs to concerns raised by patients that were work-related and ongoing, were often normalising, re-directing patients elsewhere or dismissive. The theoretical framework of ‘Candidacy’ is being utilized to explore and expand on the difficulties faced by patients in negotiating access to the consultation and work-related discussions. Patients and GPs engage in a shared and active process of legitimating consulting, which is facilitated by the recursive nature of past consultations and experiences and the biomedical approach to the consultation.

ABSTRACTS OF ORAL PRESENTATIONS Patients struggle to raise work concerns via socio-culturally informed perceptions such as; moral worthiness to consult and stigmatised constructions of sickness certification. Consequences: Difficulties experienced by patients in raising work within the consultation can restrict patients with musculoskeletal conditions from receiving appropriate and adequate support. The findings may have implications for clinical practice and highlight a need to increase patient awareness of the support systems available to them. 2C.3

Consequences: COPERS improved psychological well-being and is likely to be cost effective by current NICE criteria but failed to influence our primary outcome of pain related disability. In the absence of more effective group self-management interventions, COPERS could be used as a substitute for less well evidenced (and more expensive) pain selfmanagement programmes. However, effective interventions to improve hard outcomes, such as disability, in chronic pain patients are still needed. 2C.4 Defining latent phenotypes of patients with hand Osteoarthritis; a fresh approach to understanding musculoskeletal conditions

Effectiveness and cost-effectiveness of a group pain self-management intervention for patients with chronic musculoskeletal pain identified in primary care: COPERS randomised controlled trial.

Dan Green, Elaine Thomas, Jo Protheroe, Danielle van der Windt Primary Care Centre, Keele University, Staffordshire, UK

Steph Taylor1, Dawn Carnes1, Kate Homer1, Brennan Kahan1, Natalia Hounsome1, Karla Diaz-Ordaz2, Sandra Eldridge1, Anne Spencer3, Tamar Pincus4, Anisur Rahman5, Stephen Bremner1, Martin Underwood6 1 Barts and The London School of Medicine and Dentistry, London, UK, 2 The London School of Hygiene and Tropical Medicine, London, UK, 3 University of Exeter, Exeter, UK, 4Royal Holloway, University of London, London, UK, 5University College London Hospitals, London, UK, 6 University of Warwick, Coventry, UK

The problem: Hand problems in older people can be due to a range of conditions, with osteoarthritis (OA) being the most frequent cause of pain and disability. In a community-based study of adults aged 50 years and over, approximately 80% of older people with hand pain attending a research clinic showed radiographic change (Kellgren & Lawrence ≥2), indicating little discrimination between individuals in such a population. A functional classification of hand pain and OA could be more meaningful in community/primary care-based studies.

The problem: Chronic musculoskeletal pain is a common problem in primary care, but is difficult to alleviate. Pharmacological approaches are often ineffective and there are increasing concerns about side effects or dependence. Non-pharmacological interventions may enable patients to manage their condition better but there is limited evidence for their effectiveness and cost effectiveness. We set out to design and evaluate a theoretically grounded self-management support intervention for chronic musculoskeletal pain that might be easily implemented in primary care.

The approach: The study population stemmed from the NorStOP (North Staffordshire Osteoarthritis Project), a large population-based prospective cohort study. Information on participants with hand pain or problems was collected using a health survey and further pain questionnaire at baseline and 3 years; all participants that responded at baseline and 3 years were included for analysis. Phenotypes of hand OA were identified using Latent Transition Analysis, a longitudinal technique that is employed to not only define sub-groups of a population based on cross-sectional data, but also to incorporate change over time. The selection of variables for the analysis was driven by previous research and advice from our Research Users Group to gain a further in-depth patient perception on the elements of hand problems. The final model was further tested to investigate if the class definitions had a similar interpretation over time.

The approach: COPERS was a novel, evidence-based, manualised, group intervention consisting of 24 components delivered over 15 hours and three days with a two week top up. The course was informed by cognitive behaviour therapy principles and aimed to increase self-efficacy to manage pain. It was delivered by two trained facilitators with experience of chronic pain - a health care professional and a lay person. We conducted a pragmatic controlled trial of COPERS in participants recruited from primary care or musculoskeletal physiotherapy services in east London and the Midlands. We used varied permuted block randomisation with an allocation ratio of 1.33:1 (intervention: control). Control participants received usual care and a relaxation DVD. We collected follow up data at 12 weeks (self-efficacy only), six and 12 months. Our primary outcome was pain related disability (Chronic Pain Grade, CPG, subscale) at 12 months. We also measured: costs, health utility (EQ-5D), anxiety, depression (Hospital Anxiety and Depression Scale, HADS), coping, pain acceptance and social integration. Findings: We recruited 703 participants, mean age 59.9 years, 81% white, 67% female, 23% in employment; 85% with pain for at least three years, nearly a quarter (23%) on strong opioids. Symptoms of depression and anxiety were common (mean HADS scores 7.4 (SD 4.1) and 9.2 (4.6), respectively), On average, intervention participants received 85% of course content. At 12 months there was no significant difference between treatment groups in CPG disability (difference -1.0, intervention vs. control, 95% CI -4.9 to 3.0, p=0.63). However self-efficacy, anxiety, depression, pain acceptance and social integration were significantly better in the intervention group at six months and these differences remained significant at 12 months for depression and social integration. COPERS has a high probability (>79%) of being cost-effective compared to usual care at a threshold of £30,000 per quality adjusted life year.

Findings: 9,705 participants responded at both baseline and 3 years (54.7% were female, 64.2 mean age (SD=9.1)), of which 5,675 reported hand problems at either or both points. 15 variables were considered for analysis. Removing variables that did not improve the classification of the model or distinction between sub-groups, led to a model based on 9 variables including 3 pain, 5 function and 1 stiffness measure. The optimal model separated the population into 5 potential phenotypes at baseline: ‘least affected’, ‘poor gross function’, ‘high pain during activities’, ‘high pain and poor gross function’ and ‘severely affected’. Phenotype definitions showed only slight variations over the 3 year period, indicating progression in severity of problems. The greatest movement was seen for those with high pain on activity transitioning into the least affected group (43%), and those that developed a substantial increase in pain levels whilst maintaining their poor gross functional impairments (36%). Consequences: This works has defined phenotypes of patients with hand OA, based on self-report answers to brief pain, function and stiffness items. In addition, it provides evidence that there is movement amongst some classes, but little movement into less severe phenotypes once a member of the highly affected group. This will offer potential benefits to General Practitioners in their knowledge of the sub-groups of hand OA patients, and further understanding about the course of the condition.

ABSTRACTS OF ORAL PRESENTATIONS 2C.5 Disconnects between clinical guidelines and real world practices: a qualitative exploration of the role of NICE guidelines in primary care management of low back pain Felicity Bishop1, Alex Dima1, Jason Ngui1, Paul Little1, Rona Moss-Morris2, Nadine Foster3, George Lewith1 1 University of Southampton, Southampton, Hampshire, UK, 2Institute of Psychiatry, Kings College London, London, UK, 3Arthritis Research UK Primary Care Centre, Keele University, Keele, UK The problem: The 2009 NICE guidelines for persistent non-specific LBP (CG88) sparked considerable debate in the academic, medical and lay press when published. They remain of particular interest as they recommend treatments that have previously been widely contested (e.g. acupuncture, spinal manipulation). As has been the case for other clinical guidelines, we anticipated some difficulties around the implementation of these guidelines and wanted to understand their role in the primary care management of LBP in England. We also aimed to identify the decisionmaking processes that result in the delivery of specific interventions to patients with low back pain (LBP) in primary care. The approach: We conducted a qualitative study using semi-structured interviews and inductive thematic analysis. We interviewed a purposive sample of 53 primary care clinicians who were working either within the NHS (20), the private sector (21), or both (12) in south-west England. All major professions that provide treatments recommended by NICE for primary care management of LBP participated: 16 GPs, 10 chiropractors, 8 physiotherapists, 8 acupuncturists, 7 osteopaths, and 4 nurses. Findings: Clinicians discussed a complex process of matching individual patients (perceived in terms of their medical history, clinical presentation, personal preferences, concerns, and psychosocial context) to familiar treatments (viewed primarily in terms of likely effectiveness). National clinical guidelines comprised one of many possible inputs to this process. In particular, many clinicians found the label “persistent non-specific LBP” to have little relevance to their clinical practice. A fundamental disconnect thus exists between the label that appears on the front cover of the NICE guidelines and the (admittedly quite varied) terminology that different clinicians are more familiar with and likely to use in practice. Decisions about which treatment to prescribe, refer, or deliver to an individual patient were made while clinicians were working within a broader context informed by personal experience and inter-professional networks and constrained by structural organisational factors. Clinicians were frustrated by disparities between the idealised care pathway described in the NICE guideline and the real-world situation of short consultation times, difficult-to-access specialist services and failure to commission guideline-recommended treatments. Consequences: Our findings suggest specific actions that address each key element of clinical decision-making and could thus increase the relevance of the LBP guidelines for clinical practice. These actions involve: ensuring the terminology used in guidelines reflects that used in clinical practice; dispelling the image of guidelines as rigid and prohibiting patient-centred care; providing opportunities for GPs and nurses in particular to engage in experiential learning about guideline-recommended CAM therapies; and commissioning guideline-recommended treatments for NHS patients.

The problem: The management of long term conditions is based on guidelines for each specific disease. However, there is an increasing prevalence of patients with multimorbidity. For these patients, the single condition model for treatment may be burdensome. They attend multiple clinics which can be inconvenient and inefficient. They frequently get depressed and sometimes complain that no-one treats them as a ‘whole person’ or takes their views into account. Complex polypharmacy and treatment adherence can also be issues. An increasing body of literature describes the scale of this problem, with qualitative research disclosing the views of patients and health professionals, editorials expounding what should be done, but few large scale randomised controlled trials examining what is clinically and cost effective. This study aims to develop, optimise and evaluate a new approach to improve the management of multimorbidity in general practice. Instead of focussing on each disease in isolation, the aim is to treat the whole patient in a joined-up manner in order to improve their overall quality of life. The approach: The study is a pragmatic cluster randomised controlled trial with nested process evaluation and economic analysis of cost effectiveness. A pilot phase of 4 practices (2 in Bristol and 2 in Glasgow) will develop the intervention and study processes. The main trial will recruit 32 practices from Bristol, Glasgow and Manchester, cluster randomised to receive the intervention or usual care. A target of 1383 participants will be recruited. Eligible participants will be 18 or over with multimorbidity, defined as having 3 or more long term conditions from a predefined QOF list. The intervention is based on the Patient Centred Care model. Practices will implement strategies to maximise continuity of care and offer longer appointments for patients with multimorbidity. Instead of unco-ordinated reviews of each of their conditions, patients will have ‘3D’ reviews focussing on the ‘Dimensions of Health’ (patients’ priorities, quality of life, and disease control), ‘Depression’ (assessment and treatment) and ‘Drugs’ (pharmacist recommendations, strategies to simplify drug regimes and improve adherence). Each practice will have a linked general physician in hospital for telephone advice. Participants will be followed up for 12 months with the primary outcome being health related quality of life (EQ-5D) at 12 months. Findings: Funding from the NIHR Health Services and Delivery Research Programme (HS&DR 12/130/15) has been awarded. The study is currently in set-up phase and about to recruit pilot practices. Consequences: There is widespread interest in how to improve care for patients with multimorbidity. We hope that this study will develop and test new approaches to GP management of multimorbidity patients. If successful, this intervention could improve the quality of life of patients, their experience of care and reduce NHS and patient costs. 2D.2 Resolving conflicts in the multimorbid consultation: How do GPs balance diseases, drugs and the views of other doctors?

The 3D Study: Improving the management of patients with multimorbidity in general practice

Carol Sinnott1, Sheena McHugh2, Maria Boyce2, Colin Bradley1 1 Department of General Practice, University College Cork, Cork, Ireland, 2 Department of Epidemiology and Public Health, University College Cork, Cork, Ireland

Chris Salisbury1, Mei-See Man1, Pete Bower2, Sara Brookes1, Ali Heawood1, Sandra Hollinghurst1, Bruce Guthrie3, Cindy Mann1, Imran Rafi5, Stewart Mercer4

The problem: Multimorbidity, a global challenge in the provision of healthcare, requires innovation in primary care research and intervention design. To date, most research in multimorbidity has focused on

2D.1

1 University of Bristol, Bristol, UK, 2University of Manchester, Manchester, UK, 3University of Dundee, Dundee, UK, 4University of Glasgow, Glasgow, UK, 5Royal College of General Practitioners, London, UK

ABSTRACTS OF ORAL PRESENTATIONS organisational and patient-orientated interventions. Few studies have examined the role of professional-orientated interventions. The aim of this study was to explore how General Practitioners (GPs) currently respond to difficulties in managing multimorbidity, with a view to identifying targets suitable for professional-orientated interventions. The approach: Design: In-depth qualitative interviews incorporating chart stimulated recall, a clinical assessment tool that uses a medical chart to stimulate a physician’s recall of a case and its management. Setting: Primary care in the Republic of Ireland. Participants: GPs purposively sampled from continuing professional development groups, using sampling criteria of: length of time qualified; location (rural/urban); and practice size (single/group). Analysis: Interviews were coded using the grounded theory method of constant comparison and theory was developed iteratively. Findings: Twenty GPs were interviewed on a total of 53 cases of multimorbidity. The data demonstrate how decision making in multimorbidity requires integration of information from multiple sources, such as the patient, specialists, evidence based medicine, and the expertise of the GP themselves. The relative importance of these factors varies by case, and their inputs are ‘balanced’ by the GP to make the most appropriate decision for that patient. Difficulties arise when the factors conflict, thus unbalancing the decision making process. GPs respond to this scenario using a range of techniques. Conflicts arising from guidelines are responded to by ‘satisficing’ with respect to chronic disease targets; GP respond to their own uncertainty by ‘broadening the loop’ to increase the relative input of specialists or the opinions of the patient or their next of kin; conflicts arising from specialist input leads to ‘negotiation’ with specialist or patient, or the GP acting as ‘final arbitrator’. Conflicts most commonly arose when GPs were isolated from the support of other generalists (GPs/ specialists), or had difficulties in the doctor-patient relationship. In multimorbid patients who were ‘stable’, even those with high levels of polypharmacy, the default position for GPs was to ‘maintain the status quo’. This approach of not changing medications, unless there was clear evidence of harm or adverse drug events, arose from GP fear that future events would be linked to the act of removing medications or the act would lead to patient anxiety and the perception of withdrawal of care. Consequences: This study identified potential weaknesses in the decision making process in multimorbidity, which related to lack of GP empowerment, access to professional support, and communication with patients. These novel findings will inform the development of a professional-orientated intervention, to assist GPs in the provision of multimorbidity care.

The approach: Participants were recruited from primary care services. Fifty participants were randomised to the intervention group and received the OPTIMAL programme (n=26) or the waiting-list control group (n=24). OPTIMAL aimed to increase occupational participation and covered a range of topics that included fatigue, physical activity levels, stress/anxiety management, medication management and communication skills. The primary outcome was frequency of occupational participation. Secondary outcomes included self-perception of occupational performance, independence in activities of daily living, anxiety and depression, self-efficacy, health-related quality of life, the impact of health education, health care utilisation and individualised goal attainment. Findings: Patients in the intervention group had significant improvement across a range of occupational performance and psychosocial measures. A multiple linear regression model showed a significant improvement in frequency levels of occupational participation for the intervention group, compared to the control group (p=0.003). A significant difference was shown between the two groups for perceptions of occupational performance (p<0.001) and satisfaction with same (p<0.001), self-efficacy levels (p<0.001), independence in daily activities (p<0.001), health-related quality of life (p=0.001) and positive and active engagement in life (p=0.002). The intervention group demonstrated high levels of goal achievement, between baseline and follow-up (p<0.001). No significant differences were found in anxiety, depression or health care utilisation. Consequences: The findings from this RCT provide evidence that an occupation-based self-management programme can significantly improve occupational outcomes, self-efficacy and quality of life. GPs and practice nurses play an important role in the future implementation of the OPTIMAL programme, in raising awareness and recruiting suitable patients for the programme. The trial will contribute towards much needed evidence on the effectiveness of interventions for patients with multimorbidity and has implications for occupational therapy practice and policy in primary care. The results have supported the application for a definitive intervention trial that will test the cost effectiveness and sustainability of the OPTIMAL intervention over a wider range of practices in Irish primary care. 2D.4 Common patterns of morbidity and multi-morbidity and their impact on health related quality of life: evidence from a national survey Ruben Mujica Mota1, Martin Roberts2, Gary Abel3, Marc Elliott4, Georgios Lyratzopoulos3, Martin Roland3, John Campbell1 1 University of Exeter Medical School, Exeter, UK, 2University of Plymouth Peninsula Schools of Medicine and Dentistry, Plymouth, UK, 3Cambridge Centre for Health Services Research, Cambridge, UK, 4RAND, Santa Monica, USA

2D.3 The Impact of OPTIMAL, an Occupation-based Self-management Programme, on Occupational Participation for Individuals with Multimorbidity: a Randomised Controlled Trial Jess Garvey1, Deidre Connolly1, Fiona Boland2, Susan Smith2 1 Trinity College Dublin, Dublin, Ireland, 2Royal College of Surgeons in Ireland, Dublin, Iran The problem: Multimorbidity, defined as the co-existence of two or more long-term conditions, presents a multitude of challenges to both the individual and healthcare services. There is an identified need to develop interventions to improve outcomes for individuals with multimorbidity and exploratory work indicated that an occupational therapy intervention may be effective. This randomised controlled trial aimed to investigate the effectiveness of OPTIMAL, an occupation-based self-management programme, for multimorbidity in a primary care setting.

The problem: There is limited evidence about the impact of specific patterns of multi-morbidity on health-related quality of life from large samples of data obtained from adults. The approach: We used data from the English General Practice Patient Survey 2011-2012. We defined multi-morbidity as the presence of two or more of 12 self-reported conditions or another (unspecified) long term health problem. We investigated differences in quality of life (EQ-5D scores) associated with combinations of these conditions after adjusting for age, gender, ethnicity, socio-economic deprivation, and the presence of a recent illness or injury. Analyses were based on 831,537 responses from patients aged 18 years or older in 8,254 primary care practices in England. Findings: 23% of respondents reported two or more chronic conditions (ranging from 7.5% of those under 45 years of age to 51.1% of those 65 years or older). Multi-morbidity was more common among women, White individuals, and in respondents from socio-economically deprived areas. 45

ABSTRACTS OF ORAL PRESENTATIONS Neurological problems, mental health problems, arthritis and long-term back problem were associated with the lowest health related quality of life. The presence of three or more conditions was commonly associated with an impact on quality of life over and above the sum of the impact of the individual conditions on their own. Mental health problems had an adverse effect on quality of life, especially when occurring in combination with other physical conditions. Decrements in EQ5D scores associated with an additional condition in people with more than three physical conditions were greater for younger people than for older people. Multi-morbidity was associated with a substantially worse quality of life in diabetes than in other long-term conditions. Consequences: Multi-morbidity is associated with a larger negative difference in quality of life in diabetes than other long-term conditions. Mental health problems have strong associations with quality of life both in their own right, and also as comorbidity. Integrating mental and physical health approaches to care should be seen an essential component in the delivery of high quality care for people with long-term health conditions. 2D.5 Exploring the clustering of chronic pain, depression and heart disease in a general population-based cohort Oliver van Hecke1, Nicola Torrance1, Lynne Hocking2, Sandosh Padmanabhan3, David J Porteous4, Blair H Smith1 1 University of Dundee, Dundee, UK, 2University of Aberdeen, Aberdeen, UK, 3University of Glasgow, Glasgow, UK, 4University of Edinburgh, Edinburgh, UK The problem: For many people chronic pain can be debilitating and can arise after an operation, injury, or onset of disease; it remains one of the most common reasons why people seek medical care. Current evidence points towards a common link between chronic pain, depression and/or heart disease and subsequent poor health and premature death. Because of the complexity of the pain experience and its multi-factorial effect on health-related quality of life, it remains unknown the extent to which this co-occurrence of chronic pain, depression and heart disease is the result of a shared-aetiology, genetic effect and/or because of certain confounding factors such as health inequality. The approach: Data from Generation Scotland: the Scottish Family Health Study (chronic pain, cardiovascular and mental health parameters) were analysed using logistic regression modelling to calculate the likelihood of one, two or all three condition(s) adjusting for various confounders including relative deprivation. Findings: 24,042 participants were included. 7,162 (36%) reported any chronic pain; 3,664 (22%) had severe chronic pain; 2,009 (10%) had significant exertional chest pain (Rose Angina); 2, 771 (12%) had a history of major depressive disorder. Individuals with any chronic pain were significantly more likely to have depression and angina than those without chronic pain [adjusted OR 6.59, 95% CI 4.89-8.87]. This likelihood increased further in those with severe chronic pain [adjusted OR 9.35, 95% CI 6.82-12.82]. Consequences: Our findings reiterate previous epidemiological work that chronic pain is common within the community. Specifically this study demonstrates the close relationship between those that self-report chronic pain, depression and angina (as a risk marker for heart disease) after adjusting for confounders: having one condition significantly raises the odds of having the other condition(s) compared to those with neither. Our epidemiological strategy mapping the co-occurrence of these three conditions supports current theories of an underlying shared-aetiology mechanism and/or genetic aetiology between these conditions. The next step will be to explore the co-heritability of these conditions in family 46

members within Generation Scotland with the potential for genetic analysis. 2D.6 Multimorbidity and polypharmacy in stroke. Katie Gallacher1, David Batty2 ,5, Gary McLean1, Stewart Mercer1, Bruce Guthrie3, Peter Langhorne1, Carl May4, Frances Mair1 1 University of Glasgow, Glasgow, UK, 2University College London, London, UK, 3University of Dundee, Dundee, UK, 4University of Southampton, Southampton, UK, 5University of Edinburgh, Edinburgh, UK The problem: Owing to improvements in the management of stroke, more people worldwide are surviving and following lifelong management regimes. Many of these people will face the challenge of managing multiple conditions simultaneously, a global issue recognised by the World Health Organization. Treatment burden is a new concept that denotes the workload of healthcare for patients and the effect of this on health and wellbeing, a particular issue for those with multimorbidity. The aim of this study is to examine patterns of multimorbidity and polypharmacy in stroke, to identify if this is a group at risk from high treatment burden. The approach: A cross-sectional study carried out on a national dataset managed by the Primary Care Clinical Informatics Unit (PCCIU) at the University of Aberdeen. Based on 314 primary care practices in Scotland, the database comprises 1,424,378 individuals aged 18 years and over. Data included information on the presence of stroke and another 39 long term conditions plus prescriptions for regular medications. Differences between those with stroke and all other individuals were calculated using t-tests and logisitic regression, adjusting for gender, age and deprivation. The association between number of morbidities and regular medications was calculated using Spearmanâ&#x20AC;&#x2122;s correlation. Findings: Demographic characteristics of the sample were representative of the Scottish adult population (gender, age, deprivation). The prevalence of stroke was 2.5%. The five most frequent co-morbidities present in Stroke patients were: Hypertension (60.9%), Coronary Heart Disease (CHD)( 29.5%), Painful Condition (21.9%), Depression (20.7%), Diabetes (18.8%). Of the 39 conditions examined, 33 were more common in the stroke group. Of the stroke patients, 94.2% had one or more morbidity in addition to their stroke, whereas only 48% of the stroke-free control group had any morbidity present. A strong association was demonstrated between stroke and the presence of any morbidity (odds ratio; 95% confidence interval: 5.18; 4.95 to 5.43). Number of regular medications was strongly correlated with number of morbidities (Ă&#x; = 0.757, p<0.001). 17.8% of the stroke group had 10 or more repeat prescriptions compared to 2.1% of the control group. The prescription of any regular medication was strongly associated with stroke presence (4.53; 4.33 to 4.74). Limitations include the use of data collected for clinical rather than research purposes, difficulties in defining the presence of certain long term conditions, and a lack of weighting in the measurement of multimorbidity, although the latter was deemed suitable for a descriptive study. Consequences: Multimorbidity and polypharmacy are more common in stroke patients than in the general population, therefore people who have had a stroke are at higher risk of being over-burdened from the demands of managing their health. This is the first study of such scale that examines multimorbidity and polypharmacy in stroke. 2E.1 Antidepressant use and the risk of cardiovascular outcomes: a cohort study using a primary care database Carol Coupland1, Trevor Hill1, Richard Morriss1, Antony Arthur3, Michael

ABSTRACTS OF ORAL PRESENTATIONS Moore2, Julia Hippisley-Cox1 1 University of Nottingham, Nottingham, UK, 2University of Southampton, Southampton, UK, 3University of East Anglia, Norwich, UK

anomalies, have been associated with antidepressant use in pregnancy. However, most research has failed to adequately account for important characteristics of the women prescribed antidepressants.

The problem: Antidepressants are one of the most commonly prescribed medications in primary care, and their use is increasing. More than 50 million antidepressant prescriptions were issued in England in 2012. There is concern that selective serotonin reuptake inhibitor (SSRI) antidepressants may increase the risk of cardiovascular outcomes including stroke and arrhythmia but evidence is inconsistent and inconclusive.

Findings: During follow-up 87.7% (n=209,476) of the 238,963 patients in the cohort received one or more prescriptions for antidepressants. The majority of prescriptions were for SSRIs (71%). During the first five years of follow-up 772 patients had a myocardial infarction, 1106 had a stroke/ TIA and 1452 patients were diagnosed with arrhythmia.

The approach: We used a comparative cohort design including four cohorts of women with different antidepressant exposures using data from The Health Improvement Network (THIN) primary care database. Thus, we identified a sample of 209,135 pairs of women and their children including 5,154 women receiving SSRIs before pregnancy, 2,776 receiving SSRI in pregnancy, 992 receiving other antidepressants and 200,213 not receiving antidepressants before or during pregnancy. We also identified characteristics potentially affecting the association between use of antidepressants in pregnancy and congenital heart anomalies. These included: maternal age, obesity (BMI above 30), smoking, social deprivation, other psychotropic medications (antipsychotics, antiepileptic drugs (AED), hypnotics and anxiolytics), history of alcohol problems, history of illicit drug use and diabetes. The risk of congenital heart anomalies were compared for women who received SSRIs in the first trimester against the three other cohorts using multivariable logistic regression. We also examined the associations of congenital heart anomalies and the characteristics of the women listed above. Findings: Less than 1% of children had a record of congenital heart anomalies within five years of birth and there were no significant differences related to antidepressant exposure in pregnancy. However, independent of antidepressant prescribing, diabetes (Odds Ratio (95% Confidence Interval): 2.23 (1.79 - 2.77)) increasing age (OR (95% CI) per year: 1.01 (1.00 - 1.02)), alcohol problems (OR (95% CI): 2.58 (1.55 4.29)), illicit drug use (OR (95% CI): 1.89 (1.09 - 3.25)) and obesity (OR (95% CI): 1.38 (1.13 - 1.69)) were associated with an increased risk of having a child with congenital heart anomalies.

There were no significant associations between class of antidepressant and myocardial infarction over five years follow-up, although in an analysis restricted to the first year of follow-up patients prescribed SSRIs had a significantly reduced risk of myocardial infarction compared with no use of antidepressants (adjusted hazard ratio 0.58, 95% CI 0.42 to 0.79). Among the most commonly prescribed antidepressants the SSRI fluoxetine was associated with a significantly reduced risk of myocardial infarction over 1 year (adjusted hazard ratio 0.44, 95% CI 0.27 to 0.72).

Consequences: Our study did not support a general association between SSRIs and congenital heart anomalies. However, our findings confirm an increased risk of congenital heart anomalies associated with diabetes, obesity and history of alcohol and illicit drug use independent of the prescription of antidepressants may suggest that advising women to stop antidepressant treatment in pregnancy may be counterproductive. Future research on the associations between SSRIs and congenital heart anomalies must account for these characteristics.

There were no significant associations between antidepressant class and stroke/TIA. There was some evidence of a reduced risk of arrhythmia when SSRIs were used (adjusted hazard ratio 0.84, 95% CI 0.73 to 0.97) for the analysis if five years follow-up. Among the SSRIs only fluoxetine was associated with a significantly reduced risk of arrhythmia (adjusted hazard ratio 0.74, 95% CI 0.59 to 0.92), but citalopram was not associated with a significantly increased risk even at high doses (adjusted hazard ratio 1.08, 95% CI 0.74 to 1.57 for doses over 20mg/day).

2E.3

Consequences: This study has found no evidence that SSRIs are associated with an increased risk of myocardial infarction, stroke/TIA or arrhythmia, but some evidence that they are associated with a reduced risk of myocardial infarction and arrhythmia particularly for fluoxetine. This is reassuring in light of recent safety concerns.

The problem: Globally, mental health problems contribute massively to the burden of ill-health. Patients face long waiting lists and difficulty in accessing appropriate services. Some patients prefer to avoid mainstream mental health services. This trial tested the effectiveness and costeffectiveness of low-intensity Cognitive Behavioural Therapy (CBT), delivered as self-help group classes in community settings, for people with symptoms of anxiety and depression, to investigate an alternative to usual services.

We carried out a study to assess the associations between antidepressant treatment and the risk of three cardiovascular outcomes (myocardial infarction, stroke/transient ischaemic attack (TIA) and arrhythmia). The approach: A cohort of patients aged 20 to 64 with a first recorded diagnosis of depression between 1/1/2000 and 31/07/2011 was identified using the QResearch primary care database and followed up to 31/07/2012. Information was extracted on prescriptions for antidepressants during follow-up, and on potential confounding variables. Cox proportional hazards models were used to estimate adjusted hazard ratios for the associations between myocardial infarction, stroke/TIA or arrhythmia and antidepressant treatment over five years follow-up, including analyses of antidepressant class, dose, individual drug type and duration of use.

2E.2 Selective Serotonin Reuptake Inhibitors and congenital heart anomalies: Comparative cohort studies of women treated before and during pregnancy and their children Irene Petersen1, Stephen Evans2, Louise Marston1, Ruth Gilbert1, Irwin Nazareth1 UCL, London, UK, 2London School of Hygiene and Tropical Medicine, London, UK 1

The problem: Adverse birth outcomes, in particular congenital heart

A randomised controlled trial of a community based, group guided, self-help intervention for anxiety and depression. Jill Morrison, Chris Williams, Alex MacConnachie, Carrie-Anne McClay, Lynsey Matthews, Caroline Haig University of Glasgow, Glasgow, UK

The approach: Individuals with symptoms of depression self-referred via community adverts (including free newspapers) to Living Life ToThe Full (LLTTF) classes. The 8-week course involved weekly 90 minute classes. Class leaders (provided by the charity Action on Depression) guided participants through written self-help booklets aimed at teaching key life-skills. Participants were randomly allocated to an Immediate Access group or a Delayed Access Control group who received the classes after 6-months. Measures of depression (PHQ-9), anxiety (GAD-7) and social 47

ABSTRACTS OF ORAL PRESENTATIONS function (WSAS) were collected at baseline and 6-months. Individuals aged 16 and over with at least mild depression were recruited. Findings: 142 patients were recruited (16 above target). 68.1% had experienced depression for over five years. 49.3% were taking an antidepressant at baseline with no difference in antidepressant usage between the groups. 71.8% (102/142) provided data at six months. Significant reductions in depression (mean 3.6 points improvement on PHQ-9), anxiety (mean 2.8 points improvement on the GAD-7) and social function (mean 5.3 points improvement on the WSAS) were observed in the Immediate Access arm compared with the Delayed Access Group. Mean participant satisfaction with the LLTTF classes (IA group) was 24.3±5.1 (n=47) measured by the CSQ-8. Classes were rated as useful, quite useful or extremely useful by 81.8% (n=36 of 44) of participants who returned class feedback forms. The classes were cost-effective with a reduction in the LLTTF class arm costs from £907 to £780 (£-127), and from £802 to £740 (£-62) in the control group, in the 6 months before and after joining the study. Overall the delivery of the classes (rooms, staff, resources) was cost-neutral as delivery costs were fully offset by the savings in health care usage in the intervention arm - with significant savings in terms of admission to hospital. Consequences:This study is the first to evaluate CBT self-help resources delivered with low-intensity support via short, weekly, small group classes delivered in partnership with the voluntary sector in a community setting. The classes were effective, cost effective and acceptable in the management of depression, anxiety and impaired social function. Community-based recruitment can successfully reach individuals in need of support including those not currently receiving GP support. The LLTTF classes provide an alternative treatment option for use in primary care and community settings. 2E.4 Minimum clinically important differences on the Beck Depression Inventory (II): Implications for outcome assessment in depression trials Katherine Button1, Daphne Kounali1, Laura Thomas1, Nicola Wiles1, Tim Peters1, Nicky Welton1, A.E. Ades1, Glyn Lewis2 1 University of Bristol, Bristol, UK, 2University College London, London, UK The problem: The Beck Depression Inventory (BDI) self-administered questionnaire is widely used in clinical research on depression. However, the minimum clinically important difference (MCID) is unknown. It is also unknown whether the MCID differs according to patient characteristics such as initial depression severity or previous non-response to antidepressant medication. The aim of this research was to estimate the MCID on the BDI according to the patient’s report of feeling better, and to examine whether this estimate varied according to initial severity or previous treatment-resistance. The approach: Improvement on “global rating of change” scales were compared with changes in BDI scores using Receiver Operator Characteristics, logistic regression, and Generalised linear models (GLM), in previously collected data (n = 1039) from three large RCTs for the management of depression (GenPod and TREAD) or treatment-resistant depression (CoBalT) in primary care. The cut-point from the ROC analyses provides the optimal threshold (in terms of maximising the sum of sensitivity and specificity) above which individuals are classified as “better” and below which they are classified as “not better”. The ROC analyses therefore provide the best estimates of MCID. GLM analyses were used to explore baseline dependency and to assess whether MCID is best measured in absolute terms (i.e., difference scale) or as percent reduction in scores from baseline (i.e., ratio scale).

Findings: Improvement in BDI scores associated with reporting feeling “better” depended on initial depression severity with more severe patients requiring a bigger reduction in BDI score in absolute terms to report feeling better. GLM indicated that, given this baseline dependency, the MCID is best measured on a ratio scale. We estimated a MCID of a 17.5% reduction in scores from baseline from ROC analyses. The corresponding estimate for individuals with longer duration depression who had not responded to antidepressants was higher at 32%. The average improvement from GLM models for those with average baseline to report feeling better was 36%. The corresponding estimate for individuals with longer duration depression who had not responded to antidepressants was higher 45%. Consequences: MCID on the BDI is dependent on baseline severity and the MCID for longer duration depression that has not responded to antidepressants is larger than that for more typical depression. This has important implications for the design and interpretation of clinical trials and clinical practice. Outcome may need to be assessed in terms of percent reduction in scores from baseline, and greater improvements on the BDI are required to indicate clinically meaningful change in depression which has not responded to antidepressant medication. 2E.5 A Systematic Review of Population-Based Risk Prediction Modelling for Dementia Eugene Tang, Stephanie Harrison, Linda Errington, Louise Robinson, Blossom Stephan Institute of Health and Society, Newcastle University, Newcastle, UK The problem: Dementia is a global public health concern. The lack of effective treatments has meant that earlier stages of disease have become a research focus and target for intervention trials. The UK National Screening Committee has concluded that there is no evidence base for routine screening for dementia or even at-risk cases despite calls in the UK and Internationally to do so. A number of risk factors have been identified, including cardio-metabolic health status which could be targeted by intervention strategies. These risk factors have been incorporated into numerous risk prediction models to identify those at high risk of developing dementia. A previous systematic review in 2009, identified 25 articles relating to dementia risk prediction models. The review highlighted that unlike other disease areas such as cardiovascular health, where there are risk prediction models used in clinical practice (e.g., Framingham Risk Scores) that have had significant impact for informing treatment decisions and clinical research, no such tool exists for dementia. Since 2009, with government interest and funding calls focused on stratified medicine in the field of dementia, numerous new dementia risk prediction models have been developed. Given this, the aim of this project was to update the original review. The approach: A systematic literature search was conducted in MEDLINE, Embase, Scopus and ISI Web of Science. All articles published from 1st January 2009 to 9th September 2013 were included. Longitudinal, population-based studies examining dementia risk prediction models, including models developed in individuals with Mild Cognitive Impairment (MCI), which included measures of sensitivity/specificity and/or area under the receiver operating characteristic curve, were included. AUC values of 0.9-1, 0.7-0.89 and ≤0.7 are used to indicate excellent, good and models of questionable utility, respectively. Findings: Out of the 1093 articles screened 31 met the eligibility criteria. 11 were developed within the context of MCI and 20 were developed within the context of the whole population (e.g., non-MCI based). For non-MCI models the AUC ranged from 0.50 - 0.89 and 0.48 - 0.90 for MCI models. Only 1 new model was externally validated. The models could be broadly divided into three categories: (1) neuropsychological based models; (2) health based models and, (3) multifactorial models.

ABSTRACTS OF ORAL PRESENTATIONS Consequences: At present no dementia risk prediction model can be currently recommended for use within a screening program. The best models were ones that incorporated either cognitive factors alone or incorporated these scores with health and demographic factors. A cost-effective consensus model based on the best combination of risk factors needs to be developed to accurately identify those at high risk of future dementia. Further research is then required to evaluate whether this would be used particularly in primary care where similar tools for other conditions are already in use.

The problem: Survival rates following a diagnosis of cancer vary between countries: a difference attributed to diagnostic delay resulting in more advanced disease at diagnosis. The International Cancer Benchmarking Partnership (ICBP), a collaboration between 11 jurisdictions with primary care led health services, was set up in 2009 by Cancer Research UK. Module three investigated the primary care interval, defined as the period from when the patient first presents to primary care with a symptom of possible oncological significance until the time that a referral is made to secondary care (or a diagnosis of cancer is made in primary care).

2E.6

The approach: An online cross-sectional survey was used to identify differences between jurisdictions. Hypotheses were developed based on both the clinical experience of the academic steering group and knowledge of factors that could account for different cancer outcomes from the literature. The survey comprised a set of direct questions on primary care infrastructure and five clinical scenarios to assess clinician behaviour, and was discussed and refined concurrently and repeatedly with international partners. The survey was validated using an iterative process in England, and was then translated and amended to be relevant to the medical systems operating in other jurisdictions.

What self-harm means in general practice and how it relates to suicide prevention: a qualitative study Amy Chandler1, Chris Burton2, Steve Platt 1 University of Edinburgh, Edinburgh, UK, 2University of Aberdeen, Aberdeen, UK The problem: Self-harm is a contested term, incorporating a range of behaviours including self-injury (notably self-cutting) and self-poisoning. Research on medical care of patients who have self-harmed has focused on Accident and Emergency settings, or management of suicidal self-harm in primary care. While self-harm is understood to increase risk of completing suicide, the majority of self-harm is not thought to be suicidal. Little is known about how General Practitioners respond to self-harm in primary care, nor how they manage the challenges of assessing suicide risk in patients who have self-harmed. The approach: We conducted qualitative, semi-structured interviews with 30 GPs from two health boards in Scotland. GPs were diverse in terms of: age, gender, location of practice (rural/urban; deprived/affluent) and experience (number of years practicing). Interviews explored GP’s accounts of treating patients who had self-harmed; views on the meaning of self-harm; accounts of assessing and responding to suicidality in self-harming patients. Interviews were transcribed and analysed using an inductive, narrative informed, thematic approach. Findings: Significant differences emerged in the ways that GPs working in different areas accounted for the meaning of self-harm, and how self-harm was related to suicide risk. Those working in areas of deprivation or with marginalised groups (especially drug-using populations) described self-harm and suicide as closely related in complex ways. In contrast, those working in more affluent communities provided accounts which indicated a clearer distinction between self-harm and suicide; in some cases suggesting patients presenting with self-harm would be classed as at low risk of suicide. This reflected an understanding that GPs across the sample alluded to: that those at greatest risk of suicide were least likely to seek help for either self-harm or suicidality. Consequences: Differences in GPs accounts of how self-harm related to suicide have clear implications for suicide prevention. GPs’ suggestions of a clear split between self-harm and suicide run counter to suicide prevention messages which highlight self-harm as a key risk factor for suicide. However, such accounts also reflect a parallel ‘truth’ that most self-harm is not suicidal in nature. The challenges reported by GPs in responding to self-harming patients highlight the complex and uncertain nature of self-harm and suicidality. Suicide prevention messages need to recognise the diversity of experiences and understandings of these issues. 2F.1 Are cancer survival differences embedded in primary care? Brian Nicholson, David Mant, Peter Rose Oxford University, Oxford, UK

Findings: We will present the overall findings from module three in relation to international differences in cancer survival, including variations in: clinician views on cancer diagnosis; clinical reasoning in response to the clinical scenarios; practice administration and infrastructure; primary care access to investigations; and time taken to access specialist care. Consequences: The Module 3 findings will directly inform international efforts to reduce inequalities in cancer outcomes by providing insight into which high-yield policy changes could be made in primary care to reduce diagnostic delay. 2F.2 The risk of cancer in primary care patients with hypercalcaemia: a retrospective cohort study using electronic records. Fergus Hamilton1, Robert Carroll1, William Hamilton2, Chris Salisbury1 University of Bristol, Bristol, UK, 2University of Exeter, Exeter, UK

The problem: Hypercalcaemia is an uncommon but important problem in primary care. In secondary care, up to 60% of hypercalcaemia is found to have a malignant cause, but these figures are likely to be vastly different in primary care. There have been no large studies looking at the risk of cancer in primary care after hypercalcaemia. The approach: We undertook a retrospective cohort study in the Clinical Practice Research Datalink (CPRD). A dataset from the DISCOVERY programme was used. All calcium blood tests were extracted for the period 2000-2009, and stratified by value. For the year after the blood test, all diagnoses of cancer were extracted. A multivariable logistic regression was performed for age, gender and calcium as a continuous variable. This then repeated for subcategories of calcium elevation. Positive predictive values (PPV’s) were calculated. Findings: 54,267 participants had a calcium level during the study period. In 3% (1,674 participants), this level was high (=>2.6mmol/l). 72.5% (1214) of these participants were female. Hypercalcaemia was strongly associated with cancer, especially in men (OR 2.92 in men, 1.86 in women, p<0.001) Hypercalcaemia had a PPV of 11.5% for cancer in men, but only 4.12% in women. This disparity increased with higher calcium levels - at 2.8mmol/l the PPV was 27.9% for men and 8.7% in women. At a calcium of 3.0mmol/l, the PPV was 50% in men and 16.7% in women. Lung, myeloma, prostate, colorectal and haematological cancers were the main 49

ABSTRACTS OF ORAL PRESENTATIONS cause in men, with a similar picture in women, with breast replacing prostate. Consequences: Hypercalcaemia is strongly associated with cancer in primary care, with men being at most risk, despite hypercalcaemia being more common in women. The risk is proportional to the calcium level. This study should also help guide cliniciansâ&#x20AC;&#x2122; investigation and referral strategies. In men, 81% of the cancer associated with hypercalcaemia was caused by lung, prostate, myeloma, colorectal and haematological cancers. In men with hypercalcaemia, clinicians should consider the risk of the above malignancies, and guide their investigation appropriately. We did not study symptom reporting in our cases, though it is likely that many of the patients harbouring malignancy have additional diagnostic pointers towards that diagnosis. In primary care, there are good investigations for all of these cancers (except for colorectal cancer), so initial cancer investigation may not require specialist referral. In women, these findings require some explanation - but it is likely that primary hyperparathyroidism accounts for much of the benign hypercalcaemia. 2F.3 Systematic review of factors that predict the emotional impact of cancer screening Laura Bedford, Ben Young, Denise Kendrick, Roshan das Nair, John Robertson, Kavita Vedhara University of Nottingham, Nottingham, UK The problem: Cancer screening has been found to have negative and positive emotional consequences, for example, increased anxiety, cancer worry and positive affect. Negative emotional consequences can persist in the longer term and are associated with reduced attendance at future screening. Research has highlighted predictors of the emotional impact of cancer screening, which include components of the screening process (e.g, nature of the screening test), clinical factors (e.g., type of test result), sociodemographic factors (e.g., age), and psychosocial factors (e.g., social support). It is essential that all the evidence in this area is summarised and the most important predictors of the emotional impact of cancer screening are identified in order to develop appropriate interventions and support to reduce any emotional harm associated with screening. The approach: A systematic review was conducted to identify factors that predict the emotional impact of cancer screening. Treatment and observational studies that measured at least one emotional outcome, and predictors of that outcome, following cancer screening in adults (> 18 years) were included. Seven databases were searched for English-language studies published in peer-reviewed journals between 1980 and 2013. Studies were assessed for eligibility and a quality assessment conducted for each included study using the Downs & Black checklist. A narrative synthesis was undertaken due to heterogeneity in cancer being screened for, screening test, participants, and data collection methods. Findings: Of the 5077 articles identified, 56 met inclusion criteria: 28 reported on breast cancer, six on lung cancer, six on colorectal cancer, five on prostate cancer, five on ovarian cancer, four on cervical cancer and two on anal cancer. Studies examined a range of emotional outcomes using measures of generic distress, screen-specific distress, disease-specific distress, illness beliefs, and health-related quality of life. A range of psychosocial factors predicted emotional impacts of screening, some of which included pre-screening cancer worry, pessimism, and coping style. Younger age, previous experience on cancer screening programmes and family history of the cancer screened for were also associated with emotional consequences. Finally, a number of aspects of the screening process predicted emotional impacts, such as receipt of false positive test result and satisfaction with information provided. The results of this review will be discussed by cancer type and the most important predictor factors across cancers will be identified.

Consequences: This review has identified factors that predict the emotional impact of cancer screening. These findings can inform the development of interventions to reduce any emotional harm associated with screening and enable clinicians to identify patients who may be susceptible to emotional harm and the type of harm they may experience. These factors should be considered when developing and improving cancer screening programmes in order to limit emotional harm to patients. 2F.4 Understanding symptom appraisal and help-seeking in people with symptoms suggestive of colorectal cancer: a qualitative interview study Nicola Hall1, Linda Birt2, Katie Mills2, Jon Banks3, Jon Emery4 ,2, Margaret Johnson6, Fiona Walter2 ,5 1 Durham University, Stockton on Tees, UK, 2University of Cambridge, Cambridge, UK, 3University of Bristol, Bristol, UK, 4University of Melbourne, Melbourne, Australia, 5University of Western Australia, Crawley, Australia, 6Lay member of steering group, Cambridge, UK The problem: The UK has lower one year cancer survival for colorectal cancer (CRC) than other comparable countries. Treatment options are greater when CRC is diagnosed at an early stage. Time to seeking healthcare represents the largest proportion of the time period between symptom onset and cancer diagnosis for symptomatic CRC. Although a range of factors have been associated with this time interval, there is limited understanding of how and why people choose to seek healthcare when they do for symptoms which could suggest CRC. The approach: Semi-structured in-depth qualitative interviews with 40 participants referred to gastroenterology clinics in the North East and East of England with symptoms suggestive of CRC (17 female, 43-87 years, 18 diagnosed with CRC, 12 interviewed prior to diagnosis). Purposive sampling ensured that a range of demographics and diagnoses were represented. Patientsâ&#x20AC;&#x2122; symptom appraisal and help-seeking decisions were explored within and across cases using Framework analysis with a particular focus on the intervals, events and processes from the Pathways to Treatment model. Findings: Participants described a range of psycho-social, contextual and symptom specific influences on the processes within the appraisal and help-seeking intervals. Bodily changes were often initially attributed to age, stress, diet or viruses and were appraised within the context of usual bowel patterns, co-morbidities, and life events. For some participants, general or vague changes were only interpreted as symptoms after an association was made to other colorectal changes or after help-seeking, referral or diagnosis. A range of inhibitors and facilitators to seeking care were identified. These included the intermittent or non-specific nature of symptoms, symptom attributions, emotional reactions, significant others, the impact on daily life, lived experience of cancer, expectations of health care, and, for some participants, cancer awareness campaigns. The private or embarrassing nature of symptoms was not a common theme, but influenced healthcare seeking for a minority of participants alongside other factors. Awareness that symptoms may be due to CRC was not always sufficient, nor essential, for a decision to seek care. High cancer concern before presentation was more common amongst those without a cancer diagnosis and was usually related to rectal bleeding. Negative bowel cancer screening results and previously reassuring consultations were reported to have contributed to beliefs that symptoms were not serious and prolonged time to presentation for some participants. Consequences: Understanding the factors which influence the time to presentation for CRC can help to inform public health and primary care strategies to optimise timely symptomatic CRC diagnosis. This study uniquely allowed comparison of accounts between those with and without a cancer diagnosis enabling consideration of potential recall bias and

ABSTRACTS OF ORAL PRESENTATIONS post-diagnosis rationalisation of previous symptom experiences. 2F.5 Building the evidence base for the early symptomatic diagnosis of cancer: systematic review of the effect of time to diagnosis on cancer outcomes Richard Neal1, Barbara France1, Julia Fallon-Ferguson2, Annie Hendry1, Tekendra Rai1, Katie Shaw2, Briony Williams2, Jon Emery3 1 Bangor University, Wales, UK, 2University of Western Australia, Western Australia, Australia, 3University of Melbourne, Victoria, Australia The problem: Symptomatic diagnosis of cancer is important and has been the subject of considerable innovation and intervention in recent years; the English National Awareness and Early Diagnosis Initiative (NAEDI) has made a major contribution to this effort. There is currently no up-to-date overview of the evidence regarding symptomatic presentation and diagnosis to inform clinical practice, policy and the future direction of research. It remains unclear whether more timely cancer diagnosis brings favourable outcomes (stage, survival, other outcomes). The approach: what did you do? We are updating a previous systematic review to determine the impact of different patient, primary care and secondary care intervals (and their combinations) on outcomes. The inclusion criteria are that papers must: F Be full text and available in English F Report patients with a symptomatic diagnosis of primary cancer (screen detected cancers and biomarkers excluded) F Primarily set out to determine the association of at least one time interval to diagnosis or treatment F Report at least one time interval (patient, primary care, secondary care, or a combination) allowing assessment against accepted definitions F Report at least one of morbidity, survival or stage outcome. Searches have been conducted in Medline, Medline -in-Process, Embase, Cinahl, PsycINFO, and the Cochrane Library, 2010-present. All candidate papers have been assessed for eligibility by two researchers independently. Full texts have been obtained and data extracted by one researcher and checked by another. Extracted data includes time intervals, outcomes, and statistical comparisons including confounding by indication. If there is limited heterogeneity, meta-analysis will be undertaken. Papers reporting adjustment for the ‘waiting time paradox’ will be analysed separately as they are likely to be of higher quality. We have developed a review specific bias assessment tool. The review is being undertaken using EPPI-Reviewer software. This facilitates searching, assessment, storage of full text papers, and data extraction and is especially useful for researchers in different locations. Findings: 71 papers have been included. The findings will be reported in keeping with the PRISMA guidelines and on a cancer by cancer basis and across types of intervals (patient, primary care, secondary care, and combinations). The findings will be reported in full at the conference. Consequences: There has been significant investment in early diagnosis initiatives and research, and this is likely to continue. The findings will inform the direction and nature of future research funding, especially by Cancer Research UK, who commissioned this work. 2F.6 GPs, patients and cancer screening Trish Green1, Karl Atkin2, Una Macleod1 Hull York Medical School, Hull, UK, 2University of York, York, UK

The problem: In the UK, diagnostic delay is deemed to account for an estimated 5-10,000 extra cancer deaths each year. Primary care has been identified as an area where explanation for this might be found and is the focus of considerable research and policy activity. Although work to date has included GP perspectives in cancer specific studies there is limited work that investigates GPs’ understanding of awareness and early detection in relation to general practice. The overall purpose of our study was to understand cancer diagnosis from the perspective of GPs, and to develop a framework for conceptualising the potential of the GP role to improve cancer recognition and referral. One area of focus was to explore GPs’ perceptions of asymptomatic cancer screening programmes. Specifically we aimed to identify GPs’ understanding of their role with respect to cancer awareness, screening and early detection, to examine their attitudes to screening, and elicit views on the role of primary care in supporting screening coverage. Importantly, we aimed to explore particular challenges GPs might experience with respect to cancer awareness, screening and early detection. The approach: Individual face-to-face interviews were conducted with 55 GPs (23 women; 32 men), four of whom were GP cancer leads for their areas (2 women; 2 men). We also carried out 18 second interviews over the telephone (8 women; 10 men). Cancer leads were not re-interviewed. Participating practices were in the North and North East of England (n.45) and Greater London (n.10). Purposeful sampling ensured demographic diversity of practices. GP informants were equally diverse regarding gender, ethnicity, age, years in practice and position held. All interviews were recorded and transcribed verbatim. Repeated reading of GPs’ narratives engendered thematic analysis across the interview material, this included GPs’ understandings of developments of screening programmes; reasons for lack of uptake in some patient cohorts; and GPs’ perceptions of their role in improving screening uptake in hard-to-reach patient groups. Findings: Primary care was perceived as the most appropriate host for screening programmes. Because of its embeddedness in primary care and women’s psyche, enhancing the uptake of cervical screening was perceived as more manageable, whereas screening frameworks that were less clear created several challenges. For example, data revealed GP consensus on the issue of patients’ informed choice, but recent breast screening conflicts and debates created difficulties for GPs in face-to-face interactions with their patients. Endorsement of bowel cancer screening also raised several issues ranging from patients’ reluctance to participate, to communicating the implications of positive test results to patients. Consequences: Input from primary care would aid the development of screening programmes and GPs’ understandings of their practice populations would assist in targeting underserved areas that fall below the national average for asymptomatic screening. 3A.1 People with CFS/ME should be managed in primary care: is the NICE guideline wrong? Carolyn Chew-Graham1, Lisa Riste2, Sarah Peters2, Kerin Bayliss2, Alison Wearden2, Rebecca Band2, Louise Fisher2, Karina Lovell2 1 Keele University, Staffs, UK, 2University of Manchester, Manchester, UK The problem: Chronic Fatigue Syndrome (CFS) or Myalgic Encephalomyelitis (ME) is characterized by disabling fatigue not alleviated by rest, and lasts at least six months. Symptoms include headaches, unrefreshing sleep and post-exertional malaise, weakness, pain, sore throat, concentration or memory problems. The condition is distressing and costly in terms of health service utilization and economic burden. The NICE guideline for CFS/ME emphasises the need for an early diagnosis in primary care with management tailored to patient needs. However, GPs are reluctant to make the diagnosis and lack confidence in managing the condition; in addition, the condition is of low prevalence and seen as a contentious illness by some clinicians.

ABSTRACTS OF ORAL PRESENTATIONS Patients are dissatisfied with, and disengage from, primary care. The approach: The study team, with integrated PPI (patient and public involvement) input, developed resources for practitioners and patients to support the diagnosis and management of CFS/ME in primary care. Participating practices in NW England were asked to invite their patients with CFS/ME to take part in evaluation of the patient resources. Patients completed baseline questionnaires, including assessment of health utilisation and satisfaction with care. Patients were alerted to the availability of the resources from their GPs. GPs were invited to complete the on-line training module or participated in face to face training delivered by members of the research team. At follow-up, patients completed the assessment schedules again. Semi-structured interviews with patients and GPs were conducted to evaluate the resources, the training (for practitioners) and how patients and GPs perceived care for CFS/ME had changed. Findings: Recruitment of practices was difficult: only 21 practices (out of 400) in four Primary Care Trusts agreed to participate. 57 patients were recruited from 13 practices. GPs suggested that patients who had CFS/ME were not always READ coded. Patients were not routinely given the resources by their GP, and the resources were not used in partnership with the GP as we had intended. GPs suggested that CFS/ME is not a condition that should be managed in primary care, and were reluctant to encourage patients to attend more frequently. GPs valued the patient resources, but saw them as a means to reduce consultation, rather than re-engage the patient. Patients valued the resources but suggested that GPs continued to be sceptical about CFS/ME and reluctant to discuss management options. Consequences: Managing people with CFS/ME was not seen as a priority for GPs, which meant that there was limited engagement with the study, and either patient or practitioner resources. GPs do not see managing people with CFS/ME as a priority, and patients may remain un-diagnosed and under-treated. A shift in the guideline may be needed so that more appropriate services, outside primary care, are commissioned. 3A.2 Introducing a predictive risk tool in primary care: expectations of practitioners Mark-Rhys Kingston, Alison Porter, Bridie Angela Evans, Hayley Hutchings, Helen Snooks Swansea University, Swansea, UK The problem: A third of emergency hospital admissions are of patients with chronic conditions, many of which are not life threatening, and may result in unnecessary hospital stays. Targeted primary and community care interventions could prevent one in five emergency admissions, improving health outcomes and saving money. Predictive risk (PR) tools identify patients who might benefit from such interventions, and offer a useful resource for GPs in England and Wales to fulfil new contractual expectations to identify and target patients at high risk of emergency admission. However, little is yet known about the expectations and understanding of GPs and their colleagues in relation to PR tools. Normalisation Process Theory (NPT) suggests that for implementation of an innovation to be successful, it needs to be seen by potential users as offering advantages over current practice, while at the same time being workable and integrating with existing systems. As part of the PRISMATIC trial of the Prism PR tool in south Wales, we carried out qualitative work in primary care to explore expectations of its effect on workload, patient care and emergency admissions. The approach: We held focus groups or interviews with 30 GPs, 9 Practice Managers and 3 Practice Nurses. Fieldwork took place before the practice 52

received Prism. We asked about participants’ expectations of Prism use and impact and any concerns. We recorded the groups and interviews, with participants’ consent. We analysed transcripts thematically, informed by NPT, focussing on the how respondents understood the innovation (coherence) and factors affecting their desire to use it (engagement). Findings: Respondents had a good understanding of the Prism risk prediction tool. They generally supported the principle of identifying at-risk patients and were familiar with risk prediction from using condition specific tools. They were also willing to test Prism. Participants felt that it fitted with policy imperatives and regarded Prism as relatively straightforward to install and use. Respondents generally reported that Prism appealed to their desire to improve care of patients at risk of emergency admissions. However they felt challenged by limited capacity, within the practice and among community staff, to respond to identified needs. There were anxieties about raising patient expectations, and about implications in terms of performance management. Consequences: Though respondents were open to trying Prism, its effectiveness will depend on clinicians’ capacity to respond to identified needs and integration with wider community services. Risk prediction tools like Prism are increasingly being introduced because they effectively predict hospital emergency admissions. But their introduction is unlikely to make a difference to patients or resources without good understanding about how practitioners perceive and use them. These findings will have UK and international relevance at a time of heightened focus on chronic conditions management within constrained resources. 3A.3 Managing patients with advanced chronic kidney disease in primary care: a qualitative study with GPs. Sarah Tonkin-Crine1, Miriam Santer1, Geraldine Leydon1, Fliss Murtagh2, Ken Farrington3, Fergus Caskey4, Hugh Rayner5, Paul Roderick1 1 University of Southampton, Southampton, UK, 2Kings College London, London, UK, 3Lister Hospital, Stevenage, UK, 4Southmead Hospital, Bristol, UK, 5Heart of England NHS Foundation Trust, Birmingham, UK The problem: Chronic kidney disease (CKD) can be seen as a relatively new condition for general practice since the introduction of NICE guidelines in 2006. Older adults are the fastest growing population with CKD and are likely to have other comorbidities; as such GPs may experience difficulties when deciding on the best approach to management and referral to secondary care. Exploring GPs’ views of this patient population can help to inform links between nephrology and primary care in order to promote timely referral rates. The approach: Semi-structured interviews with GPs recruited from primary care practices in four areas of the UK. GPs were asked about their experience of managing older adults with CKD stage 4 and 5 and decisions about referring such patients to secondary care. Interviews were audio recorded and transcribed verbatim. Transcripts were analysed following thematic analysis. Findings: Nineteen interviews were completed. Few GPs had experience managing patients with CKD 4 or 5 and consequently welcomed guidance, in particular from their local renal unit. There was variation in GPs’ accounts regarding referral decisions. Some GPs referred based on the renal function alone, some used wider criteria including age and co-morbidity and some learnt from experience that patients referred according to national guidelines were not infrequently discharged immediately. GPs with more experience of managing older adults felt more comfortable referring patients later, or sometimes not at all, if there were no

ABSTRACTS OF ORAL PRESENTATIONS additional problems such as uncontrolled hypertension. GPs reported that they were happy to treat patients with CKD approaching end of life but stressed that they had little experience of placing patients on a palliative care register for renal disease alone. GPs reported that they would not feel confident managing patients who had opted not to have dialysis unless there was clear communication from the renal unit regarding this.

Consequences: Anxiety remains a hidden symptom in people with osteoarthritis. The use of case-finding for anxiety and depression in people with chronic physical conditions will only be of use if the importance of identification and management of anxiety and depression are recognised by GPs, who then prioritise management of the patient holistically, regardless of age or gender.

Consequences: GPs make decisions about referrals to nephrology services on an individual patient basis and consequently not all CKD 4 and 5 patients are likely to be known to secondary care. While GPs have access to national guidelines they felt better supported by additional advice from their local renal unit. GPs may further benefit from guidance from their local unit to clarify referral criteria and guidance on the management of older adults with comorbidities.

3A.5

3A.4 Prioritising the physical: a qualitative study into case finding for depression and anxiety in osteoarthritis in primary care Bernadette Bartlam1, Carolyn Chew-Graham1, Daniel Green1, Caroline Mitchell2, Barbara Nicholl3, Jane Richardson1, Christian Mallen1 1 Keele University, Keele, UK, 2University of Sheffield, Sheffield, UK, 3 University of Glasgow, Glasgow, UK The problem: Osteoarthritis is a common reason for primary care consultation in the UK, with approximately one million seeking treatment annually. Depression is up to four times more common in individuals with persistent pain. Treating depression improves outcomes, including pain and function, but detection is poor, with only half of all cases being identified. Moreover, the majority of work on comorbid mood disorders in primary care has focused on depression despite anxiety being more common amongst patients with osteoarthritis. NICE guidelines recommend case finding for depressive and anxiety symptoms in people with long-term conditions yet this is not routinely carried out for those with osteoarthritis. The approach: This qualitative study was nested in a cluster randomised trial, which used electronic templates to prompt GPs to ask patients with osteoarthritis up to 5 questions in a primary care consultation. In the intervention arm GPs used validated case finding questions for depression (PHQ-2) and anxiety (GAD-2), with a fifth question on pain intensity. In the control arm GPs were prompted to ask about pain intensity only. A brief questionnaire exploring attitudes to the study and its impact on clinical practice was distributed to all participating GPs, resulting in a 50% response rate (82/165). Twenty-five GPs, from across both arms, who consented to further contact, were interviewed to explore their perspectives on the trial, and on the detection and management of anxiety and depression in patients with osteoarthritis. Applied thematic analysis with constant comparison was used to analyse the interview data; the questionnaire data were descriptively analysed. Findings: Participation in the trial appears to have raised GPs’ awareness of depression in people with osteoarthritis; although less so, anxiety. It appears that low mood in patients with osteoarthritis is normalised by GPs, particularly in older people; and depression associated with osteoarthritis was seen as ‘understandable’, and predictable and not needing treatment, unlike depression without an underlying physical cause. The importance of detecting and managing anxiety was generally unrecognised. Prioritization was given to controlling the physical symptoms of pain due to osteoarthritis over the detection and management of co-existing depression or anxiety, the latter being described as ‘small fish’ in comparison. GPs described differences in both their communication and management approaches to patients, depending on age and gender, with women and older people seen as less in need of active intervention either for osteoarthritis or mental health problems.

Feasibility Study: Blood pressure (BP) telemonitoring for people with previous stroke/ transient ischaemic attack (TIA) Janet Hanley1, Hilary Pinnock2, Cathie Sudlow2, Paul Padfield2, Lucy McCloughan2, Mary Paterson2, Allison Todd3, Peter Fairbrother3, Brian McKinstry2 1 Edinburgh Napier University, Edinburgh, UK, 2University of Edinburgh, Edinburgh, UK, 3NHS Lothian, Lothian, UK The problem: BP telemonitoring in primary care is an effective way of improving control. However, it cannot be assumed that the benefits seen in the general hypertensive population will necessarily apply in people with stroke or TIA (even though they should have more to gain from good BP control), since they may be more debilitated and take more medications than the populations previously included in telemonitoring trials. Furthermore, clinicians may be more cautious about lowering BP in these patients and, while only a minority of stroke and TIA survivors have severe residual disability, minor physical or cognitive disability may reduce ability to use a BP monitoring system. The approach: We aimed to test the feasibility of mounting a pragmatic randomised controlled trial based of an established health technology (BP telemonitoring) in patients who had previous stroke/TIA in a UK primary care setting, using a protocol which had been effective in our previous trial which had excluded this client group. This included: a 6 month pilot trial in 6 general practices involving 23 stroke and 22 TIA patients whose most recent surgery BP measurement and subsequent daytime average ambulatory systolic BPs both exceeded 130 mmHg, randomised in a ratio of 3:1 to telemonitoring or usual care; interviews with 18 pilot trial patients and 3 nurses; focus groups with 23 patients and carers attending stroke support groups. Findings: 44% of those approached were randomised; 93% were fully followed up and 93% of those who were telemonitoring were fully compliant with the monitoring regimen. Only 2 of those randomised needed assistance with the equipment. There was a mean reduction of 10.1 mmHg in the daytime average ambulatory systolic BP in the telemonitoring group compared with 3.8 mmHg in the control group. As this pilot was not conducted or powered to detect effects on BP, we did not carry out formal tests of statistical significance. The qualitative study showed that in the stroke support groups there were a number of people who could not use the equipment and would not have access to help. Both they and the trial participants worried about high BP causing further stroke but many said their healthcare practitioners did not engage them in their BP management, often not even telling them what the readings were. They found telemonitoring empowering, reassuring and much more convenient than surgery visits. Consequences: A trial of BP telemonitoring for stroke/ TIA patients in primary care is feasible and likely to be welcomed by this patient group However, should BP telemonitoring become standard care, there is a small group of patients who would be unable to use the equipment, would not have access to help and would require alternative arrangements. 3A.6 Modification of cardiovascular risk scores for use with home blood pressure measurements

ABSTRACTS OF ORAL PRESENTATIONS Sarah Stevens1, Richard Stevens1, Richard McManus1, Ben Cairns2 1 Dept. of Primary Care Health Sciences, University of Oxford, Oxford, UK, 2 Cancer Epidemiology Unit, University of Oxford, Oxford, UK The problem: Established cardiovascular risk calculators (e.g. Framingham (1991), QRisk2) were developed using blood pressure (BP) measurements taken by a physician or nurse in a clinic setting and may not be suitable for use with BP measurements taken by a patient at home. However, for patients with masked or “white-coat” hypertension, home BP monitoring is needed to reliably establish their underlying BP and risk calculators should be able to accommodate this variation in BP monitoring techniques. The approach: We studied how the SCORE, Framingham (1991) and QRisk2 risk equations could be adapted for use with home blood pressure measurements, allowing for a) systematic differences between clinic and home blood pressure on average, and b) altered within-person variability. Data on 754 patients from a cross-sectional, BP monitoring study (BP-Eth) was analysed using multi-level mixed effects linear regression. The between-person variability in BP, two measures of within-person variability (one each for clinic and home BP) and the average difference between home and clinic BP was estimated. Regression dilution methods were then used to modify the coefficients for BP in the risk equations and an offset was calculated to ensure that the average population level risk remained constant between the existing and modified scores. Findings: We estimated that the coefficient for systolic BP in the SCORE risk equation should be multiplied by 1.33 (95% CI 1.17 to 1.48). This implies that if the SCORE risk calculator had been developed using cohort data with home systolic BP measurements, the coefficient would be approximately 33% greater than in the existing equation. Additionally, we estimated that home BP measurements are 2.38mmHg lower than clinic BP measurements on average. The difference between the existing and modified SCORE equation can be demonstrated through an individual patient example. According to the existing SCORE equation, a 57 year-old, non-smoking male without diabetes, with total cholesterol = 6.5 mmol/L and clinic systolic BP of 157mmHg, has a 5.7% risk of fatal cardiovascular disease within 10 years. Assuming a “white coat” effect and home systolic BP of 133mmHg, the modified SCORE equation estimates 10-year risk to be 3.9%. Modification of the Framingham and QRisk2 risk equations is near completion and a validation study is in its development stages. Consequences: The modified risk equations will provide GPs with an alternative method of cardiovascular risk prediction when a patient’s home BP measurements are available. This will be particularly useful for risk prediction in patients susceptible to masked or “white-coat” hypertension, whose underlying BP is more accurately measured through homemonitoring. A more accurate risk assessment in these patients will lead to better patient management decisions, for example initiation of antihypertensive or statin therapies. 3B.1 Age and gender variations in cancer diagnostic intervals: analysis of data from the UK General Practice Research Database Nafees Din1, Obioha Ukoumunne2, Greg Rubin3, William Hamilton2, Ben Carter4, Sal Stapley2, Richard Neal1 1 Bangor University, Wrexham, UK, 2University of Exeter, Exeter, UK, 3 Durham University, Durham, UK, 4Cardiff University, Cardiff, UK The problem: Time from symptomatic presentation to cancer diagnosis (diagnostic interval) is an important, and modifiable, part of the patient’s cancer journey, and can be affected by factors such as age, gender and type 54

of presenting symptoms. Using data from general practitioner records, we quantified inequalities in the diagnostic interval related to these factors. The approach: Data from the UK General Practice Research Database for 15 types of incident cancer (bladder, breast, cervical, colorectal, endometrial, gastric, head & neck, kidney, lung, leukaemia, lymphoma, myeloma, oesophageal, pancreatic, testicular) diagnosed between 2007 and 2010 were analysed. Symptom lists for each cancer were categorised into either NICE qualifying (NICE) or not (not-NICE) based on NICE Urgent Referral Guidelines for Suspected Cancer criteria. Multivariable (adjusted) linear regression models were fitted to examine the relationship of diagnostic interval (outcome) with age, gender and symptom type (predictors). Analyses were carried out both separately for each cancer site and then pooled across cancers. Findings: 18,618 newly diagnosed cancer patients aged ≥40 who had a recorded symptom in the preceding 12 months were included. Age showed a linear relationship with diagnostic interval for all cancers, except colorectal. For the latter cancer site, age was categorised into five age groups based on quintiles and the youngest used as reference in the regression analyses. Diagnostic interval increased with increasing age for five cancers: bladder (P<0.001), kidney (P<0.001), leukaemia (P<0.001), lung (P<0.001), and colorectal (P<0.001); and decreased for gastric (P=0.03) and pancreatic (P=0.03) cancer. A longer diagnostic interval was found for females in six of the gender non -specific cancers: bladder (P=0.04), colorectal (P=0.001), gastric (P=0.03), head and neck (P=0.02), lung (P=0.02), and lymphoma (P=0.01). Diagnostic intervals were shorter (P<0.05) for NICE symptoms in 10 (bladder, breast, cervical, colorectal, endometrial, gastric, kidney, oesophageal, pancreatic, testicular) of the 15 cancers independent of the age or gender of patients. The pooled analysis across all cancers demonstrated strong evidence of differences overall with longer diagnostic intervals for older patients (7.9 days per year; 6.5 to 9.2; P<0.001); females (10.2 days; 6.9 to 13.6; P<0.001); and not-NICE symptoms (27.7 days; 23.7 to 31.6; P<0.001). Conclusions: Variations were found in cancer diagnostic intervals for age and gender in some cancers. Further research is needed to find out why these differences occur. Interventions should be developed to provide equitable speed of diagnosis across all age and sex groups. 3B.2 A Description of the EarlyCDT-Lung test in Scotland (ECLS) Trial : a case-finding method in lung cancer. Frank Sullivan2, Frances Mair1, Denise Kendrick3, Alistair Dorward7, Stuart Schembri8, John Robertson3, Kavita Vedhara3, Herb Sewell3, Shaun Treweek5, Colin McCowan1, Chris Robertson6, Mike Sproule7, Andy Briggs1, Lewis Ritchie5, Robert Milroy7, Tom Taylor8, Roberta Littleford4, David Brewster9 1 University of Glasgow, Glasgow, Scotland, UK, 2University of Toronto, Toronto, Canada, 3University of Nottingham, Nottingham, England, UK, 4 University of Dundee, Dundee, Scotland, UK, 5University of Aberdeen, Aberdeen, Scotland, UK, 6University of Strathclyde, Glasgow, Scotland, UK, 7 NHS Greater Glasgow and Clyde, Glasgow, Scotland, UK, 8NHS Tayside, Dundee, Scotland, UK, 9Scottish Cancer Registry, Scotland, UK The problem: Lung cancer is the leading cause of cancer death worldwide. The majority of lung cancer cases are detected at a late stage, so the prognosis is poor. The National Lung Screening Trial (NLST) recently reported on using Computed Tomography (CT) with a 20% reduction in lung cancer mortality. However CT is expensive as a primary screening modality and leads to substantial morbidity in individuals whose tests are false positives. The EarlyCDT-lung test detects autoantibodies to tumour

ABSTRACTS OF ORAL PRESENTATIONS antigens in the earliest stages of the disease with a specificity of 93%. This randomised controlled trial (RCT) aims to determine whether using the EarlyCDT-Lung test in high-risk patients can reduce the incidence of late-stage lung cancer (3 & 4) or unclassified presentation (U) at diagnosis, compared with standard practice. The approach: We are conducting a RCT of 10 000 participants in areas of Scotland within the lowest quintile of deprivation. Adults aged 50 to 75 who are at around a 2% risk of lung cancer over the next 2 years are eligible to participate. They should be healthy enough to undergo curative interventions. We will compare the EarlyCDT-lung test and follow-up imaging with standard clinical practice. The primary outcome is the difference, after 24 months, between the rates of patients with stage 3, 4 or unclassified lung cancer at diagnosis. Participants who develop lung cancer will be followed-up via electronic record-linkage to cancer registry to assess both time to diagnosis and stage of disease at diagnosis. The secondary outcomes are cost-effectiveness, and a range of psychological measurements. There is a nested qualitative study of the psychological effects test results have on participants. Findings: In the first 6 months of the trial 1477 patients have been recruited (representing a 20% recruitment rate). Seventy nine positive blood results have been identified; 56 CXRs and 50 CT scans have been undertaken; and 22 abnormalities identified which are undergoing further investigation. Twelve per cent the intervention group tested have had a positive EarlyCDTlung test. The Consequences: The study will determine the EarlyCDT-Lung test’s clinical and cost effectiveness. It will also assess potential morbidity arising from the test and potential harms and benefits of a negative EarlyCDT-Lung test result. The study therefore has the potential to have a major influence on future screening strategies. 3B.3 Development and validation of a clinical prediction rule to identify suspected breast cancer: a prospective cohort study Rose Galvin1, Doireann Joyce2, Fiona Boland1, Tom Fahey1, Arnold Hill2, Kirsty O’Brien1 1 Royal College of Surgeons in Ireland, Dublin, Ireland, 2Department of Surgery, Beaumont, Dublin, Ireland The problem: The number of primary care referrals of women with breast symptoms to symptomatic breast units (SBU) has increased yet the number of new cases with breast cancer has remained the same. The aim of this study is to develop and validate a clinical prediction rule (CPR) to identify women with breast cancer using routine data from an Irish national SBU so that a more evidence based approach to referral can be developed. The approach: We analysed routine data from a prospective cohort of consecutive women reviewed at a SBU with breast symptoms. The dataset was split into a derivation and validation cohort. Regression analysis was used to derive a CPR from the patient’s history and clinical findings. Validation of the CPR consisted of estimating the number of breast cancers predicted to occur compared with the actual number of observed breast cancers across deciles of risk. Findings: A total of 6,590 patients were included in the derivation study and 4.9% were diagnosed with breast cancer. Independent clinical predictors for breast cancer were: increasing age by year (adjusted odds ratio 1.08, 95%CI 1.07-1.09); presence of a lump (5.63, 95%CI 4.2-7.56); nipple change (2.77, 95%CI 1.68-4.58) and nipple discharge (2.09, 95% CI 1.1-3.97). Validation of the rule (n=911) demonstrated that the probability of breast cancer was higher with an increasing number of these

independent variables. The Hosmer-Lemeshow goodness of fit showed no overall significant difference between the expected and the observed numbers of breast cancer (ß2HL: 6.74, p-value: 0.5647). Consequences: This study derived and validated a CPR for breast cancer in women attending an Irish national SBU. We found that increasing age, presence of a lump, nipple discharge and nipple change are all associated with increased risk of breast cancer. Further validation of the rule is necessary as well as an assessment of its impact on referral practice. 3B.4 Symptom interpretation and decision-making processes in patients with lung or colorectal cancer Sarah McLachlan1, Gemma Mansell1, Tom Sanders1, Sarah Yardley1, Danielle van der Windt1, Carolyn Chew-Graham1, Paul Little2, Lucy Brindle2 1 Keele University, Staffordshire, UK, 2University of Southampton, Southampton, UK The problem: Lung and colorectal cancer are two of the most common cancers diagnosed in men and women in the UK and both have high mortality rates, relative to other forms of cancer. Evidence suggests that early diagnosis is critical in reducing cancer mortality. Delay in presentation of symptoms in primary care is widely reported in the literature on lung and colorectal cancers, and perceived barriers to symptomatic presentation are higher in the UK than in other countries. The Department of Health national cancer strategy (2011) emphasised the importance of encouraging people to recognise symptoms and seek help quickly to facilitate early diagnosis. This study aimed to explore patients’ interpretation of symptoms and decision-making to gain a richer understanding of the processes leading to help-seeking. The approach: Semi-structured interviews were conducted with nine patients with lung cancer and 20 patients with colorectal cancer who had been diagnosed within the previous 12 months. Patients were asked about symptoms experienced in the period preceding diagnosis, how these were interpreted, and triggers for help-seeking. Interviews were audio-recorded and transcribed verbatim. Thematic analysis (using constant comparative methodology) was conducted and comparisons were drawn within and across patient groups to explore patterns in the data. This interview study was part of CANDID, a large prospective cohort study to develop cancer diagnosis decision rules led by the University of Southampton. Findings: Emergent themes and patterns were used to develop a model of symptom interpretation and help-seeking, from an initial pre-symptom perception phase through to patients’ reflections with hindsight following diagnosis. Findings indicated that patients took an active and rational approach to addressing symptoms. Many had engaged in a “watch and wait” exercise in monitoring symptoms, similar to the approach taken by GPs. In seeking to make sense of symptoms, patients developed alternative, non-cancer explanations which were logical and based on knowledge, past experience and discussion with others. Patients tended to adopt benign explanations, or to link symptoms with comorbidities, side-effects of medications, or ageing. In addition to sanctioning help-seeking, significant others sometimes contributed to delays by endorsing alternative accounts of symptoms. Many patients sought help when alternative explanations for symptoms became unviable, often as a result of symptom persistence, exacerbations, or ineffective attempts to self-medicate. Perceptions of risk and fear that symptoms may indicate cancer also influenced help-seeking decisions when alternative explanations could not be maintained. Consequences: Findings suggest that prompting patients for further detail on symptoms within consultations may provide critical contextual information to aid referral decisions. The model of symptom interpretation 55

ABSTRACTS OF ORAL PRESENTATIONS and help-seeking developed from this study could be used by primary care practitioners to facilitate this process. The results also have implications for public health campaigns for lung and colorectal cancers.

3B.6 Explaining cancer screening attendance decisions in the UK: A systematic search and meta-synthesis of qualitative evidence

3B.5 “The tests always came back negative you know” The influence of a negative bowel screening result among patients who subsequently develop colorectal cancer. Karen Barnett1, David Weller1, Stephen Smith2, Robert Steele3, Sheina Orbell4, Jane Melia5, Sue Moss6, Julietta Patnick7, Peter Vedsted8, Christine Campbell1 1 University of Edinburgh, Edinburgh, UK, 2University Hospitals Coventry & Warwickshire NHS Trust, Rugby, UK, 3University of Dundee, Dundee, UK, 4 University of Essex, Essex, UK, 5University of Cambridge, Cambridge, UK, 6 Queen Mary University of London, London, UK, 7NHS Cancer Screening Programmes, Sheffield, UK, 8University of Aarhus, Aarhus, UK The problem: Colorectal cancer is a significant cause of mortality and morbidity in the UK. Bowel cancer screening, which has the potential to significantly reduce mortality from colorectal cancer, has been introduced across the UK. However, approximately 40% of cancers will not be detected by the Faecal Occult Blood test (FOBt), the national screening test, and therefore there is a need for awareness of the symptoms of colorectal cancer among screening participants as well as the general population. The aim of the study is to examine how screening participants who develop interval cancers (a diagnosis of colorectal cancer after a negative FOBt) respond to any bowel symptoms and whether their response is influenced by their negative screening result. Specifically, we are exploring if these individuals ‘downplay’ any symptoms, or defer taking action, through a misunderstanding of the significance of a negative result, or conversely, if participation in the screening programme raises awareness about symptoms of bowel cancer and makes people more inclined to consult a GP with symptoms. The approach: Semi-structured qualitative face-to-face interviews are being carried out with individuals who have been diagnosed with an interval bowel cancer following symptomatic presentation to primary care (approximately 20 in Scotland, and 20 in the Midlands). Participants are being recruited through hospitals associated with the Scottish Bowel Screening Centre, and the Midlands and North West Screening Hub. Interviews are carried out within 6 months of diagnosis to minimise recall bias. Interviews last approximately one hour, and are transcribed and subsequently analysed following the principles of thematic analysis. Findings: Sixteen interviews have been carried out to date (9 in Tayside, Scotland and 6 in North-West Midlands, England; 8 female, 8 male; ages 52-77). Interviews and analyses are ongoing. Emerging findings indicate a number of themes which are being explored. Interviewees in general obtain reassurance from a negative screening result: many are unaware of the limitations of screening, and some are explicit that this led to them downplaying the importance of subsequent symptoms. Most had poor recall of the advice on bowel symptoms provided with their negative result. Despite this many remain supportive of screening. Consequences: This study will provide an in-depth understanding of how bowel screening participants interpret a negative result in relation to their risk of bowel cancer, and how a negative result influences how people respond to bowel symptoms. Results from the patient interviews will be triangulated with interviews with GPs, and with results from a database linkage study and focus group study which are also investigating the response to bowel symptoms following a negative FOBt. Together, these will inform information provision strategies within bowel cancer screening programmes and primary care.

Ben Young, Laura Bedford, Denise Kendrick, Roshan das Nair, John Robertson, Kavita Vedhara University of Nottingham, Nottingham, UK The problem: There are an estimated 7.6 million deaths caused by cancer annually worldwide. Screening is a key strategy for the reduction of cancer mortality through early detection. To be effective in reducing mortality and/ or morbidity, screening tests must achieve high uptake and test methods must be acceptable to the population. Those at highest risk of a disease are least likely to attend for screening. Rates of uptake of routine cervical, breast and bowel cancer screening in England are 78%, 73% and 55% respectively. Our understanding of why a significant proportion of people choose not to be screened for cancer is drawn largely from a body of quantitative research and isolated qualitative studies. To date, no attempt has been made to synthesise the available qualitative evidence, in order that it can make a contribution to improving the uptake of cancer screening tests. The approach: A systematic search of the literature was conducted for all qualitative evidence explaining individuals’ decisions about participation in routine cancer screening or a cancer screening trial in the UK. Six databases and a range of journals and electronic sources were searched for relevant articles. Titles and abstracts were scanned for relevance and full text papers were retrieved and assessed for inclusion. Included studies were assessed for quality. Quotes from participants and analytic commentary by authors were extracted as data of interest and synthesised using the approach of meta-ethnography. Findings: 10,230 papers were screened for inclusion and 32 studies were included in the meta-synthesis. Included studies reported reasons for attendance or non-attendance at screening tests for breast, cervical, bowel, prostate, ovarian and lung cancer. Common explanations for attendance decisions included factors relating to perceptions of one’s health, relationships with the health care system, fear, embarrassment and practical reasons. Previous screening experiences were commonly cited and cultural factors were also important for some groups. Additional factors revealed a complex picture of the perceptions and concerns of individuals on receipt of cancer screening invitations. Themes will be presented along with extracts of raw data cited in papers and the original authors’ commentary. Consequences: The findings will be of interest to all planners and implementers of cancer screening trials and programmes worldwide in order to meet the global challenge of preventing cancer through early detection. An in-depth understanding of cancer screening attendance decisions could inform modifications to the nature and content of screening invitations to address common perceived barriers to screening, particularly in hard to reach groups. Organised screening activities, such as the NHS National Screening Programmes for breast, cervical and bowel cancer, could draw on this knowledge to improve uptake, leading to potential advances in their effectiveness and cost effectiveness. 3C.1 Derivation and validation of an algorithm to predict hospitalisation in children presenting to primary care with acute cough and RTI: the ‘TARGET’ prospective cohort study Alastair D. Hay1, Andrew M. Lovering2, Barry Vipond3, Brendan Delaney4, Hannah Christensen1, Hannah V. Thornton1, John P. Leeming2, Margaret Fletcher5, Matthew Thompson6, Niamh M. Redmond1, Paul Little7, Peter Muir3, Sophie Turnbull1, Tim J. Peters1, Peter S. Blair

ABSTRACTS OF ORAL PRESENTATIONS University of Bristol, Bristol, UK, 2North Bristol NHS Trust, Bristol, UK, Public Health England, Bristol, UK, 4Kings College London, London, UK, 5 University of the West of England, Bristol, Bristol, UK, 6University of Oxford, Oxford, UK, 7University of Southampton, Southampton, UK

Family Medicine, Uppsala University, Uppsala, Sweden, 7Bio Imaging and Laboratory Medicine Unit, Södra Älvsborg Hospital, Borås, Sweden, 8 Department of Infectious Diseases, Institute of Biomedicine, the Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden

The problem: There is international concern regarding antimicrobial resistance. Children with respiratory tract infections (RTIs) are the single most frequent patient group using primary care. Lack of evidence for which children are at risk of poor outcome in the days following primary care assessment is cited by GPs and NICE as a key reason for the over (defensive) use of antibiotics. The TARGET study aimed to derive and validate an algorithm to identify children at risk of future hospitalisation.

The problem: There are several risk factors for the colonisation, infection and spreading of antibiotic resistant bacteria among elderly residents of nursing homes. An updated estimate of the native prevalence of antimicrobial resistance in uropathogens among Swedish nursing home residents is needed.

1 3

The approach: TARGET was a multicentre (Bristol, London, Oxford and Southampton), prospective, cohort study of children (3 months to 16 years) presenting to primary care with a cough and RTI. Parent-reported symptoms, clinical signs and demographic data were recorded at presentation and primary care medical notes reviewed for RTI hospitalisations in the 30 days post recruitment (the primary outcome). An algorithm was derived using logistic regression and internally validated using bootstrapping and prognostic utility assessed using the area under the ROC curve (AUROC). Findings: 8413 children were recruited with primary outcome data available for 100%. 3113 (37%) were prescribed an immediate or delayed antibiotic and 78 (0.9%) were hospitalised (15 on day 0, 15 day 1, 10 day 2 and 38 thereafter). Seven factors were found to be independently associated with hospitalisation: age <2 years; odds ratio 3.52 (95% CI 2.13 to 5.81, p<0.001); illness duration ≤3 days OR 2.89 (1.81 to 4.57, p<0.001); parent reported ‘severe fever’ OR 2.42 (1.28 to 4.58, p=0.007); parent reported moderate/severe vomiting in the last 24 hours OR 2.60 (1.51 to 4.45, p=0.001); asthma OR 4.03 (2.21 to 7.34, p<0.001); inter/ subcostal recession OR 4.26 (2.42 to 7.52, p<0.001); and wheeze OR 2.17 (1.27 to 3.70, p=0.004), with an AUROC curve of 0.82 (0.78 to 0.87). Since the seven factors had similar strengths of association (and did not change with bootstrapping), a ‘rule of thumb’ was constructed with each factor contributing a value of 1. This was used to stratify the odds of hospitalisation into: low (72% children scored ≤1 (35% of these were prescribed an antibiotic); hospitalisation odds 1:332); medium (26% scored 2 or 3; (42% antibiotic) hospitalisation odds 1:57); and high (2% scored ≥4 (59% antibiotic); hospitalisation odds 1:7). Consequences: We have identified symptoms and signs with good prognostic utility to stratify the risk of hospitalisation within 30 days. The algorithm is being incorporated into the development of an intervention to improve the use of antibiotics. Assuming antibiotic prescribing was halved for low risk children, remained the same for medium risk children and increased to 90% for high risk children, overall antibiotic prescribing would fall by 10%. 3C.2 Antimicrobial resistance in urinary pathogens among Swedish nursing home residents remains low: a cross-sectional study comparing antimicrobial resistance from 2003 to 2012 Pär-Daniel Sundvall1 ,2, Marie Elm3, Ronny Gunnarsson4 ,1, Sigvard Mölstad5, Nils Rodhe6, Lars Jonsson7, Peter Ulleryd8 1 Department of Public Health and Community Medicine, Institute of Medicine, the Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden, 2Research and Development Unit, Primary Health Care in Southern Älvsborg County, Borås, Sweden, 3Health Care Unit, Borås Municipality, Borås, Sweden, 4Cairns Clinical School, School of Medicine and Dentistry, James Cook University, Cairns, Australia, 5 Department of Clinical Sciences, General Practice, Lund University, Malmö, Sweden, 6Department of Public Health and Caring Sciences,

The approach: Urine specimens were collected for culture and antimicrobial susceptibility testing against mecillinam, ampicillin, cefadroxil, trimethoprim, nitrofurantoin and quinolones from the residents of 32 and 22 nursing homes, respectively. The residents were capable of providing a voided urine sample in 2003 and 2012. In 2012 urine specimens were also collected from residents with indwelling urinary catheters. Any antibiotic treatment during the previous month was registered in 2003 as well as hospitalisation and any antibiotic treatment during the previous six months in 2012. Findings: The proportion of positive urine cultures was 32% (207/651) in voided urine specimens in 2003, 35% (147/421) in 2012, and 46% (27/59) in urine samples from catheters in 2012. There were growths of mixed bacterial flora in all but one of the cultures obtained from catheters, classified as negative. Escherichia coli (E. coli) was the most commonly occurring bacteria. Resistance rates in E. coli (voided urine specimens) in 2012 were; ampicillin 21%, trimethoprim 12%, mecillinam 7.7%, ciprofloxacin 3.4%, cefadroxil 2.6% and nitrofurantoin 0.85%. There were no significant changes in the average resistance rates in E. coli for antibiotics tested 2003-2012. There was a trend towards higher resistance rates in E. coli in urine specimens from catheters compared to voided urine for ampicillin (p=0.079) and ciprofloxacin (p=0.11). In 2012, two isolates of E. coli produced extended spectrum beta-lactamase enzymes (ESBL) and one with plasmid mediated AmpC production. Any antibiotic treatment during the previous month increased the risk for resistance in E. coli, adjusted for age and gender; for mecillinam with an odds ratio (OR) of 7.1 (2.4-21; p=0.00049), ampicillin OR 5.2 (2.4-11; p=0.000036), nalidixic acid OR 4.6 (1.4-16; p=0.014) and trimethoprim OR 3.9 (1.6-9.2; p=0.0023). Hospitalisation during the previous six months increased the risk for antibiotic resistance in E. coli to ampicillin, ciprofloxacin and any antimicrobial tested, adjusted for age, gender and antibiotic treatments during the previous six months. Consequences: There were still comparatively low levels of antimicrobial resistance in urinary pathogens among Swedish nursing home residents with no major changes between 2003 and 2012. This may be partly due to successful efforts in Sweden to lower antibiotics usage and the choice of narrow spectrum antibiotics in favour of e.g. ciprofloxacin. Any antibiotic courses during the previous month predicted higher resistance rates in this study. It is important to use antibiotics properly and continue analysing antimicrobial resistance in nursing homes due to the potentially high risk for increasing antibiotic resistance in this population. 3C.3 Interleukin-6 concentrations in the urine and dipstick analyses related to bacteriuria but not symptoms in the elderly: a cross sectional study of 421 nursing home residents Pär-Daniel Sundvall1 ,2, Marie Elm3, Peter Ulleryd4, Sigvard Mölstad5, Nils Rodhe6, Lars Jonsson7, Bengt Andersson8, Mirjana Hahn-Zoric8, Ronny Gunnarsson9 ,1 57

ABSTRACTS OF ORAL PRESENTATIONS Department of Public Health and Community Medicine, Institute of Medicine, the Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden, 2Research and Development Unit, Primary Health Care in Southern Älvsborg County, Borås, Sweden, 3Health Care Unit, Borås Municipality, Borås, Sweden, 4Department of Infectious Diseases, Institute of Biomedicine, the Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden, 5Department of Clinical Sciences, General Practice, Lund University, Malmö, Sweden, 6Department of Public Health and Caring Sciences, Family Medicine, Uppsala University, Uppsala, Sweden, 7Bio Imaging and Laboratory Medicine Unit, Södra Älvsborg Hospital, Borås, Sweden, 8Department of Clinical Immunology, the Sahlgrenska Academy, the University of Gothenburg, Gothenburg, Sweden, 9Cairns Clinical School, School of Medicine and Dentistry, James Cook University, Cairns, Sweden 1

The problem: A complementary test in order to discriminate between symptomatic urinary tract infections and asymptomatic bacteriuria is needed. Previous studies indicate that concentrations of Interleukin-6 (IL-6) in the urine might be used as such a test. The aim of this study was to investigate the association between laboratory findings of bacteria in the urine, elevated concentrations of IL-6 in the urine, dipstick urinalyses and new or increased symptoms in residents of nursing homes for elderly. The approach: In this cross sectional study voided urine specimens were collected from all those able to provide a urine sample, regardless of the presence of symptoms, for urine cultures, urine dipstick and IL-6 analyses. The following new or increased symptoms occurring during the previous month were registered; fatigue, restlessness, confusion, aggressiveness, loss of appetite, frequent falls, not being herself/himself, dysuria, urinary urgency and urinary frequency. Voided urine specimens and symptom forms were provided from 421 elderly residents of 22 nursing homes. Findings: Recent onset of nonspecific symptoms (symptoms not specific for the urinary tract) was common among residents at nursing homes (85/421). Urine cultures were positive in 32% (135/421), and Escherichia coli was by far the most common bacterial finding. There were no significant differences in the proportion of positive urine cultures among those with or without nonspecific symptoms. Residents with positive urine cultures had higher concentration of IL-6 in the urine (p<0.001). In residents exclusively with bacteriuria there were no significant differences in concentrations of urine IL-6 when comparing those with or without a new or increased symptom; fatigue (p=0.24), restlessness (p=0.40), confusion (p=0.38), aggressiveness (p=0.66), loss of appetite (p=0.27), frequent falls (p=0.15), not being herself/himself (p=0.90), having any of the nonspecific symptoms (p=0.69), dysuria (p=0.13) and urinary urgency (p=0.82). Neither were there any differences in urine dipstick analyses regarding nitrite and leukocyte esterase ≥3+. Consequences: Residents with positive urine cultures had higher concentrations of IL-6 in the urine. However among residents with positive urine cultures there were no differences in IL-6 concentrations between those with or without nonspecific symptoms. Thus IL-6 concentrations are not useful when assessing elderly residents with new or increased symptoms and bacteria in the urine. Neither are urine nitrite and leukocyte esterase dipsticks useful to differentiate between asymptomatic bacteriuria and cystitis in residents with bacteriuria. Recently onset nonspecific symptoms were common among elderly residents of nursing homes. Residents without nonspecific symptoms had positive urine cultures as often as those with nonspecific symptoms suggesting that nonspecific symptoms are not caused by bacteria in the urine. The implications of current treatment policies for nonspecific symptoms in nursing homes will be discussed.

3C.4 Antibiotic treatment for acute sore throat in primary care: antibiotic choice duration and effect Michael Moore1, Beth Stuart1, Richard Hobbs2, Chris Butler2, Alastair Hay3, John Campbell4, Brendan Delaney5, David Mant2, Hazel Everitt1, Ian Williamson1, Paul Little1 1 University of Southampton, Southampton, UK, 2University of Oxford, Oxford, UK, 3University of Bristol, Bristol, UK, 4University of Exeter, Exeter, UK, 5Kings College, London, UK The problem: In the management of acute sore throat current UK (HPA & SIGN) guidelines recommend avoiding antibiotics. When antibiotics are indicated a 10 day course of phenoxymethylpenicillin is suggested, based on limited trial evidence from children. Evidence suggests patients frequently do not complete antibiotics and reserve left overs for future use. It is not known what current prescribing patterns are nor whether the duration of prescribed antibiotic influences symptomatic outcomes or re-consultation. The approach: We report data from a prospective cohort study of sore throat management in primary care. Participants were aged 16 or more with acute uncomplicated illness (≤14 days), who presented with sore throat as the main symptom and had an abnormal examination finding of the pharynx were recruited by GPs who completed a simple one page proforma collecting baseline clinical symptoms and signs. A structured notes review was completed after one month to document re-consultation and suppurative complications. One in ten participants completed a validated symptom diary. Findings: Overall 14 610 adolescents and adults were recruited between 10 November 2006 and 1 June 2009 from 616 practices and diary data is available in 1629. Immediate antibiotics were issued for 6088(42%) and delayed antibiotics for 1784(12%). The dominant antibiotic used was phenoxymethylpenicillin (5700/7557) and the majority of prescriptions were for 5 (20%) 7 (57%) or 10 (22%) days. Those receiving antibiotics experienced on average 1 day less symptoms than those not treated ( None: median 4 days IQR 2 to 7 days; Immediate: median 3 days IQR 2 to 5 days; Delayed : median 3 days IQR 2 to 6 days.) From the diary data of those reporting taking the antibiotics; those prescribed a course of 10 days reported taking antibiotics for 9.33 days (s.d.1.73); those prescribed a course of 7 days reported taking antibiotics for 7.01 days (s.d. 2.14) and those prescribed a course of 5 days reported taking them for 5.75 days s.d. (2.04). Controlling for baseline severity symptomatic outcomes were unrelated to duration of prescription issued (severity day 2-4, duration of moderately bad symptoms). Re-consultation in the first month was unrelated to duration of prescribed medication. Compared to those given a prescription for 10 days; 5 days: OR 1.16 (95% CI 0.78, 1.70; p=0.464); 7 days: OR 1.27 (0.82, 1.96; p=0.293) Consequences: When antibiotics are prescribed penicillin is the antibiotic of choice but half of prescriptions are for 7 days and only one fifth receive ten days as per current guidelines. There is no evidence that the duration of prescribed medication influences resolution of symptoms or risk of re-consultation. When antibiotics are indicated shorter penicillin courses may be sufficient in adults. 3C.5 Can point-of-care rapid HIV testing in primary care increase early identification of HIV? The RHIVA2 cluster randomised controlled trial

ABSTRACTS OF ORAL PRESENTATIONS Werner Leber1, Heather McMullen1, Nadine Marlin1, Stephen Bremner1, Sally Kerry1, Kambiz Boomla1, Danna Millett2 ,3, Sifiso Mguni2, Sarah Creighton2 ,3, Jose Figueroa3, Richard Ashcroft1, Graham Hart4, Valerie Delpech5, Alison Brown5, Graeme Rooney5, Maria Sampson6, Adrian Martineau1, Andreia Santos7, Jane Anderson2, Chris Griffiths1 1 QMUL, London, UK, 2Homerton University Hospital NHS Foundation Trust, London, UK, 3NHS City and Hackney, London, UK, 4University College London, London, UK, 5Public Health England, London, UK, 6Barts Health NHS Trust, London, UK, 7The London School of Hygiene and Tropical Medicine, London, UK

Adrian Martineau2, Fern Terris-Prestholt1, Chris Griffts1 Global Health Department, The London School of Hygiene and Tropical Medicine, London, UK, 2Centre for Primary Care and Public Health, Queen Mary University of London, London, UK, 3Department of Sexual Health, Homerton University Hospital NHS Foundation Trust, London, UK, 4NHS City and Hackney, London, UK, 5School of Law, Queen Mary University of London, London, UK, 6Faculty of Population Health Sciences, University College London, London, UK, 7Centre for Infectious Disease Surveillance and Control, Public Health England, London, UK, 8Department of Virology, Barts Health NHS Trust, London, UK

The problem: Early detection of HIV remains a major challenge. Many people with living HIV are unaware of their infection, being at risk of ill health and premature death due to delayed diagnosis. Early treatment saves lives and prevents onward transmission of the virus.

The problem: Hackney is a high prevalent area for HIV in London, with an estimated prevalence of 7.75/1,000 for people aged 15 to 59 years in 2012. Early diagnosis and treatment can save lives, prevent transmission of the virus and save health care resources.

The approach: We completed a pragmatic cluster randomized trial in the London borough of Hackney. Forty of 45 (89%) general practices were randomised to receive a programme promoting point-of-care rapid HIV testing during the registration health check or at first consultation, or to continue standard care. Using minimisation, and maintaining allocation concealment, practices were randomly allocated to intervention or control groups using the following criteria: practice list size, indices of multiple deprivation and male HIV testing rate. Participants aged 16 and above were eligible. Patients unable to communicate in English who presented without a suitable interpreter, and known HIV positive patients, were excluded. The study period was from April 2010 to August 2012. The primary outcome measure was mean CD4 count at diagnosis. The main secondary outcome was rate of new diagnoses per practice.

The approach: We assessed the cost-effectiveness of determining new human immunodeficiency virus (HIV) diagnoses using a HIV point-of-care test. Effectiveness was measured trough a pragmatic cluster randomized controlled trial. The trial took place in the London borough of Hackney invited 45 surgeries to take part in to the study. Forty of 45 (89%) general practices were randomised to receive a programme promoting point-ofcare rapid HIV testing during the registration health check or at first consultation, or to continue standard care. The study period was from April 2010 to August 2012. The primary outcome measure was mean CD4 count at diagnosis. We compared two strategies for HIV screening: 1) current practices of detection (non-routine, provider, or patient-initiated testing and presentation with opportunistic infections); versus 2) routine annual screening. Cost of routine care and screening intervention were estimated. The analysis was undertaken from the perspective of the NHS and thus only considered costs directly incurred by the NHS. Future costs were discounted at 3%. We used the Cost-Effectiveness of Preventing AIDS Complications (CEPAC) model, a widely-published microsimulation of HIV disease, to forecast mean costs, life expectancy, and quality-adjusted life years (QALYs) in both HIV-infected and HIV-uninfected individuals under each strategy. QALYs were discounted at 1.5% since for this patient group they potentially arise over a long time horizon. The Transmission Module uses deterministic methods to link the effects of expanding HIV screening and earlier treatment initiation to both community HIV viral load levels and the subsequent number and timing of secondary HIV infections. Findings: The mean cost of a point-of-care was estimated in £10.24. Under the current practice scenario, the discounted quality adjusted life years was estimated in £32,103 per QALY while the screening scenario was estimated in £17,392 per QALY. In a series of one-way sensitivity analyses, we varied each of the key parameters across plausible ranges, while holding all other input values at baseline. The four parameters that caused the greatest change in the ICER for the screening intervention were: the number of new diagnosis, linkage to care, the costs of ART, and HIV screening costs, but the ICER for the screening intervention remained below NICE threshold of £30,000 per QALY.

Findings: During the study 44,971 adults registered with 20 intervention practices, and 38,464 with 20 control. Mean age was 35 years, and 45% were men in each group. Ethnicity was 49% white and 17% black (intervention) versus 57% and 15% (control). Intervention practices offered 11,180 rapid tests, with 4,978 (44.5%) accepted (14 tests reactive, 11 confirmed HIV positive). Opportunistic serology testing identified 21 (intervention) and 14 (control) further cases. Of the 32 new diagnoses in the intervention group, 19 (59%) were men, 20 (63%) were black African, and six (19%) were men who have sex with men (MSM). Mean (SD) CD4 count at diagnosis was 356 cells/μL (254) in intervention versus 270 (257) in control; adjusted difference in square root CD4 cell count was 3.1, 95% confidence interval, -1.2 to 7.4; p=0.16. It was estimated the percentage of cases diagnosed below 350 cells/μL fell by 18% (-4% to 40%) from 73% to 55%. In a pre-planned sensitivity analysis excluding seven patients diagnosed by routine antenatal screening, this difference was 6.4, 1.2 to 11.6, p=0.017. Rate of diagnosis per 10,000 patients per year increased from 0.07 in control to 0.30 in intervention (adjusted ratio of geometric means: 4.51, 1.27 to 16.05, p=0.021). All people diagnosed via rapid testing were successfully transferred into specialist care. Consequences: Our program offering point-of-care HIV testing at general practice registration led to increased and earlier detection of HIV, and identified patients from non-MSM risk groups. To reduce undiagnosed and late presentation of disease, we recommend large-scale HIV testing be implemented in UK general practices located in high prevalence areas.

Consequences: Routine HIV screening is a cost-effective intervention for high endemic areas as Hackney; the intervention is justified on both clinical and cost-effectiveness grounds. Screening focusing on high endemic areas reduces undiagnosed and late presentation of disease, and save health care resources for the NHS.

3C.6 3D.1 Point-of-care rapid HIV testing in primary care is a cost-effective intervention in high endemic areas: The results of RHIVA2 cluster randomised controlled trial Andreia C Santos1, Werner Leber2, Heather McMullen2, Nadine Marlin2, Stephen Bremner2, Sally Kerry2, Kambiz Boomla2, Danna Millett3 ,4, Sifiso Mguni3, Sarah Creighton3 ,4, Jose Figueroa4, Richard Ashcroft5, Graham Hart6, Valerie Delpech7, Alison Brown7, Graeme Rooney7, Maria Sampson8,

The Canadian Primary Care Sentinel Surveillance Network - How Canada is Leveraging Electronic Medical Record (EMR) Data for National Surveillance and Research Rachael Morkem, David Barber Queen’s University, Kingston, Ontario, Canada

ABSTRACTS OF ORAL PRESENTATIONS The problem: Using electronic medical records (EMR) as a source of data for primary care research and surveillance has fast become a keystone in academic primary care. However, amalgamating EMR data in a healthcare system that is made up of numerous silos is a vast undertaking. The Canadian Primary Care Sentinel Surveillance Network has built a robust and secure national infrastructure that extracts and uses EMR data to provide high-quality health information about Canadian primary care. The approach: Describe the formation of CPCSSN and the infrastructure that has enabled primary care data to be pulled from 10 different EMR products in 8 Canadian provinces. Key hurdles and barriers that were overcome in order to facilitate the formation of this world-class network of networks will be discussed. Issues such as privacy, data sharing, governance and inter agency collaborations will be explored to showcase how Canada is becoming a model in primary care EMR surveillance and research. Findings: CPCSSN is a network of practice based research networks (PBRN) that was formed as an entity of the College of Family Physicians of Canada (CFPC) in partnership with the Public Health Agency of Canada and the Canadian Institute of Health Information. This pan-Canadian network was built upon 7 existing academic primary care networks and collects de-identified patient health information from associated family practitioners that use EMRs. CPCSSN has since expanded and is now collecting data from 578 primary care providers, or Sentinels, who serve over half a million Canadians. Data is extracted every three months and stored in a CPCSSN database in a highly secure facility. As CPCSSN is funded to provide surveillance on chronic disease and neurologic conditions, case definitions were developed and validated for: diabetes, chronic obstructive pulmonary disease, hypertension, osteoarthritis, depression, Alzheimer disease and related dementias, epilepsy, and Parkinson disease. Consequences: Not only is CPCSSN providing surveillance for chronic conditions, CPCSSN is also offering unique feedback to contributing Sentinels on their practice. In just five short years CPCSSN is set to become the premier resource for primary care research in Canada. 3D.2 Patterns of statin prescribing for the primary prevention of cardiovascular disease in people with severe mental illness Ruth Blackburn, David Osborn, Irene Petersen, Kate Walters, Irwin Nazareth University College London, London, UK The problem: Severe mental illness (SMI) such as bipolar disorder and schizophrenia is associated with excess cardiovascular disease (CVD) morbidity and mortality, which has significant financial and social cost implications. This healthcare need is recognised within UK primary care and, since the introduction of GP incentives in 2004, rates of CVD screening have increased amongst individuals with SMI. However, the impact on statin prescribing is unknown. The approach: We used data from The Health Improvement Network (THIN) primary care database from 2007 to 2012 to describe differences in new statin prescriptions in UK patients with and without SMI with no pre-existing CVD conditions. Individuals aged 30-99 years with a diagnosis of schizophrenia or bipolar disorder were selected and age and gender frequency-matched to ten control patients in the same practice who did not have an SMI diagnosis. Individuals were excluded if they had pre-existing CVD or a previous statin prescription. Prescribing differences between individuals with SMI and controls were investigated in Poisson regression models with new statin prescription as the outcome. Models were adjusted for calendar time or estimated CVD risk. Findings: A total of 8,530 individuals with schizophrenia and 7,561 with bipolar disorder were included in the study: 85,088 and 74,446 frequency60

matched controls without SMI were also selected for comparison with individuals with schizophrenia or bipolar disorder (respectively). Time-adjusted incidence rate ratios (IRR) for statin prescribing amongst 30-59 year olds were; 1.92 (95% CI: 1.74-2.11) for schizophrenia and 1.98 (95% CI: 1.78-2.20) for bipolar disorder compared to controls without SMI. Individuals aged 60-99 years with schizophrenia had lower rates of statin prescribing than controls; IRR 0.75 (95% CI: 0.67-0.84), but there were no such difference for bipolar disorder; IRR 0.99 (95% CI: 0.88-1.11). After adjusting for estimated CVD risk, individuals with schizophrenia (but not bipolar disorder) were less frequently prescribed statins than controls; IRR 0.93 (95% CI: 0.87-0.98). Consequences: The introduction of GP incentives for CVD screening has successfully increased statin prescribing amongst younger individuals with SMI. However, evidence of disparity in statin prescribing is apparent for older individuals with schizophrenia (relative to controls without SMI) may suggest an unmet need for statin prescribing in this group. Further work is required to ascertain the impact of CVD interventions in people with SMI, and how the uptake of effective measures can be improved. 3D.3 How do caregivers in England describe their health-related quality of life and experiences in primary care? Evidence from 195,364 carers responding to the 2012 General Practice Patient Survey. Gwilym Thomas, Catherine Saunders, Martin Roland, Charlotte Paddison Cambridge Centre for Health Services Research, University of Cambridge, Cambridge, UK The problem: Informal carers comprise over 10% of the UK population, and perform important social and economic roles, with an estimated monetary value to society of ÂŁ119billion/year. Carers have been shown to have poorer health than non-carers, although some studies suggest that small amounts of caring may have health benefits for the carer. Previous research examining carersâ&#x20AC;&#x2122; experiences of primary care has focused on their caregiver role: much less is known about the primary care experiences that carers report for themselves as patients. This primary care-based study aims to explore the health-related quality of life and patient experience of informal carers in England. The approach: Responses from the 2011-12 General Practice Patient Survey were analysed using mixed effect linear regressions to describe health-related quality of life (measured using EQ-5D) of 195,364 informal carers. Logistic regressions were used to describe the primary care experience (seven questions covering: access, continuity, communication and overall experience) of carers, compared with non-carers. Analyses included a random effect for practice and fixed effects to control for age, gender, ethnicity, deprivation and, for experience only, EQ-5D. Additional analyses explored differences by age and EQ-5D domain (mobility, self-care, usual activities, pain and anxiety/depression). Findings: Carers reported poorer health-related quality of life than non-carers of similar age, gender, ethnicity and level of social deprivation. Carers of 50+ hours a week reported health 0.05 (95% CI 0.05 to 0.04) points worse on the standardised EQ-5D scale of -0.59 to 1. On average, carers reported less positive patient experiences in primary care than non-carers across all domains (OR range 0.84-0.97). Carers with higher time commitments reported poorer health than those with lower commitments, consistent across age groups and EQ-5D domain. There was heterogeneity in the association between caregiving and health by age (interaction p-value<0.0001). Compared to non-carers, younger carers reported poorer health for all time commitments, whereas those at older ages (particularly 85+) reported better health. Carers of 50+ hours/week reported more pain, OR=1.53 (1.50-1.57), p<0.0001, and

ABSTRACTS OF ORAL PRESENTATIONS anxiety/depression, OR=1.69 (1.66-1.73), p<0.0001, than non-carers, but older carers reported better mobility and better ability to carry out their daily activities. Consequences: Carers experience a double disadvantage of poorer health and poorer patient experience in primary care. We find no evidence for health benefits of caring. Older carers report better physical health than non-carers; this could be the result of being in sufficiently good physical health to be able to undertake a caring role, rather than any positive health benefit from caring itself. As practical steps we recommend practices focus on identifying and managing pain, anxiety and depression among those with high caregiving commitments, monitor the overall health of younger caregivers and act to improve the primary care experiences of patients who are carers. 3D.4 Does faster nicotine metabolism affect chances of stopping smoking in pregnancy? Evidence from the SNAP trial cohort Luis Vaz1, Tim Coleman1, Paul Aveyard2, Sue Cooper1, Jo Leonardi-Bee1 University of Nottingham, Nottingham, UK, 2University of Oxford, Oxford, UK

The problem: Outside of pregnancy, nicotine replacement therapy (NRT) helps smokers quit, but there is no evidence that standard dose NRT is effective in pregnancy. As the rate of nicotine metabolism increases in pregnancy, this could reduce NRT efficacy. In a sample of pregnant women who took part in a RCT investigating the efficacy of NRT in pregnancy, we investigated relationships between women’s rates of nicotine metabolism and smoking cessation. The approach: Women’s rates of nicotine metabolism were measured by estimating the trans-3’-hydroxycotinine: cotinine ratio (nicotine metabolite ratio, NMR) in saliva samples donated at trial entry. We used linear and logistic regression to investigate relationships between NMR and selected baseline variables and validated smoking cessation at one month after study enrolment and at delivery. We also tested for an interaction between NMR and treatment allocation (i.e. nicotine versus placebo patches). Findings: The NMR was almost normally distributed. The number of cigarettes smoked prior to pregnancy was positively associated with NMR (ß: 0.003, 95%CI: 0.001-0.004, p=0.006). At both one month (OR: 0.87, 95%CI: 0.76-0.99, p=0.043) and at delivery (OR: 0.79, 95%CI: 0.66-0.95, p=0.010) the NMR was inversely associated with validated cessation. There was no evidence for an interaction between NMR and allocation to NRT at either one month (p=0.556) or at delivery (p=0.065). Consequences: This study provides the first evidence that, in a pregnant population of smokers attempting to quit, those with faster nicotine metabolism had reduced chances of stopping smoking. In pregnancy, for NRT to have efficacy, higher than standard doses of NRT may be required to overcome increases in the rate of nicotine metabolism which are known to occur in the second trimester. 3D.5 Exposure to oral hormonal contraceptives and risk of venous thromboembolism: a nested case-control study using the CPRD database. Yana Vinogradova, Carol Coupland, Julia Hippisley-Cox University of Nottingham, Nottingham, UK The problem: Many studies have found an increased risk of venous thromboembolism (VTE) associated with the use of oral combined hormonal contraceptives, but a variety of methodologies were used in study designs. The first methodological issue relates to definition of VTE

cases - either including or excluding cases without records of anticoagulant therapy. A second issue relates to selection of patients with or without co-morbidities and conditions associated with increased VTE risk. Further, all earlier studies were based on data up to 2009 and use of oral contraceptives has continued to change. This study has focused on common oral hormonal contraceptives, including rarer compositions with drospirenone and cyproterone, and has used sensitivity analyses to compare findings with previous studies. The approach: Two nested case-control studies based on the general population using records from UK general practices within Clinical Practice Research Datalink and QResearch. Cases were female patients aged 15 to 49 with clinical records of primary VTE diagnosed between 2001 and 2013, matched by age, year of birth and practice to 5 female controls alive and registered with the practice at the time of case diagnosis (index date). Pregnant or sterilised women were excluded. Exposure to different hormonal contraceptives was based on prescriptions in the year before the index date. Effects of duration and the length of any gap since last use were investigated. Conditional logistic regression was applied to calculate odds ratios adjusted for smoking, ethnicity, body-mass index, co-morbidities associated with increased risk of VTE, and use of non-oral and progestogen-only contraceptives. The results from both studies will be combined after completing the analysis of the QResearch study. Findings: 5068 cases and 19669 controls were included to the CPRD study. The last use of any oral combined contraceptive within 28 days before the index date (current use) was associated with a significantly increased risk of VTE - odds ratio of 2.90 (95% confidence interval 2.64 to 3.18) compared with no use in the last year. The risk varied between different combinations containing: norethistiron 2.34 (1.82 to 3.01), levonorgestrel 2.29 (2.03 to 2.59), desogestrel 3.30 (2.47 to 4.42), norgestimate 2.10 (1.69 to 2.62), gestodene 4.54 (3.64 to 5.65), drospirenone 4.32 (3.38 to 5.52) and cyproterone 4.33 (3.39 to 5.53). Risks associated with use of cyproteron, drospirenone, desogestrel and gestodene were statistically significantly higher than those of norethisteron, levonorgestrel and norgestimate. Although the risk associated with past use (more than 28 days before the index date) for any oral combined contraceptive was slightly increased (1.31, 1.10 to 1.57), risks for the individual contraceptives were not statistically significant. Consequences: Combined oral contraceptives remain a risk factor for VTE in women. This study provides more information, particularly on newer drugs which may help prescription practice. 3D.6 Screening for atrial fibrillation in general practice: A national, crosssectional study of an innovative technology. Gerard Bury1, Davina Swan1, Walter Cullen2, David Keane1, Helen Tobin1, Mairead Egan1, David Fitzmaurice3, Crea Carberry1, Cecily Kelleher1 1 University College Dublin, Dublin, Ireland, 2University of Limerick, Limerick, Ireland, 3University of Birmingham, Birmingham, UK The problem: Atrial fibrillation (AF) is responsible for up to a quarter of all strokes and is often asymptomatic until a stroke occurs. 12 lead ECG is the screening method of choice but is demanding of time and resources and has not been accepted as the basis for screening. This project tested the use of three lead monitoring as a screening tool for AF in general practice. The approach: 26 randomly selected general practices identified 80 randomly selected patients aged 70 or older from their databases and excluded those known to have AF or who not attended for three years. Patients were invited to attend for screening and a two minute three-lead ECG was recorded and collected centrally for expert cardiology assessment. Risk factor data was gathered by chart review and confirmed at the screening visit. 61

ABSTRACTS OF ORAL PRESENTATIONS Outcomes: (i) point prevalence of atrial fibrillation (ii) proportion of ECG tracings which were adequate for interpretation, (iii) uptake rate by patients (iv) acceptability of the screening process to patients and staff (reported separately). Findings: Of 1447 current patients, 1003 were eligible for inclusion, 639 (64%) agreed to take part in screening and 566 (56%) completed screening. The point prevalence rate for AF was 10.3% made up of 2.1% new cases (12 of 566 who were screened) and 9.3% existing cases (134 of 1447 eligible patients). Only four of 570 (0.7%) screening visits did not record a usable ECG and 11 (2.6%) three lead ECGs required a clarifying 12 lead ECG.

to estimate mortality. Weaknesses include we cannot rule out benefit from other unidentified parallel interventions, such as change in dietary behaviour- although if present these are desirable effects, or the effects of taking part in a clinical trial. The size of this perceived placebo benefit may well be smaller in usual practice outside of a clinical trial. 3E.2 Patients anticipated actions in the event of symptoms of a transient ischaemic attack: a qualitative study. Parker Magin1, Terry Joyce1, Christopher Levi1, Dan Lasserson2 University of Newcastle, Newcastle, NSW, Australia, 2Univesity of Oxford, Oxford, UK 1

Consequences: Three lead screening for AF is feasible and effective and offers an alternative to pulse taking or 12 lead ECGs. The availability of this technology may facilitate more effective screening, leading to reduced stroke incidence. 3E.1 Meta-analysis of the estimated placebo effect in primary prevention of cardiovascular mortality Kamalpreet Cheema, Antonia Hardcastle, Nicholas Steel, Robert Fleetcroft University of East Anglia, Norwich, UK The problem: For some conditions the placebo effect of the doctor-patient relationship may be the most powerful intervention in the medical bag.(1) The perceived placebo effect is a function of the true placebo effect, the natural course of the illness, regression to the mean, other time effects and unidentified parallel interventions. (2)This is additional to any treatment effect gained from the pharmacological effectiveness of a drug. We aimed to measure and compare the size of the placebo effect with the size of the direct pharmacological benefit from statins, in the primary prevention of cardiovascular disease (CVD). The approach: We conducted a systematic literature review of the literature for randomised controlled trials using statin therapy for the primary prevention of cardiovascular mortality. We used the control arms of statins in primary prevention trials to estimate the placebo effect, by comparing actual CVD mortality in the trials and predicted CVD mortality calculated using the SCORE risk assessment algorithm. SCORE is the European cardiovascular disease risk assessment model, validated for different populations. Findings: We identified six primary prevention trials meeting the inclusion criteria with a total of 10,471 patients. The relative risk of actual CVD mortality experienced in the placebo arm compared to predicted mortality using SCORE was 0.80 (95% CI 0.66 - 0.97). The relative risk of CVD mortality in the statin arm compared to the placebo arm (the treatment effect of statins) was 0.79 (95% CI 0.64 - 0.97). The total relative risk of CVD mortality in the statin arm compared to predicted SCORE mortality was 0.62 (0.80 multiplied by 0.79). Consequences: Placebo effects produce statistically and clinically significant improvement on the reduction in CVD mortality, which is both in addition and of similar proportion to the pharmacological benefits gained by taking a statin. The implication is that equal importance should be attached to both the intervention (prescribing a statin) and the setting in which the process of care takes place. In this example the placebo responses were observed in clinical settings, which might not be transferable to (for example) purchasing the drug ‘over the counter’. There is however a dearth of research into the placebo response, and more is needed. The strengths of this study are that estimate error due to the natural course of disease and regression to the mean are unlikely as we used mean values 62

The problem: Transient Ischaemic Attack (TIA) requires urgent investigation and management to prevent stroke, but non-presentation or delays in patient presentation to health services have been found to compromise timely management. Previous studies have addressed the behaviour of patients who have presented with TIAs. But the problematic (and common) scenarios of patients failing to present following symptoms, or only presenting following subsequent stroke, have not been adequately studied. In this study we aimed to address this research gap by exploring general practice patients’ anticipated responses to TIA symptoms. Thus, the anticipated responses, rationale and context of those who anticipate not acting upon TIA symptoms (as well as those of patients who do access health care) would be explored. The approach: We conducted a qualitative study employing semistructured telephone interviews. Participants were recruited from respondents in an earlier quantitative study based in Australian general practices. Maximum variation purposive sampling of patients (on the basis of age, rurality, gender and previous experience of stroke/TIA) continued until thematic saturation was achieved. After initial interviews explored knowledge of TIA and patients’ potential responses, subsequent interviews further explored anticipated responses via TIA and non-TIA clinical scenarios. Transcribed interviews were coded independently by two researchers. Data collection and analysis were concurrent and cumulative, using a process of iterative thematic analysis and constant comparison. A schema explaining participants’ anticipated actions emerged from analysis of the data during this process and was iteratively tested in later interviews. Findings: Thirty-seven interviews were conducted and a ‘spectrum of action’ was established. The spectrum was from watchful waiting (only responding if symptoms recurred), through accessing GP or Emergency Department care with increasing levels of urgency, to summoning an ambulance immediately. The substrate for decision-making along this spectrum operated via three constructs: the ‘individual set’ of the participant (their inherent disposition towards action in response to health matters in general), their ‘discriminatory power’ (the ability to discriminate TIA symptoms from non-TIA symptoms) and their ‘effective access’ to health-care services. The individual set (ranging from ‘can’t be too careful’ to ‘can’t jump at everything’) was influenced by age, prior health experiences and individual personality. Discriminatory power was based on stroke knowledge, and the clinical phenotype and temporal pattern of the event. Effective access depended upon knowledge of services and a number of determinants of services availability. Consequences: Policies to improve patients’ accessing care (and accessing care urgently) post-TIA should address these three determinants of anticipated action. Influencing an ‘individual set’ is inherently unlikely. ‘Discriminatory power’ is the target of stroke public awareness media campaigns but their extension to TIA may be problematic. A targeted approach may be appropriate. Our findings suggest facilitating better GP practice triage of potential TIA may be appropriate.

ABSTRACTS OF ORAL PRESENTATIONS 3E.3

3E.4

The REACH Heart Failure Manual: Development of a complex, home-based intervention to support self care for people with heart failure.

Treating uncomplicated stage 1 hypertension in primary care: an expensive mistake?

Colin Greaves1, Carolyn Deighan2, Jenny Wingham3, Patrick Doherty4, Wendy Armitage2, Michelle Clark2, Jenny Elliott2, Heather Cursiter2, Jackie Austin5, Kevin Paul6, Louise Taylor2, Rod Taylor1, Hayes Dalal3 1 University of Exeter Medical School, Exeter, UK, 2NHS Lothian Heart Manual Department, Edinburgh, UK, 3Royal Cornwall Hospitals Trust, Truro, UK, 4University of York, York, UK, 5Aneurin Bevan Health Board, Abergavenny, UK, 6Service user representative, Truro, UK The problem: Heart failure (HF) affects almost a million people in the UK. Cardiac rehabilitation is effective and recommended by NICE guidelines. However, fewer than 5% of people with HF currently access rehabilitation programmes in the UK. A range of self-care activities such as improving fitness, managing medications, appropriate help-seeking and managing stress can affect symptoms and well-being. A home-based rehabilitation and self-care support package therefore offers the opportunity to increase access to care and to substantially increase health-related quality of life in people with HF. Improved self-management of HF may also reduce hospital readmissions. The approach: We used formal Intervention Mapping methods to develop a complex exercise training and self-care support intervention for people with HF. This included 1) Needs Assessment: Interviews and a ‘scaffolding questionnaire’ survey with patients, service providers and caregivers; systematic reviews of quantitative and qualitative evidence; reviewing guidelines; consultation with experts in cardiology, cardiac rehabilitation, self-help manual development, cardiac nursing, health psychology, primary care and a 9-member patient participation group. Data from these disparate sources were synthesised using framework synthesis (Oliver et al., 2008). 2) Construction of ‘mapping matrices’: We constructed tables of intervention objectives and modifiable determinants of these objectives 3) Intervention techniques and strategies were selected to modify the determinants identified, based on existing intervention theory and taxonomies of behaviour change techniques as well as expert opinion and experience. Findings: The above processes generated three main resources: A ‘Heart Failure Manual’ for use by patients, a Caregiver Resource and a training programme for intervention facilitators. Facilitation consists of 4-6 contacts (at least two face-to-face) with a cardiac nurse or physiotherapist. The intervention has four primary self-care objectives (maximise physical fitness, manage stress /anxiety, manage and take medications, manage fluid balance (and seek help appropriately when this changes)), as well as 19 secondary objectives (e.g. manage low mood, get vaccinated, improve sleep quality). The intervention is tailored to address both patient priorities and clinical priorities on an individual basis. It aims to increase both understanding of the condition and mastery of self-care activities through practice, self-monitoring of progress and facilitated problem-solving. It also aims to strongly support caregiver involvement in self-care. Our service user group participated strongly in developing the intervention objectives, reviewing text, providing ‘Patient Voices’ that were used throughout the manual and helping to deliver the facilitator training. Consequences: We have rigorously developed a comprehensive self-care support intervention for people with heart failure and their caregivers. This is grounded in the evidence base, behaviour change theory, clinical guidelines, wide-ranging expertise and patient perspectives. The intervention has potential to improve health-related quality of life for people with HF. A NIHR-funded feasibility study is in progress and a multi-site RCT will follow.

James Sheppard1, Kate Fletcher2, Richard McManus1, Jonathan Mant3 University of Oxford, Oxford, UK, 2University of Birmingham, Birmingham, UK, 3University of Cambridge, Cambridge, UK

The problem: Hypertension is a key risk factor for the development of cardiovascular disease. Classifications of hypertension are by definition arbitrary, and treatment of uncomplicated (i.e. low cardiovascular disease risk) stage 1 hypertension (i.e. sustained blood pressure between 140/90-159/99mmHg) has long been considered controversial due to the paucity of evidence to support prescription of therapy. Despite this, even the most up to date guidelines continue to recommend treatment and in the UK, pay-for-performance targets for general practitioners were recently revised to encourage blood pressure control to a target of 140/90 mmHg, regardless of underlying cardiovascular disease risk. This study set out to estimate the prevalence of this condition in an untreated population and establish the costs of initiating treatment in such patients. The approach: This was a cross-sectional study of anonymised records from patients aged 40-74 years in 19 general practices in the West Midlands, UK. Data relating to patient demographics, existing cardiovascular disease and risk factors (blood pressure, total cholesterol) were extracted from patient records. Patients with a blood pressure between 140/90-159/99mmHg, no cardiovascular disease and <20% 10-year cardiovascular risk were classified as having uncomplicated stage 1 hypertension. The prevalence of untreated, uncomplicated stage 1 hypertension was estimated using descriptive statistics and extrapolated using national data. The cost of achieving blood pressure control in this population was examined in a cost-impact analysis using published costs. Findings: Of the 34,975 patients in this study, untreated, uncomplicated stage 1 hypertension was present in 2,867 individuals (8.2%). This is equivalent to 1,892,519 patients in England and Wales for whom the additional cost of controlling blood pressure according to international guidelines was estimated to be between £106 million and £229 million per annum depending on the healthcare professional delivering care. Consequences: Untreated, uncomplicated stage 1 hypertension is relatively common, affecting 1 in 12 middle aged patients in Primary Care, but treatment is expensive with little prospect of benefit either at an individual or population level. Guidelines should be cautious in encouraging treatment in this group without robust trial evidence. 3E.5 A multicentre ethnographic study of unplanned hospital admission for heart failure. Rosemary Simmonds1, Sarah McLachlan3, Margaret Glogowska2, Rachel Johnson1, Helen Cramer1, Tom Sanders3, Umesh Kadam3, Daniel Lasserson2, Sarah Purdy1 1 University of Bristol, Bristol, UK, 2University of Oxford, Oxford, UK, 3 University of Keele, Keele, UK The problem: Unplanned hospital admissions are a problem for the NHS, they are expensive and can be distressing for patients. In England heart failure (HF) is a common cause of unplanned admissions. Despite existing interventions to prevent admissions these have not made a significant impact. The aims of this study are to understand in more detail pathways that lead to unplanned admissions for people with HF and to identify opportunities to reduce or prevent future admissions.

ABSTRACTS OF ORAL PRESENTATIONS The approach: We adopted a multi-centre, longitudinal, qualitative design using an ethnographic approach including: non-participant observation, in-depth interviews with patients, carers and health professionals, patient/ carer diaries and documentary analysis. The study was conducted in three contrasting geographical locations in England. Our sample included 31 patients with severe or difficult to manage HF, 9 carers and 24 health professionals. Patients were observed for up to 11months on their individual journeys through healthcare. Verification of findings was achieved through regular research and PPI meetings where interpretations of data were discussed in depth. Findings: Findings indicated local variations in how care for people with HF was organised and delivered. Some localities provided services that were more conducive to admission avoidance. In one locality care for HF was well thought out and co-ordinated across service providers. Patients were confident about seeking help on exacerbation of symptoms and had good access to specialist HF teams across primary and secondary care. In contrast, in another locality patients experienced fragmented and chaotic services with little coherence and inequitable access to specialist services, such as HF specialist nurses. Levels of HF awareness were very variable amongst patients and some GPs. Patients and carers wanted more detailed lifestyle advice and guidance on self-care, especially on diagnosis and discharge from hospital. Clinically, symptoms of HF were often misdiagnosed or ‘backgrounded’ leading to unplanned admissions. Improvements in the early detection of HF and diagnostic procedures were recommended together with a more holistic approach to care across health conditions and specialties. Disease focused approaches in primary care monitoring was not helpful for patients with HF and co-morbid conditions such as diabetes. A number of patients experienced emotional distress precipitating unplanned admissions. Both patients and clinicians recommended improved emotional support and understanding for patients with HF to avoid anxiety related admissions. Support from community matrons and HF specialist nurses helped patients to stay out of hospital by addressing their overall needs as people living with chronic conditions. Consequences: Findings suggest that improvements in diagnostic procedures, disclosure and ‘patient activation’ should help patients and clinicians in managing and treating HF. Whilst some patients will need to be admitted to hospital unexpectedly, our findings suggest there is scope to reduce those admissions precipitated by misdiagnoses, patient anxiety and lack of knowledge. 3E.6 Atrial Fibrillation: is screening effective in identifying patients at risk of stroke? Deborah McCahon1, David Fitzmaurice1, Jennifer Baker1, Ellen Murray1, Sue Jowett1, Hardeep Sandhar1, Roger Holder1, Richard Hobbs2 1 University of Birmingham, Birmingham, UK, 2University of Oxford, Oxford, UK The problem: Atrial fibrillation (AF) is an important independent risk factor for stroke and oral anticoagulation therapy provides a highly effective treatment to reduce this risk. Active screening strategies improve detection of AF in comparison with routine care, however, whether screen detected patients have stroke risk profiles favouring anticoagulation is unclear. Objective: To determine whether active screening is effective in identifying patients at risk of stroke. The approach: Data from the screening for AF in the elderly (SAFE) study, a 64

large scale multi-centred, cluster randomised controlled trial involving 50 UK primary care centres, randomised to either screening (intervention) or no screening (control) were analysed. Secondary analysis of data derived from the 25 SAFE screening practices was undertaken. Cohorts of 200 patients within the 25 screening practices were randomly selected and allocated to either opportunistic (pulse and ECG) or systematic screening (postal invitation for ECG). Stroke risk assessment was undertaken using baseline data extracted from medical records and CHADS2 criteria. CHADS2 scores were compared between the screening groups. Findings: 149 new cases of AF were detected via screening, 75 via opportunistic screening and 74 via systematic screening. CHADS2 scores were ≥1 in 83% (95% CI; 72.6-89.6) of patients detected via opportunistic screening and 78% (95% CI; 67.7-86.2) detected via systematic screening. There were no significant differences in stroke risk profiles of patients detected via opportunistic and systematic screening. Consequences: These findings suggest that active screening (either opportunistic or systematic) for AF in patients aged ≥65 years in primary care is effective for identifying patients at increased risk of stroke, with 78-83% of patients identified eligible for anticoagulation treatment. 4A.1 Community participation in primary care - an analysis of levers and barriers to implementation using Normalisation Process Theory Edel Tierney1, Rachel Evoy2 ,1, Anne MacFarlane1 Graduate Entry Medical School, University of Limerick, Limerick, Ireland, 2 Health Service Executive, Mayo, Ireland

The problem: International health policies recognise that communities should be involved in shaping primary care. However researchers and practitioners continue to debate definition, models, approaches and operational challenges of community participation in primary care. The enactment of this policy ideal has proved problematic. Using Normalisation Process Theory (NPT) we explored levers and barriers to the implementation of community participation in primary care as a routine way of working The approach: We conducted two complementary qualitative studies in Ireland based on an instrumental case of the Joint Community Participation in Primary Care Initiative - an initiative funded by the Health Services Executive and the Combat Poverty Agency. Based on principles of participatory research approach, we utilised documentary evidence (study 1), semi-structured interviews (study 1 & 2) and focus groups. We developed a purposeful sample (N=72) comprised of community representatives (n=43), health professionals (n=3), HSE personnel (n-25) and project evaluator (n=1). All data were analysed using NPT constructs as the basis for thematic analysis by an inter-disciplinary and inter-agency research team. Findings were validated through data analysis clinics and feedback sessions with participants. Findings: For many participants community participation in Irish primary care is regarded as a new and valuable way of working. Participants reported a need for all stakeholders to have a clear, shared understanding of the aims, objectives and benefits of this way of working. However, many reported that this was not their experience when working ‘on the ground’ in inter-agency groups. Drivers or champions were seen as integral to initiating the implementation process as were strong working partnerships between community and health service personnel. Some participants identified the specific benefits of established relationships in initiating and sustaining this type of work. There were many examples of successful community projects and health programmes enacted as a result of the Joint Initiative. Participants emphasised the benefits of training and networking to support these activities. Meaningful community representation on

ABSTRACTS OF ORAL PRESENTATIONS inter-disciplinary Primary Care Team proved a particular challenge with little progress reported. Participants were broadly positive about the impacts of community participation but very concerned about scope to sustain the work outwith the ‘protected’ space and resources of the Joint Initiative.

4A.3 Non-alcoholic Fatty Liver Disease: What are GPs doing about it? A review of current practice in two London boroughs ahead of implementation of a novel referral pathway.

Consequences: Findings are being used to create an interactive framework to stimulate individual and collective work among stakeholders to progress implementation of community participation in primary care in the future. Our use of NPT provided a valuable conceptual lens to synthesis data about levers and barriers to this implementation work and the use of theory means the synthesis will have relevance for other settings.

Ruth E Gailer1, Ankur Srivastava2, Alex Warner1 ,3, WMC Rosenberg2 Primary Care and Population Health Department, University College London, London, UK, UCL Institute for Liver and Digestive Health, London, UK, 3Camden Clinical Commissioning Group, Liver Group, London, UK

4A.2

The problem: Non Alcoholic Fatty Liver Disease (NAFLD) has a prevalence of 15-40%. NAFLD can be considered a feature of the metabolic syndrome, however it now accounts for around 20% of all liver transplants.

Can telephone triage systems help to improve satisfaction among patient groups who tend to report lower satisfaction? Fiona Warren1, Raff Calitri1, Emily Fletcher1, David Richards1, Tim Holt3, Val Lattimer5, Suzanne Richards1, Chris Salisbury2, Rod Taylor1, Vicky Bowyer4, Katherine Chaplin2, Jamie Murdoch5, Julia Roscoe4, Anna Varley5, John Campbell1 1 University of Exeter Medical School, Exeter, UK, 2University of Bristol, Bristol, UK, 3University of Oxford, Oxford, UK, 4Warwick University, Warwick, UK, 5University of East Anglia, Norwich, UK Previous surveys of patient satisfaction in Primary Care have indicated lower satisfaction among patients from certain demographic groups, including ethnic minorities, young adults, and patients with poorer health. The reasons for this are unclear, but alternative methods of patient management may improve satisfaction for some patient groups. The ESTEEM randomised controlled trial investigated two systems of telephone triage for management of urgent consultation requests, compared with usual patient management. Practices were randomised to one of three treatment arms: usual care (UC); GP triage (GPT); nurse triage (NT). All patients were able to communicate effectively by telephone. The approach: Patients were invited to rate their overall satisfaction with care on a Likert scale of 1 (very satisfied) to 5 (very dissatisfied). Scores were linearised on a scale of 0-100 for analysis. A series of linear hierarchical models were fitted, adjusting for socio-demographic factors (including age, gender, ethnicity dichotomised as `White’ and `Other’, presence of a long-standing health condition and ease of attending the practice during working hours) and practice-level factors. A random effect on practice was included in all models. Interactions between treatment arm and patient socio-demographic factors were investigated by including one interaction term in each model. Findings: The baseline hierarchical model (N=11 282) with no interactions indicated that young adults (16-24) were less satisfied with their care vs. patients aged 25-59 (mean difference 4.34, 95% CI 2.72; 5.97), as were patients in the `Other’ ethnic group vs. `White’ patients (mean difference 4.82, 95% 2.77; 6.88). Also, patients with a long-standing health condition were less satisfied (mean difference 1.91, 95% CI 1.06; 2.76), as were patients who were unable to attend the practice during working hours (mean difference 5.53, 95% CI 4.00; 7.06) and those who could only do so with difficulty (mean difference 2.58, 95% CI 1.41; 3.75). Interaction terms were largely non-significant, although there was some weak evidence to indicate increased satisfaction of ethnic minority patients with GP triage vs. usual care (GPT vs. UC. -4.76, 95% CI -9.74; 0.22, p-value 0.061). Consequences: The reduced satisfaction among ethnic minority patients cannot be attributed to difficulties with communicating in English. GPT appeared to be associated with higher satisfaction than UC in ethnic minority patients, but the evidence for an association was weak. Patients who found it difficult or impossible to attend the practice during working hours did not appear to find telephone triage more satisfactory than usual care.

Focus is now on stratifying patients with NAFLD through use of biomarkers and risk scores. Joint working between hepatologists and GPs is essential to identify appropriate referral to Hepatology. For such referral pathways to improve management, the current practice of GPs in this area must be assessed. How are we as GPs managing this growing group of patients in the absence of clear guidance and what patient and GP factors may account for any variation? The approach: A novel NAFLD referral pathway is being implemented between 2014-2015 in Camden & Islington, London. This covers over 400,000 patients who are registered with a GP. Those identified as ‘high risk’ will be referred to Hepatology, chiefly at the Royal Free Hospital London. There are three questions: What are the existing attitudes of GPs about NAFLD prior to pathway implementation? This is being conducted by online questionnaire to all 76 practices via the CCG. F Who have GPs referred over the 12 months prior to pathway implementation? This will be done by retrospective analysis of NAFLD referrals. F What variation is seen in referral rates between GP practices? What demographic, physician or health system related factors predict this? Statistical analysis of rates of referral and practice demographics will be performed. F

Findings: In February 2014, all GPs in Camden & Islington have been contacted to assess attitudes towards NAFLD. Preliminary findings showed that whilst 84% agree NAFLD is a significant clinical problem, GPs underestimate the prevalence of NAFLD in their patient groups (mean estimate 10.8%). Clinical management is a concern for GPs as whilst 69.5% feel confident diagnosing NAFLD, 27.3% feel confident managing it. F Between 2012-2013 there were 612 referrals to Hepatology. Preliminary findings show 21.01% of these were for NAFLD, of which 72.7% were direct GP referrals. Detailed information about the patients and referring practices are being collected for analysis. Initial findings show patients had an average age of 47.9 years, a mean BMI of 30.06kg/m2 and 60.8% had hypercholesterolaemia. F

Consequences: Questionnaire findings confirm the necessity of GP guidance and joint primary/secondary care working in this area, echoing findings in the USA. Characteristics of patients referred with NAFLD are unsurprising due to the association with the metabolic syndrome and show that GPs are referring patients who represent the typical phenotype of NAFLD. 65

ABSTRACTS OF ORAL PRESENTATIONS Detailed analysis of practice level factors which influence referral will identify the practices that can be targeted for future education around NAFLD and the use of innovative referral pathways.

Clinically-led commissioning: what ‘added value’ do GPs bring to Clinical Commissioning Groups?

The problem: Chronic pain is a common condition and frequent reason for seeking medical help. Mindfulness approaches are used increasingly for management of chronic conditions such as chronic pain. Mindfulness is self-regulated attention that can be developed through meditation practice. A systematic review carried out by the author revealed that low uptake and high attrition rates are limiting factors in RCTs of mindfulness approaches for patients with chronic pain.

Kath Checkland1, Neil Perkins1, Anna Coleman1, Stephen Peckham2 1 University of Manchester, Manchester, UK, 2University of Kent, Canterbury, UK

The approach: The aim was to explore the barriers and facilitators to engagement, of chronic pain patients, with a mindfulness programme to enable a modified mindfulness programme to be developed that addresses these.

The problem: The NHS in England recently underwent a significant and disruptive reorganisation. Amongst many other changes, Primary Care Trusts were abolished and replaced with new GP-led Clinical Commissioning Groups (CCGs). CCGs are responsible for commissioning that majority of care for patients in England, spending a budget of £65 billion. Many claims have been made about the ‘added value’ that clinicians bring to commissioning; this study explores the views of those involved and sets these in the context of the wider literature on GP-led commissioning

A computer search of practice prescribing records identified adults with chronic pain, screened by GPs against inclusion and exclusion criteria. The 480 patients on the screened list were invited to take part. Those who consented to attend a pre-programme interview were given questionnaires that included demographics, medical history and quantitative measures of pain, psychological health, quality of life, pain acceptance, mindfulness and psychological flexibility using validated measures. Participants were invited to take part in the eight week mindfulness programme, consisting of standard components, including completing daily diaries. Following programme completion, participants were sent post-programme questionnaires and were invited to attend a face-to-face interview and a three-month follow-up telephone call with the lead researcher. These were conducted and analysed using interpretative phenomenological analysis (IPA) methodology. This approach was chosen to enable the researcher to explore how the participant ascribes meaning to their experience, so that themes and psychological concepts can be identified based on understanding the deeper experience of the participant. Descriptive frequency statistics were conducted using SPSS for quantitative data. Qualitative data were stored using NVIVO software. For quality assurance, a second coder reviewed a subset of transcripts.

4A.4

The approach: The study undertook semi-structured interviews with 39 clinicians and managers in seven CCGs across England. The study took place in 2013, and forms the second phase of a three-phase longitudinal project which is following the development of CCGs. The interviews focused upon respondents perceptions of clinician ‘added value’, and explored the factors that were influencing their work. Interviews were transcribed and analysed using the Nvivo qualitative data analysis software. Findings: Claims to GP-added value in commissioning centred on two aspects of their experience. It was claimed that GPs bring ‘fine grained’ knowledge to the commissioning process. This is knowledge that is rooted in the experiences of individuals, and it was claimed that GPs are in a unique position to aggregate this individual knowledge and thereby provide an overview of the system as a whole. It was also claimed that GP knowledge is ‘concrete’, in that it is rooted in actual patient experiences. Respondents reported concerns about the sustainability of the current arrangements, whilst also highlighting the good and productive relationships they were building with managers. Consequences: A focus on ‘fine-grained’ and ‘concrete’ clinical knowledge brings with it three major concerns. Firstly, the number of GPs closely involved with CCGs is necessarily limited. This means that either they need to build excellent systems for gathering the views of their wider members, or there is a danger that the views of a small number of activists will drive commissioning decisions. Secondly, focus on front line clinical knowledge implicitly denigrates the wider population knowledge previously supplied by PCT public health (PH) experts, who are no longer centrally involved with commissioning following their move to local government. Previous incarnations of clinically-led commissioning have struggled to move beyond the concerns of a small number of activists to embrace wider population needs; our study suggests that CCGs may have similar issues. Finally, sustainability of CCGs is a significant issue, with GPs and managers alike describing a burgeoning workload. This was coupled with concerns that CCGs were failing to bring forward new clinicians to take on the roles, suggesting that succession planning for the future is urgently required. 4B.1 A Mindfulness-based programme for patients with chronic pain: Exploring the barriers and facilitators to engagement. Fathima Marikar Bawa1, Christine Bond1, Stewart Mercer2, Rachel Atherton3 1 University of Aberdeen, Aberdeen, UK, 2University of Glasgow, Glasgow, 66

UK, 3NHS Highland, Inverness, UK

Findings: Of the 480 patients who were invited, 56 agreed to attend a pre-programme interview. Following interview, 28 confirmed an interest in attending the mindfulness programme. 22 attended week one of the programme and of these, 12 attended 5 or more sessions. All 28 were invited to attend a post-programme interview. Emerging themes included internal barriers (lack of: sense of autonomy; perceived competence at taking part; sense of how the programme relates to them), external barriers (lack of support from family, time commitments), internal facilitators (gaining clarity of values and a sense of fulfilment during the course), and external facilitators (support from family and encouragement from programme facilitators). Consequences: Internal and external barriers and facilitators to engagement with mindfulness programmes were identified. These will be addressed in designing a modified mindfulness programme to improve uptake and reduce attrition rates. This can then be evaluated for use in a large scale RCT of mindfulness approaches for patients with chronic pain. 4B.2 Diagnosis and monitoring of chronic kidney disease in primary care systematic review and meta-analysis of bias of glomerular filtration rate estimating equations. Daniel Lasserson1 ,2, Jennifer Hirst1, Richard Hobbs1 ,2, Brian Willis3, Christopher O’Callaghan2 ,4 1 Nuffield Department of Primary Care Health Sciences, University of Oxford, Oxford, UK, 2NIHR Oxford Biomedical Research Centre, Oxford, UK, 3Department of Primary Care Clinical Sciences, University of Birmingham, Birmingham, UK, 4Nuffield Department of Medicine, University of Oxford, Oxford, UK

ABSTRACTS OF ORAL PRESENTATIONS The problem: Chronic Kidney Disease (CKD) increases cardiovascular risk and progression to end stage renal failure. Estimated glomerular filtration rate (eGFR) underpins the detection, staging and monitoring of CKD in primary care. In reviewing its CKD guidance NICE is considering whether to recommend changing from the current MDRD equation for eGFR reporting to the newer CKD-EPI equation which some studies suggest may be more accurate. We undertook a systematic review and meta-analysis to address this question. The approach: MEDLINE, EMBASE and the Cochrane library were searched for diagnostic accuracy studies comparing eGFRs calculated with the MDRD or CKD-EPI equation to a gold standard radionuclide, inulin or iohexol measured GFR. Included studies either recruited patients in primary care or reported metrics on a distribution of measured GFR that allowed comparison to a primary care population. Quality scores were carried out using QUADAS2. For both equations, random effects meta-analysis of the observed bias (estimated - measured GFR) and meta-regression of bias against study parameters were undertaken using the metafor package in R software (version 3.0.2). Finding: Of 5978 studies retrieved on initial search, we included 155 studies examining the MDRD equation (61,574 patients) and 64 examining the CKD-EPI equation (35,801 patients). Very few studies reported sensitivity or specificity for diagnosing CKD stages, but the majority reported mean bias of eGFR compared with measured GFR. Sufficient data for meta-analysis was presented in 84 MDRD studies (19,687 patients) and 38 CKD-EPI studies (8,280 patients). Both equations under-estimated measured GFR but MDRD had a greater weighted mean bias of -7.0 ml/ min/1.73m2 (95% CI -9.7 to - 4.3) compared with CKD-EPI mean bias of - 4.7 ml/min/1.73m2 (95% CI -8.1 to -1.3). Meta-regression of eGFR bias against mean measured GFR in each study demonstrated that the bias with MDRD worsens with increasing measured GFR i.e. at higher levels of renal function (beta coefficient = -0.15, 95%CI -0.26 to -0.05, p=0.004), but CKD-EPI bias is not affected by measured GFR (beta coefficient = -0.05, 95% CI -0.19 to 0.09). Meta-regression of eGFR bias against mean age and percent female participants in each study demonstrated that the bias in both formulae is not affected by age or sex. Consequences: CKD-EPI and MDRD both under-estimate true renal function, but this bias is smaller for CKD-EPI and unlike MDRD, the CKD-EPI bias does not worsen at the higher levels of renal function which are typically found in primary care populations. Therefore, CKD-EPI will be more accurate than MDRD in primary care - a finding of direct relevance to the current NICE CKD guidance review. Nevertheless, substantial under-estimation bias of renal function remains, highlighting the need for novel biomarkers or new estimating equations to improve CKD detection and monitoring in primary care. 4B.3 Do Patient’s Characteristics Predict Failure to Receive Care for Type 2 Diabetes? Luke Mounce1 ,2, Suzanne Richards1, Antonia Hardcastle3, William Henley2, Max Bachman3, John Campbell1, Allan Clark3, David Melzer4, Nicholas Steel3 1 Primary Care Research Group, University of Exeter Medical School, Exeter, Devon, UK, 2Health Statistics Research Group, University of Exeter Medical School, Exeter, Devon, UK, 3Norwich Medical Scool, University of East Anglia, Norwich, Norfolk, UK, 4School of Public Health and Epidemiology, University of Exeter Medical School, Exeter, Devon, UK The problem: The increasing prevalence of type 2 diabetes is a global public health crisis that poses major care and economic challenges for both developed and developing countries. Diabetes complications are

potentially avoidable, yet poor care is common. Understanding what factors drive variation in the receipt of core diabetes care is vital for informing the development of quality improvement strategies. Studies have identified provider factors predicting failure to receive care, but less is known about patient factors. Prior work has been limited to investigating basic patient descriptors that can be extracted from clinical records. We examined the proportions of patients offered indicated diabetes care in a large cohort of older people and modelled a broad range of sophisticated, baseline patient characteristics to predict subsequent failure to receive care two years later. The approach: English Longitudinal Study of Ageing participants (aged 50 years or older) in 2008/9 and 2010/11 self-reported on receipt of four components of diabetes care assessed by quality indicators (QIs) that were previously developed for the Assessing Care of Vulnerable Elders project. Annual checks for HbA1c, proteinuria and foot examination formed a care bundle. A further QI assessed diabetes plus other cardiac risk factors for ACE inhibitors/receptor blockers. Baseline (2008/9) patient predictors included: socio-demographic, lifestyle and health characteristics; diabetes self-management knowledge and health literacy; and whether care was received. Two logistic regression models were developed to predict failure to receive care in 2010/11; one for all-or-nothing achievement of the care bundle (n=902) and the other for the ACE inhibitor indicator (n=757). Findings: A third of participants (2008/9=32.8%; 2010/11=32.2%) did not receive one or more of care bundle components and nearly half of those eligible were not offered ACE inhibitors/receptor blockers (2008/9=44.6%; 2010/11=44.5%). In logistic regressions, having not received the care at baseline was the strongest predictor of not receiving care two years later for both the care bundle (OR 3.32, 95% CI 2.28 - 4.84) and the ACE inhibitor indicator (OR 6.94, 95% CI 4.54 - 10.61). Participants lacking diabetes self-management knowledge or cognitively impaired were more likely to not receive the full bundle. Individuals who were younger, single, not obese or who had low health literacy were more likely to subsequently not be offered ACE inhibitors/receptor blockers. Consequences: Non-receipt of indicated care is entrenched over time. Quality improvement strategies should target those who previously did not receive indicated care, who could be highlighted by routine clinical record audit. Increasing patients’ knowledge of the care they should expect to receive may aid quality improvement. 4B.4 Healthy behaviours reduce 10-year incidence of diabetes: a population cohort study Gráinne Long1, Ingegerd Johansson2, Margareta Norberg2, Patrik Wennberg2, Eva Fhärm2, Olov Rolandsson2, Simon Griffin1, Rebecca Simmons1, Lars Weinehall2 1 University of Cambridge, Cambridge, UK, 2Umeå University, Umeå, Sweden The problem: In several randomised trials, intensive behavioural interventions focusing on diet and physical activity halved type 2 diabetes incidence among people with impaired glucose tolerance. However, it is not feasible in general practice to screen populations using oral glucose tolerance tests or to deliver prolonged and intensive lifestyle interventions. Whether a comparable risk reduction might be realised if middle-aged individuals in the general population could be encouraged to meet similar behavioural goals to those targeted in diabetes prevention trials is unclear. The approach: Population-based prospective cohort study of 32120 Västerbotten Intervention Programme (VIP) participants aged between 35-55 years. VIP comprises a population- and individual-based health promotion intervention in Northern Sweden, incorporating an oral glucose tolerance test, clinical measures and completion of diet and physical 67

ABSTRACTS OF ORAL PRESENTATIONS activity questionnaires. We quantified the association between the achievement of six behavioural goals at baseline VIP assessment - BMI <25 kg/m2, moderate physical activity, non-smoker, fat intake <30% of total energy, fibre intake ≥15g/4,184kJ and alcohol intake ≤10g/day - and diabetes incidence using Poisson regression. Findings: Median (IQR) follow-up time was 9.9 (0.3) years; 2211 individuals (7%) developed diabetes (6.84 per 1000 person years at risk; 95% CI=6.56 to 7.13). Only 22% of VIP participants met 5 or 6 goals. Compared to individuals who achieved 6 goals, diabetes incidence was 2.81 (95% CI 1.88 to 4.21) times higher in participants meeting 0 or 1 health goals at baseline, adjusting for sex, age, calendar period, education, baseline blood pressure, family history of diabetes, history of myocardial infarction and self-reported long-term illness lasting at least 6 months. Consequences: Diabetes incidence was inversely associated with the number of behavioural goals achieved. While type 2 diabetes and its determinants demand public health action, the burden of diabetes-related morbidity and mortality would be significantly reduced if primary care teams could promote the achievement of behavioural goals in the general population and contribute to encouraging small shifts in population distributions of health behaviours. 4C.1 The use of a Simulated General Practice placement for undergraduate teaching: Student perceptions of the benefits and drawbacks. Russell Hearn King’s College London, London, UK The problem: This innovation to the King’s College London undergraduate curriculum was born out of the need for increased placement capacity for undergraduate medical students and the desire to capitalise on the published benefits of simulation. The use of simulation in medical education has multiple reported benefits to student and several aspects of the simulated environment are known to enhance student learning: opportunity for feedback; multiple learning strategies; presentation of clinical variations; a controlled environment and individualised learning. The usefulness and increasing use of simulation has been documented in the hospital setting however, much of the published literature relates to the use of simulation to teach or reinforce hospital based activities and specialities. There is little published information on the use of simulation in undergraduate primary care education. The approach: The study aims to evaluate the impact of placement at a simulated general practice on the learning of undergraduate medical students. Undergraduates gain varying experiences in their GP placements based on the nature of the local practice and tutor they are placed with. A pilot programme was designed and implemented in which students attend a ‘simulated’ general practice in place of a traditional placement. These students see a mixture of simulated patients, expert patients and patients recruited from general practice with the aims of maximising curriculum coverage and providing a standardised high quality educational experience for the learner. To analyse the impact online questionnaires were collected from the 80 students who participated. This informed a topic guide for focus groups for a subsection of 12 students. Qualitative thematic analysis was then performed utilising NVivo software based on focus group and free text data from questionnaires

Findings: Students identify several aspects of a simulated placement which are beneficial including standardised experiences, high quality feedback, reduction in interruptions and benefit of feedback from simulated patients. Thematic analysis of data reveals that students perceive the programme to be of educational benefit. Students appreciate the controlled environment; standardisation of teaching and patient presentations; and feedback received in this setting. Students also felt that some simulated patients did not represent real life presentations and that the patient group lacked clinical signs. Consequences: Simulating general practice placements is a viable option for student general practice placements with benefits to student learning that echo evidence published regarding evidence in other arenas. Sound logistics and robust patient scenario writing are essential in maximising student perceived educational impact. Simulated placements offer a method of increasing capacity and quality of undergraduate primary care education. 4C.2 Longitudinal Integrated Clerkships in the UK? Results of a four-year pilot in Exeter. Alexander Harding University of Exeter Medical School, South-West England, UK The problem: Longitudinal integrated clerkships (LICs) represent an interesting innovation in the organisation of medical student placements. Students are placed in one location for up to 12 months and follow a ‘panel’ of patients. LICs are now increasingly common in large countries such as the USA - for small amounts of students. The educational outcomes have been described: F More clinical experience F Improved satisfaction F The same or better assessment results F No loss of empathy towards patients F ‘Transformative’ learning experience To date, no descriptions or outcomes from LICs in smaller, more densely populated countries have been published. The aims of this study were therefore to see if an LIC could be adapted for use in such a setting and to see if similar outcomes were met. In this study, I describe the results of such a project in a general practice setting, using a novel analytic technique for clinical education settings. The approach: I developed and implemented an LIC for up to 20 students and evaluated both the process and outcomes of the pilot. To do this I carried out sequential focus groups involving both intervention and control students. In addition, I undertook periods of quasi participant observation for both intervention and control groups. The work was informed by Actor Network Theory (ANT). Here, both the social and material aspects of learning are important. Contemporary social theories of learning may underestimate the importance of more material aspects of learning (such as operating computers). Findings: The findings suggest that it is possible to implement an LIC in small densely populated countries that is acceptable to both students and teachers - albeit in an adapted form. Adaptations included a shorter period of immersion in the LIC - but offset by developing bespoke teaching methods for the LIC. The outcomes are similar to those of conventional LICs. However ‘transformation’ of students seen in many LICs was not found. Compared to control students LIC student learning networks were more durable, the students talked more about ongoing relationships, both with teachers and

ABSTRACTS OF ORAL PRESENTATIONS patients. The LIC students also showed more proficiency with important material factors in their education such as knowing how to use computers. Consequences: There is concern regarding the loss of empathy exhibited by medical students. Recent breakdowns in healthcare provision have often been attributed in part to the lack of compassion shown by carers. To date, LICs are the only educational intervention that appear to show that it is possible to maintain empathy within the current structure of medical education. However, there are problems with implementing LICs in small densely populated countries. With appropriate modifications, it may be possible to implement LICs in European countries. Development of bespoke teaching methods may facilitate this process. 4C.3 Enabling senior medical students to engage with social and exercise prescribing, and address the lifestyle risk factors associated with obesity: what needs to be in place? Kathleen Leedham-Green, Ann Wylie King’s College London School of Medicine, London, UK The problem: With the global rise in non-communicable diseases associated with modifiable behaviours, health promotion has become an increasingly important part of undergraduate medical training, often delivered in a primary care setting. As GP tutors (GPTs) may not themselves have been trained to use evidence-based methods to address obesity and the associated lifestyle risk factors, should medical schools be simultaneously and systematically supporting GPTs to update their practice in order to provide appropriate role modelling and support to students? Or is it sufficient for this teaching to be provided on campus? The approach: Prior to their eight week GP placement, final year medical students were provided with campus-based teaching on health promotion and the social determinants of health, and required to submit 4 short reflective cases studies on patient encounters related to obesity and other social or community needs. These case studies were compulsory although not formally assessed. As part of an ethically approved research programme, randomly selected obesity-related case studies (30 out of 140 per rotation since Sep 2011) were analysed using mixed methods, including content and thematic analysis, looking for changing student engagement and confidence in proactively broaching and addressing obesity in response to progressive teaching enhancements. Teaching evolved from providing basic information, to more comprehensive behavioural change training (dietary change, exercise and social prescribing) including role play. In May 2013 a subset of 60 (out of 200+) GPTs attended a training day which was designed to enable them to adapt and update their practice as well as their teaching. A comparative analysis was made of case studies by students based with these trained GPTs (30 out of 30 students) and students on the same rotation based elsewhere (30 out of 110 students). Findings: There have been progressive positive changes in student case studies since 2011 with additional improvements evident for those students who were supervised by GPTs that had attended the training day. These students reported confidence in using the approach taught, were significantly more likely to report broaching the subject of weight proactively, felt able to respond to their patients and knew where and how to access appropriate resources. GPTs also reported significant, lasting, increased confidence in both teaching and broaching obesity issues. Consequences: The global rise in morbidity associated with modifiable health behaviours means that qualifying doctors will need to be confident and competent in broaching and addressing patient lifestyle risk factors. This study suggests that GPTs benefit from support and training in these less traditional but essential aspects of medical education and that their

training has a positive impact on students’ learning and skill development, with the added benefits of potentially improving the clinical practice of the GPTs as well as their future trainees. 4C.4 Facilitation of GP Trainer revalidation in the West Midlands Jane Coomber1, Rodger Charlton1, Martin Wilkinson2 University of Nottingham, Nottingham, UK, 2Health Education West Midlands, West Midlands, UK

The problem: General practitioner (GP) Trainers will be required to provide evidence that they are fit to practice in their educator extended practice role from December 2012 onwards for revalidation and by July 2016 for General Medical Council’s recognition as a medical educator (ie GP Trainer revalidation). The aim of the study was to explore the feasibility of GP Trainers to collect medical educator evidence mapped against the Academy of Medical Educators’ (AOME) A Framework for the Professional Development of Postgraduate Medical Supervisors for Trainer Revalidation. The approach: GP Trainers in the West Midlands were invited to submit medical educator supporting information they had collated from April 2012 to March 2013 that could be mapped against the AOME’s postgraduate medical educator framework and complete an issues log to report the ease of collecting this evidence. Findings: 17/46 West Midlands GP Trainers who consented to participate in the study submitted 140 items of educator supporting information, the majority of which was evidence of continual professional development (41.4%) followed by teaching activities (24.3%) and GP trainee feedback (13.6%). The least amount of evidence had been collated for trainee assessment (2.1%), trainee supervision (2.1%) and trainee career progression (1.4%). The participants reported it was easier to source evidence where organisational support and systems were in place. Consequences: GP Trainers were already collecting medical educator supporting information as part of their whole practice appraisal. Modification of the recording of existing educational activity and organisational support could increase the comprehensiveness of educator evidence which could be presented at annual appraisal with minimal additional resources. 4D.1 Understanding support for patient self-management: a proposed new taxonomy to help identify and design interventions - The Self-Management Support Taxonomy. Steph Taylor1, Gemma Pearce2, Hannah Parke3, Eleni Epiphaniou1, Claire Bourne2, Hilary Pinnock4 1 Barts and The London School of Medicine and Dentistry, London, UK, 2 Coventry University, London, UK, 3King’s College London, London, UK, 4 Edinburgh University, Edinburgh, UK The problem: Promoting effective self-management is a key response of healthcare systems globally to the increasing prevalence of long term conditions. Interventions supporting self-management are already well established as the mainstay of treatment for asthma and Type 2 diabetes, and supporting effective self-management across a wide range of long term conditions is a core activity in primary care. However interventions (often complex and multi-component) to support self-management are currently poorly characterised due to the absence of a comprehensive taxonomy, or classification, of self-management support components. We developed a taxonomy of self-management support components based on the research literature. 69

ABSTRACTS OF ORAL PRESENTATIONS The approach: As part of a project synthesising the evidence on selfmanagement support across a range of long term conditions we held an expert stakeholder workshop and invited participants to suggest the potential components of self-management support interventions. Components were also extracted from systematic reviews of selfmanagement support interventions across a diverse range of 12 long term conditions identified by the workshop participants and purposively selected for variety. From these sources the taxonomy was developed iteratively, involving several rounds of testing, developing and refining. Findings: Fourteen distinct components were identified which may be delivered directly to patients and/or caregivers. In addition, five components were identified which could be delivered indirectly, either at the health or social care professional level, or at the organisational level. In the taxonomy the components are presented with elaboration and examples from the literature. The 14 direct components are: 1) education about condition and management; 2) information about available resources; 3) provision of/agreement on specific action plans and/or rescue medication; 4) regular clinical review; 5) monitoring of condition with feedback to the patient; 6) practical support with adherence (medication or behavioural); 7) provision of equipment; 8) safety netting; 9) training/rehearsal to communicate with health care professionals; 10) training/rehearsal for activities of daily living; 11) training/rehearsal for practical self-management activities; 12) training/rehearsal for psychological strategies; 13) social support; 14) lifestyle advice and support. The taxonomy also recognises overarching characteristics of selfmanagement support interventions: different modes of delivery, e.g. face to face, remote, telehealthcare, web-based; different personnel delivering or facilitating the support: e.g. lay people (with or without training), different types of healthcare professionals; different duration and intensity of the interventions; and different targeting: e.g. generic, culturally-specific or individually tailored. Consequences: We hope this classification will aide in the understanding and replication of interventions, the interpretation of intervention trials, the synthesis of evidence, and the development or implementation of interventions. Funding: NIHR Health Services and Delivery Research programme (11/1014/04) Department of Health Disclaimer: The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the HS&DR programme, NIHR, NHS or the Department of Health. 4D.2 Does self-management support increase the risk of mortality amongst patients with COPD? A systematic review of the evidence Kate Jolly1, Saimma Majothi1, Alice Sitch1, Elizabeth Bates1, Nicola Heneghan1, Richard Riley1, Malcolm Price1, Alice Turner1, Sue Bayliss1, David Moore1, Sally Singh2, Peymane Adab1, David Fitzmaurice1, Rachel Jordan1 1 University of Birmingham, Birmingham, UK, 2Coventry University, Coventry, UK The problem: Self-management support is recommended worldwide for patients with COPD, although interventions are heterogeneous and evidence of their effectiveness is inconsistent. In the United States a large RCT was stopped prematurely because of increased mortality among patients in the self-management arm raising concerns about the role and content of self-management interventions for COPD. Some commentators have questioned whether participants failed to follow the action plans for 70

early response to the symptoms of an exacerbation. We undertook a broad, comprehensive systematic review to evaluate whether self-management support increases mortality. The approach: RCTs were sought to May 2012 using a comprehensive search strategy with no language restrictions. Population: patients with COPD; intervention: any intervention including self-management components compared with usual care/control; outcome: all-cause mortality; duration: follow-up at least 24 weeks. Pooled estimates of hazard ratios were obtained from random effects meta-analysis. Pre-specified subgroup analyses explored the role of disease severity, previous hospitalisation for COPD and the inclusion of action planning within the intervention. Findings: From 13477 records, 822 full text articles were assessed and 44 RCTs were included. There were 447 deaths among 5522 patients. Trials varied in size (n=26 to n=743), intervention (from single components to integrated care) and generally had few events. Overall there was no clear evidence of a difference in mortality, although the point estimate favoured the self-management interventions (HR 0.90 (95%CI 0.74, 1.09)). Sub-group analyses among trials of patients with more severe disease, prior admissions (HR 0.93 (0.72, 1.19)) or provision of action plans (HR 0.89 (0.70, 1.12)) did not differ. However, interventions with the most scheduled contact with healthcare professionals showed reduced mortality (HR 0.72 (0.55, 0.95)). I2=0% (no heterogeneity) for all results. Consequences: There is no evidence that self-management support increases mortality. Only the most supported interventions (e.g .disease/ case management/home rehabilitation) with substantial scheduled contact with health care professionals may reduce mortality. 4D.3 Implementing self-management for long-term conditions: a systematic review of implementation studies Hilary Pinnock1, Eleni Epiphaniou2, Hannah Parke1 ,3, Gemma Pearce2 ,4, Stephanie Taylor2 1 Centre for Population Health Sciences, The University of Edinburgh, Edinburgh, UK, 2Centre for Primary Care and Public Health, Queen Mary University of London, London, UK, 3Health Services and Population Research Department, Kingâ&#x20AC;&#x2122;s College London, London, UK, 4Applied Research Centre in Health and Lifestyle Interventions, Coventry University, Coventry, UK The problem: Promotion of self-care is a core response of healthcare systems globally to the challenge of an aging population and increasing prevalence of long-term conditions (LTCs). In the UK, however, despite promotion in policy and a plethora of initiatives, implementation of self-management remains patchy. The PRISMS programme of systematic overviews of self-management support interventions has the over-arching aim of synthesising evidence to inform commissioners and providers of services for people with LTCs. In this systematic review we identified studies evaluating implementation strategies for supported self-management in 13 exemplar conditions (identified in a multidisciplinary workshop at the outset of the programme). The approach: We searched/screened records from seven electronic databases, and performed snowball and manual searches, selecting studies in which the self-management intervention was delivered as part of routine clinical service. We quality assessed all eligible papers, and extracted and synthesised data. Outcomes of interest were use of healthcare services, disease-specific clinical outcomes, and process outcomes. We used the whole systems approach as a framework for our narrative analysis. This considers interventions from a multilevel perspective (patients, professionals and organisation).

ABSTRACTS OF ORAL PRESENTATIONS The findings: We included 61 papers reporting implementation studies in asthma (n=19), diabetes (n=28), depression (n=6), hypertension (n=2), chronic obstructive pulmonary disease (n=2), chronic kidney disease (n=2), low back pain (n=1), stroke (n=1) delivered in primary, secondary, community and managed care settings in 14 countries. Study designs varied (8 randomised, 18 longitudinal database studies, 35 uncontrolled observational studies) reflected in a range of quality assessments. We grouped the studies by whether the intervention was 1) primarily professional training, 2) primarily patient education, 3) primarily organisational change or 4) a whole systems approach with components explicitly operating at all three levels. Complex whole systems interventions in which motivated, informed patients and trained professionals operate within an organisation with a culture of supported self-management are associated with more beneficial clinical and process outcomes than single component interventions. Success requires strong clinical leadership and commitment at healthcare system level, integration of self-management principles into routine clinical care, professionals motivated to acquire new skills and change current practices, and regular oversight and evaluation to sustain the programme. Consequences: Implementing complex self-management support interventions in routine clinical care is challenging and requires high-level ,multidisciplinary commitment, time and flexibility to adjust to the specific organisations’ processes, priorities, and mentality, as well as procedures for assessing and facilitating fidelity. There are substantial methodological challenges in evaluating implementation of complex interventions which are currently poorly defined. Funding: NIHR Health Services and Delivery Research programme (11/1014/04) The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the HS&DR programme, NIHR, NHS or the Department of Health

of PSM and its use in the long term. Interviews were audio taped and transcribed verbatim and analysed using constant comparative analysis. Findings: Patients generally had a positive perception of PSM prior to engaging with the trial. Patients described expectations of lifestyle benefits including more freedom to travel, improved consistency of monitoring, and a reduced time burden associated with routine care. Patients also described altruistic feelings in relation to reducing the burden of health care professionals. Patients’ perceptions remained positive after undertaking PSM, and for many PSM exceeded expectations. The perceptions of patients who experienced PSM fall into three main themes: autonomy, convenience and peace of mind. Long term PSM was largely acceptable to patients who engaged with PSM, and the main barrier to long term PSM was the cost associated with the coagulometer and test strips. On the other hand some patients were more comfortable with routine care mainly because of a perceived reassurance in being cared for by health professionals and a few valued the social interaction involved in routine testing. Consequences: These findings suggest patients on OAT are open to PSM and would engage in long term PSM. The costs associated with PSM and a perceived lack of supervision and support from health care professionals is however a barrier to PSM. Warfarinised patients are more likely to view PSM as a viable alternative model of care if these issues are addressed. 4E.1 Financial incentives to providers to improve smoking cessation in pregnancy and breastfeeding outcomes: acceptability and perceived consequences. Pat Hoddinott1 ,2, Heather Morgan2, Gill Thomson3, Nicola Crossland3, Graeme MacLennan2, Kate Sewel4, Linda Bauld1, Marion Campbell2 1 University of Stirling, Stirling, UK, 2University of Aberdeen, Aberdeen, UK, 3 University of Central Lancashire, Preston, UK, 4Ipsos MORI, Edinburgh, UK

“The best thing since sliced bread”: patients’ perceptions of self monitoring of warfarin

The problem: Smoking in pregnancy and/or not breastfeeding increase the risk of adverse infant and maternal health outcomes. These behaviours are challenging to change and sustain. Financial incentives are proposed as a potential solution - but are controversial. There is little evidence for provider incentives for proven smoking cessation and only one intervention with promise for breastfeeding.

Patricia Apenteng1, Deborah McCahon1, Katie Murray2, Ellen Murray1, David Fitzmaurice1 1 University of Birmingham, Birmingham, UK, 2Imperial College and West London Mental Health NHS Trust, London, UK

The approach: Aims: To use mixed methods to explore the acceptability and potential consequences of incentives for smoking cessation and breastfeeding amongst general public, health professionals and childbearing women to inform trial design.

The problem: The clinical efficacy of patient self-monitoring (PSM) of warfarin therapy with a portable coagulometer has been widely demonstrated. NICE recently recommended that PSM is considered as an alternative model for oral anticoagulation therapy (OAT) service delivery for patients with atrial fibrillation. However in the UK it is estimated that only 1-2% of the 1.2 million people receiving warfarin therapy engage with this model of care. The reasons for this have not been fully elucidated. We report findings from a qualitative study exploring patients’ perceptions of PSM conducted as part of a primary care based randomised controlled trial (SMART) evaluating the clinical and cost effectiveness of PSM in comparison with routine care.

Methods: Seven promising incentive strategies were identified from evidence syntheses and qualitative interview data from service users and providers. Two were incentives to providers; both comprised payments to local health services for reaching (i) smoking cessation in pregnancy targets and (ii) breastfeeding targets. Five were incentives to women, including conditional shopping vouchers; a free breast pump. An Ipsos MORI survey used area quota sampling and home-administered computer-assisted questionnaires and agreement was measured on a 5-point scale (1144 UK general public respondents). The same survey with free-text questions about positive and negative consequences was sent (email link to www.surveymonkey.com) to health professionals across Scotland and North West England. The 497 respondents included 156 doctors (132 were GPs); 197 midwives/health visitors/maternity-care staff; 144 other staff (which included public health, nurses, managers, support staff). Survey analysis used multivariable ordered logit models.

4D.4

The approach: Semi-structured face to face interviews were undertaken with 23 participants of the SMART study. Purposive sampling was employed to ensure a maximum variety sample that included patients randomised to PSM who completed 12 months of PSM, patients who were withdrawn from PSM before the end of the 12 months trial period and patients randomised to continue with routine care. The topic guide comprised a series of open ended questions that drew reflections on previous care in comparison with PSM, expectations of PSM, perceptions

We conducted qualitative interviews (16 focus groups; 55 face-to-face; 19 telephone) with 88 pregnant women/recent mothers/partners; 53 service providers; 24 experts/decision makers, 63 conference attendees. Iterative mixed methods analysis (NVivo10) used a thematic Framework approach. 71

ABSTRACTS OF ORAL PRESENTATIONS Findings: Net agreement with the seven incentive strategies was mixed (range 34-69%). Provider incentives for smoking cessation targets: public net agreement 39.4% (451/1144); net disagreement 37.2% (426/1144). Professional net agreement 52.9% (263/497); net disagreement: 28.6% (142/497). Those more likely to agree were aged under 54; midwives/health visitors/ maternity-care staff (compared to doctors): OR 2.35 (95% CI 1.51, 3.64; p < 0.001). Provider incentives for breastfeeding targets: public net agreement 36.4% (416/1144); net disagreement 38.6% (441/1144). Professional net agreement 44.1% (219/497); net disagreement 38.6% (192/497). Those more likely to agree were: aged under 45; ethnic minorities; women (compared to men) OR 1.81 (1.09, 3.00; p =0.023) and midwives/health visitors/maternity-care staff (compared to doctors): OR 2.54 (95% CI 1.65, 3.91; p < 0.001). Positive consequences included health benefits; increased resources; “a shared aim”. Negative consequences included relationships with patients: “pressure” and “blame” and between staff: “low morale”; increased health inequalities; the burden of proof; opportunity costs; gaming. Consequences: As acceptability is mixed, high quality evidence, preferably from randomised interventions, is needed prior to policy implementation. The cost-effectiveness is unknown and the potential for adverse unintended consequences is considerable. Funding: NIHR/HTA: 10/31/02 4E.2 Parent’s perceptions of barriers to and facilitators for keeping children safe at home; a qualitative study Joanne Ablewhite1, Isabel Peel2, Adrian Hawkins3, Lisa McDaid4, Denise Kendrick1 1 University of Nottingham, Nottingham, UK, 2Nottingham University Hospitals, Nottingham, UK, 3Newcastle Upon Tyne Hospitals NHS Foundation Trust, Newcastle, UK, 4University of East Anglia, East Anglia, UK The problem: Injuries in childhood cost the NHS approximately £240 million per year and are a leading cause of death in children aged 1-4 years. Falls, poisonings and burns or scalds are the most common childhood injuries resulting in emergency department attendance and hospital admission. The majority of these injuries occur within the home as that as where the child spends most of their time. The aim of this study was to identify key facilitators and barriers for parents in keeping their children safe from unintentional injury within their home. A further aim was to develop an understanding of parents perceptions of what might help them to implement injury prevention activities. The approach: A total of 64 parents with a child aged between 0-4 years, across four centres (Nottingham, Bristol, Norwich and Newcastle) were recruited. Forty-eight parents were recruited as their child had attended an emergency department following a fall, poisoning or scald. Sixteen parents were recruited whose child did not have an injury (four per centre). Semi-structured interviews were conducted in the homes of parents. Interview data was transcribed verbatim, thematic and content analysis was used to analyse the data. Findings: Barriers to injury prevention included; parents not anticipating injury risks or injuries resulting from some risk taking behaviours, an acceptance of some injuries as an inevitable part of child development, impeded parental supervision due to distractions, maternal fatigue, the 72

number of children in the household and the presence of older siblings. Limitations related to safety equipment were described as was the cost of safety equipment. The timing of safety information in relation to the ages and stages of child development was described by some parents as being received “too late”. Facilitators for injury prevention included parental supervision and teaching children about injury risks. This included parents allowing children to learn about injury risks through controlled risk taking, using “safety rules” and supervising the child to ensure safety rules are adhered to by the child. Adapting the home by installing safety equipment or removing hazards was also key to injury prevention practices. Some parents described how learning about injury events through other parent’s experiences may be a good way to increase the anticipation of injury risks within the home. Consequences: The issues highlighted provide a better understanding of what influences safety behaviours and what can be done to help parents. Appropriately targeted information to parents in line with the ages and stages of child development may help parents to anticipate child injury risks. One method may be through case studies of parents whose child has experienced an injury so that other parents can learn about risks and injuries which may result from these. 4E.3 Parental home safety practices and risks for childhood poisonings: findings from a multi-centre case-control study. Gosia Majsak-Newman1, Penny Benford2, Clare Bryan2, Denise Kendrick2 1 Norfolk and Norwich University Hospitals NHS Foundation Trust, Norwich, UK, 2School of Medicine, Division of Primary Care, Nottingham, UK The problem: Childhood poisonings persist as one of leading causes of unintentional injury in under 5’s and are a serious public health problem. They place a substantial burden on the NHS, with more than 26000 emergency department (ED) attendances more than 5000 hospital admissions in 0-4 year olds each year and an estimated annual cost of more than £2 million for poisoning injuries in under 15s. There is limited evidence about whether poisoning prevention practices reduce poisoning rates. The approach: A case-control study was undertaken in 7 NHS trusts across the UK. Cases were aged 0-4 years with a medically attended poisoning or suspected poisoning from medicines or household products, occurring at home. Cases were matched on age and sex with community controls. We used children attending with other types of injury as unmatched hospital controls. Parents completed a questionnaire to measure home safety behaviours, use of safety equipment and potential confounding factors. Responses were validated by home observations in a sample of participants. Conditional and unconditional logistic regression models were used in matched and unmatched analyses respectively. Analyses adjusted for deprivation and distance from hospital (plus age and sex in the unmatched analysis) and other confounders as identified by directed acyclic graphs. Findings: 567 poisoning cases, 2253 hospital and 2320 community controls participated. Results from the matched analysis suggest parents of cases were significantly more likely to not store medicines at adult level or above (OR 1.59, 95%CI 1.21, 2.09) and to not store medicines safely (locked away or at adult eye level or above) (OR 1.67, 95%CI 1.23, 2.27). Parents of cases were more likely not to have taught their children safety rules about what to do when finding medicines (OR 1.44, 95%CI, 1.00, 2.07). When comparing cases with hospital controls, further risk factors were identified including not having child resistant containers (CRCs) for all medicines (OR 2.31, 95%CI 1.74, 3.07) and cleaning products (OR 1.43, 95%CI 1.14, 1.81), having things child could climb on to reach worktops, shelves or cupboards (OR 1.48, 95%CI 1.16, 1.88) and not teaching safety rules about medicines (OR1.97, 95%CI 1.45,2.69) and household products (OR 1.61, 95%CI 1.21, 2.13). Children 1-4 years old

ABSTRACTS OF ORAL PRESENTATIONS and boys were at higher risk for poisoning when medicines and household products were not locked or stored safely. Consequences: A range of safety behaviours significantly increase the risk for unintentional poisonings in young children. Primary care practitioners can help parents reduce the risk of poisoning by advising on safe storage, use of CRCs, avoiding having things children can climb on to reach substances and teaching children safety rules. Advice should be provided in child health surveillance contacts, when prescribing for families with young children and after poisoning events. 4E.4 Why telling parents “antibiotics don’t work on viruses” doesn’t reduce expectation for antibiotic treatment for children with RTIs: a qualitative study Christie Cabral1, Jenny Ingram1, Patricia Lucas1, Niamh Redmond1, Joe Kai2, Alastair Hay1, Jeremy Horwood1 1 University of Bristol, Bristol, UK, 2University of Nottingham, Nottingham, UK Background: Acute respiratory tract infections (RTI) in children are the most common reason parents’ consult primary care in the UK. There is clinical uncertainty around identifying those cases for which antibiotics are appropriate which contributes to over-prescription. Parents report being confused by RTI information and advice received during consultations and feel their concerns are not always addressed. Misunderstandings in the consultation can contribute to unnecessary antibiotic prescription and future consultations. This study examined real-time interaction in primary care consultations for children with RTI and the parent and clinician reflections on the communication in order to understand how shared understanding or misunderstandings came about. Methods: Recruitment was conducted in 6 primary care practices selected from socioeconomically contrasting localities. 60 primary care consultations for children aged 3 months to 12 years were videoed. Parents completed pre- and post-consultation questionnaire, including measures of expectation, satisfaction and enablement. Video elicitation interviews were carried out with 13 clinicians and a sub-sample of the parents. 27 parents were selected using purposive sampling to capture maximum variation in terms of parent satisfaction and enablement. Parents and clinicians were shown the consultation video and asked to describe their thought processes and feelings at key points. The audio recordings of the interviews and consultation were transcribed and imported into NVivo8 for coding and a thematic analysis was conducted. Results: Parents were very familiar with the message that antibiotics are not effective against viral infections and were aware that clinicians are trying to reduce the number of antibiotic prescriptions. However, in most cases, this does not affect parents’ beliefs about the need to consult for cough in their child or expectations for antibiotic treatment because these are based on the perceived severity of symptoms. The perception that clinicians are less willing to hand out antibiotics and the little understood explanation has given rise to a range of lay theories to explain the perceived changes in clinicians antibiotic prescribing behaviour. One theory attributed attempts to reduce antibiotic prescribing to resource rationing, with the consequence that some parents believed they had to argue more strongly for their child to be given antibiotics. Another lay belief that taking antibiotics undermined the development of the child’s ‘natural’ immunity contributed to a parent’s decision not to give their child antibiotics which had been prescribed. Conclusions: The message that antibiotics don’t work on viruses has been successfully disseminated to parents. Unfortunately it has had little impact on expectation of antibiotics because these expectations were based on perceived severity, rather than the little understood difference between causal microbes. Lay theories offer alternative explanations for the reduced

prescribing of antibiotics and some have unfortunate consequences including both stronger antibiotic seeking and (if only rarely) antibiotic avoiding even when they are clinically indicated. 4F.1 Advanced Care Planning decisions amongst Irish Care Home Residents. Does it help to ‘Think Ahead’? Hugh Brady1, Joseph Marry1, Deborah Martin1, Connor Gallagher1, Catherine Darker1 ,2, Brendan O Shea1 ,3 1 TCD HSE GP Training Scheme at The Trinity Teaching Centre, Tallaght Hospital Dublin 24, Dublin, Ireland, 2The Department of Public Health and Primary Care, Trinity College Dublin, Dublin, Ireland, 3Kildare and West Wicklow Doctors on Call Ltd (K Doc), Kildare, Ireland, 4The Adelaide Society, Tallaght Hospital, Tallaght, Dublin 24, Dublin, Ireland The problem: Inadequate end of life planning in nursing homes leads to unnecessary hospital transfers, distressing patients and families, and is expensive for healthcare systems. Clear documentation of end of life preferences is likely to reduce inappropriate transfers. Documentation is important for on call GPs, unfamiliar with the patient. The aim of the study is to assess the impact of a GP led multidisciplinary educational intervention on end of life planning in the nursing home setting, utilising a mixed methods waiting list controlled intervention design. Outcomes include rates of advanced care planning in nursing home medical records, before, and 4 months after educational intervention, and focus group outcomes (GPs and Nurses). The approach: Eight purposively selected nursing homes (Ireland) in the catchment of a GP out of hours co-operative were randomly assigned to intervention (5) and control (3). Medical records were examined for evidence of end of life planning. Results were categorised for clarity of planning in the medical record. Category A included explicit direction, category B some evidence, and C represented no evidence of planning. Cognition was simply determined by a senior member of nursing staff, and categorised as normal or impaired to a mild, moderate or severe degree. Patient demographics but no identifiers were recorded. Ethical approval was obtained from the TCD HSE GPTS Ethics Committee. The intervention included a focused package of learning, utilising the ‘Think Ahead’ end of life planning tool, to all Co Operative GPs, and intervention nursing home staff and attending GPs. A second survey of files, and results from focus groups with GPs and nursing home staff will be captured (April 2014). If improved rates of planning are evident, the educational package will be then be delivered to control nursing homes. Findings: 489 resident files were surveyed, 65% female, average age 81 (range 28-100). Only 23% had detailed documentation (A) of advance healthcare decisions, and 40% had none (C); 36% of residents had normal or mild impaired cognition; 41% had severely impaired cognition. Only 28% with normal cognition (compared with 44% of with severely impaired cognition) had planning documented. Consequences: Almost 3 out of 4 patients with normal cognition did not have planning documented. Those most able to communicate preferences regarding end of life were least likely to have this recorded. ‘Think Ahead’ (www.thinkahead.ie) is an innovative end of life planning tool, which may be helpful in this important area of care. Final results will determine if it is effective in the nursing home setting. If this is the case, our results may have broader applicability in the Nursing Home setting. 73

ABSTRACTS OF ORAL PRESENTATIONS 4F.2 Anticipatory prescribing at the end-of-life in South Edinburgh care homes Anne Finucane1, Dorothy MacArthur2, Hilary Gardner1, Scott Murray3 1 Marie Curie Cancer Care, Edinburgh, UK, 2NHS Lothian Pharmacy Service, Edinburgh, UK, 3University of Edinburgh, Edinburgh, UK The problem: Common symptoms at the end-of-life include pain, breathlessness, anxiety, respiratory secretions, and nausea. National end-of-life care strategies advocate anticipatory prescribing to manage these symptoms, enhance patient care and reduce unnecessary hospital admissions. NHS Lothian palliative care guidelines have drawn on this advice and recommend that four classes of â&#x20AC;&#x153;as requiredâ&#x20AC;? medicines are prescribed and available for all patients in the last days of life. These are analgesics, anxiolytics, and anti-secretory and anti-emetic medication. However, the extent to which anticipatory prescribing guidelines are being implemented in care homes is unclear. This study explored the extent to which residents in eight South Edinburgh care homes had anticipatory medications prescribed prior to death. The approach: Data was collected as part of a larger project to improve palliative care in South Edinburgh care homes. The manager or lead nurse in eight care homes completed an after death audit following the death of each of the last ten care home residents who had died. This included information on anticipatory prescribing. The findings: Data on 77 residents was received, 71 of whom died in the care homes (6 died in hospital). Of the 71 who died in the care homes, 59% had some anticipatory planning medications in place: morphine for analgesia (52%); midazolam for anxiety and distress (38%); hyoscine butylbromide for respiratory secretions (21%); and levomepromazine or haloperidol for nausea (23%). Overall, 41% no anticipatory medications prescribed in the last days of life. Only 15% had prescriptions for all four medications. Consequences: Many care home residents do not have the recommended anticipatory medications in place in the last days of life and thus may experience inadequate symptom control. Barriers to anticipatory prescribing in care homes include difficulties diagnosing dying, lack of knowledge about the guidelines, cost concerns relating to wastage of medicines and concerns relating to storage. Research that explores these issues is required. In the meantime, interventions that build the skills and confidence of care home staff to diagnose dying and facilitate such prescribing are recommended. 4F.3 Integrating Palliative Care into national health systems in four African Countries: developing models of best practice Liz Grant1, Mhoira Leng2, Liz Namukwaya2, Julia Downing2, Emilly Kemigisha3, Scott Murray1 1 Primary Palliative Care Research Group, The University of Edinburgh, Edinburgh, UK, 2Makerere University, Kampala, Uganda, 3African Palliative Care Association, Kampala, Uganda The problem: Though palliative care encapsulates the very essence of humane care - care of those struggling to care for themselves because of the severity of their illness or frailty - it is rarely considered a priority or funded accordingly. High numbers live with life limiting illnesses, particularly in many low income countries, yet culturally appropriate, holistic palliative care (that includes effective pain and symptom management) provision is inconsistent, rarely integrated into mainstream health systems, often delivered sporadically. Islands of excellence in palliative care delivery exist, but the processes for scale-up to achieve 74

national coverage are missing. Ministries of Health in Africa note the lack of effective, persuasive, evidence based models to illustrate how such care can be delivered nationally in a cost effective and equitable way. The approach: To counter the dearth of policy shaping evidence, we established an action research programme to build and integrate palliative care into 12 different Government funded hospitals and their community catchment areas in four African countries (Kenya, Uganda, Zambia and Rwanda) and to capture lessons from these 12 models of integrated care. Funded by the UK Department for International Development the programme delivered training to multi-cadred professionals, created palliative care clinical hubs, established mentorship demonstration programmes, implemented protocols, trained pharmacists and established drug programmes, mapped pathways to care and built new referral systems for patients who had palliative care needs. For each hospital a matrix was developed capturing sector of care provision, approximate caseload, population service, staffing, setting, services provided, types of patients seen, referral process, patient focus, management of patients, access to essential medicines, instruments used and clinical protocols, and standards of care. Findings: Previously documentation of service delivery was rare, but rarer still was the documenting of system and service processes. Capturing comparable models of care creates new spaces and opportunities for practitioners, hospital managers and community workers to come together to explore their practices, and the needs of patients. Implementation of new knowledge requires social, emotional, political and resource changes which are rarely dealt with. Training was most effective when supported by a mentorship to enabling staff to embed knowledge into practice. Mobile technology offers an innovative way to provide services and map activities. Consequences: The 12 introduced and models mapped out across 12 African hospitals provides Governments with good practice systems, skilled practitioners and most importantly rigorous evidence of the impact of care. To make a long term difference in palliative care in low income countries, essential clinical training, change management of services and systems needs to be coupled with rigorous data collection, interpretation of data, and logistical support to enable integration into the main health system. 4F.4 Different dyings: living and dying with cancer, organ failure and physical frailty Marilyn Kendall1, Emma Carduff1, Anna Lloyd1, Barbara Kimbell1, Debbie Cavers1, Susan Buckingham2, Kirsty Boyd1, Liz Grant1, Allison Worth2, Hilary Pinnock2, Aziz Sheikh2, Scott Murray1 1 Primary Palliative Care Research Group, The University of Edinburgh, Edinburgh, UK, 2Allergy and Respiratory Research Group, The University of Edinburgh, Edinburgh, UK The problem: Providing good end-of-life care for people with all illnesses is a recognised global challenge. Better understanding of how different patients perceive their deteriorating health and approaching death can inform appropriate palliative care developments. We aimed to establish whether there are typical narratives of living with and dying from different advanced illnesses, that might provide insights into how best to provide effective care for people dying of all conditions. The approach: We synthesised data from eight longitudinal studies we had previously conducted: Three in cancer - lung, glioma and colorectal; three in organ failure - heart failure, chronic obstructive pulmonary disease (COPD) and liver failure; one in frailty; and one with South Asian participants from all three trajectories. Each study used purposive sampling to gain a broad range of perspectives. The patients were interviewed up to four times over 12 to 18 months. We developed an analytic framework based on three phases - becoming ill,

ABSTRACTS OF ORAL PRESENTATIONS living with advanced illness, and dying. Researchers from each study used this framework to review the data from their own study, and generate key themes, with illustrative quotations and a narrative summary. Participantsâ&#x20AC;&#x2122; narratives were initially synthesised by illness trajectory (i.e. cancer, organ failure or frailty) and then compared across the three phases of the illness narrative. The findings were iteratively discussed through dedicated analysis workshops. Findings: The dataset comprised 828 in-depth interviews with 156 patients, 114 family caregivers and 170 health professionals. The cancer narratives typically had a clear beginning, middle and anticipated end. They gave a well-rehearsed account of the illness, with the dual themes of hope for recovery alongside fear of dying. In contrast, people with organ failure struggled to pinpoint when their illness began, or see meaningful patterns between acute events. Fewer spoke about death, hoping instead to avoid further deterioration. Frail older people were frustrated by their progressive loss of independence, fearing dementia or nursing home admission more than dying. Consequences: Patients with different illness experienced different thoughts, fears and patterns of physical decline. Consequently the cancer-based model of palliative and end-of-life care seems poorly suited to the needs of those dying from other long term conditions. Indeed â&#x20AC;&#x2DC;Palliative careâ&#x20AC;&#x2122; as a concept may be too inextricably linked to cancer and imminent death to be acceptable and appropriate for other patients. Rather all professionals caring for people with advanced conditions will need core skills to enable them to assess supportive care needs and participate in discussions about the future. Palliative care specialists could train and support generalists, and work collaboratively with other specialists, to provide person-centred supportive care for people living with the unpredictability of long-term conditions and general frailty.

PRESENTER INDEX

Ablewhite, Joanne

EP-2A.01 Children/young people/women’s health

Friday

Ablewhite, Joanne

4E.2

Children and young people

Friday

Agur, Karolina

EP-2E.01 Multimorbidity/user experience

Ahmad, Shaleen

1C.2

Ahmed, Faraz

EP-2C.04 Patient-GP/person centred care/older people

Alcock, Elizabeth L

2C.2

Musculoskeletal

Thursday am

Apenteng, Patricia

4D.4

Self-management

Friday

Apenteng, Patricia

EP-1D.12 Prescribing and patient safety

Wednesday

Ashworth, Mark

EP-1C.07 Cardiovascular

Wednesday

Barnett, Karen

3B.5

Cancer

Thursday pm

Bartlam, Bernadette

3A.4

Managing long term conditions

Thursday pm

Bartlam, Roisin

EP-1A.12 Cancer and palliative care

Wednesday

Bates, Elizabeth

EP-1B.09 Organisation and delivery of primary health care

Wednesday

Bedford, Laura

2F.3

Benford, Penny

EP-2A.04 Children/young people/women’s health

Friday

Berdunov, Vladislav

EP-1D.10 Prescribing and patient safety

Wednesday

Birtwhistle, Richard

EP-2D.03 Information technology/infection/sexual health

Friday

Bishop, Felicity

2B.2

Generic issues

Thursday am

Bishop, Felicity

2C.5

Musculoskeletal

Thursday am

Blackburn, Ruth

EP-2F.04

Mental health

Friday

Blackburn, Ruth

3D.2

Population-based practice research

Thursday pm

Blockley, Kelly

EP-1E.02 Diabetes/managing long term conditions/clinical consultation

Boardman, Kerry

1F.1

Education

Wednesday

Bond, Christine

EP-1F.03

Research methods/epidemiology/public health

Wednesday

Bond, Christine

EP-1A.05 Cancer and palliative care

Boon, Veronica

1F.2

Bowen, Leah

EP-2A.05 Children/young people/women’s health

Friday

Bower, Peter

EP-2E.03 Multimorbidity/user experience

Friday

Bower, Peter

EP-1F.02

Boyd, Matthew J

EP-2G.06 Pot pourri 2

Brown, Lynsey

1A.3

Brunt, Eleanor

EP-2A.07 Children/young people/women’s health

Friday

Bryan, Clare

EP-2A.02 Children/young people/women’s health

Friday

Burton, Christopher

EP-2D.02 Information technology/infection/sexual health

Friday

Bury, Gerard

EP-2B.10 Education/ethics/governance

Friday

Bury, Gerard

3D.6

Butterworth, Joanne

EP-2C.05 Patient-GP/person centred care/older people

Button, Katherine

2E.4

Mental health

Thursday am

Cabral, Christie

4E.4

Children and young people

Friday

Calanzani, Natalia

EP-2D.05 Information technology/infection/sexual health

Friday

Calitri, Raff

EP-1B.08 Organisation and delivery of primary health care

Wednesday

Calitri, Raff

4A.2

Organisation and delivery of primary health care 2

Friday

Campbell, John

2D.4

Multimorbidity

Thursday am

Health promotion

Cancer

Education

Research methods/epidemiology/public health Research methods

Population-based practice research

Friday Wednesday Friday

Thursday am

Wednesday

Wednesday Wednesday

Wednesday Friday Wednesday

Thursday pm Friday

PRESENTER INDEX Chalder, Melanie

EP-1F.04

Research methods/epidemiology/public health

Wednesday

Chalder, Melanie

1A.1

Research methods

Wednesday

Chandler, Amy

2E.6

Mental health

Thursday am

Charlton, Rodger

EP-1G.04 Pot pourri

Charlton, Rodger

4C.4

Education 2

Friday

Checkland, Kath

4A.4

Organisation and delivery of primary health care 2

Friday

Cheema, Kamalpreet

EP-1C.06 Cardiovascular

Cheema, Kamalpreet

3E.1

Cardiovascular

Thursday pm

Cheraghi-Sohi, Sudeh

2A.4

Quality and patient safety

Thursday am

Chew-Graham, Carolyn

EP-2F.09

Mental health

Friday

Chew-Graham, Carolyn

EP-2F.06

Mental health

Friday

Chew-Graham, Carolyn

3A.1

Managing long term conditions

Thursday pm

Clark, Christopher

EP-1C.10 Cardiovascular

Clarke, Clare

1E.2

Older people

Wednesday

Clyne, Barbara

1B.1

Prescribing

Wednesday

Clyne, Barbara

EP-1D.04 Prescribing and patient safety

Wednesday

Cook, Amelia

EP-2A.03 Children/young people/womenâ&#x20AC;&#x2122;s health

Friday

Coomber, Jane

EP-2B.01 Education/ethics/governance

Friday

Cooper, Lucy

1A.2

Research methods

Wednesday

Coupland, Carol

2E.1

Mental health

Thursday am

Coxon, Domenica

EP-1A.03 Cancer and palliative care

Wednesday

Crawford, Carole

EP-1C.02 Cardiovascular

Wednesday

Crowther, Jacqueline

EP-1A.11 Cancer and palliative care

Wednesday

Davies, Hilary

EP-1F.12

Research methods/epidemiology/public health

Wednesday

Din, Nafees

3B.1

Cancer

Thursday pm

Drinkwater, Jessica

2B.3

Generic issues

Thursday am

Earthy, Sarah

EP-2E.06 Multimorbidity/user experience

Friday

Eccles, Sarah

EP-2B.03 Education/ethics/governance

Friday

Essam, Nadya

EP-2C.06 Patient-GP/person centred care/older people

Friday

Finucane, Anne

EP-1A.08 Cancer and palliative care

Wednesday

Finucane, Anne

4F.2

Fitzmaurice, David

EP-1C.11 Cardiovascular

Fleetcroft, Robert

1D.2

Foskett-Tharby, Rachel

EP-1D.09 Prescribing and patient safety

Gailer, Ruth E

4A.3

Organisation and delivery of primary health care 2

Friday

Gallacher, Katie

2D.6

Multimorbidity/user experience

Thursday

Garrison, Scott

EP-1G.07 Pot pourri

Gilkes, Alexander

EP-1F.10

Research methods/epidemiology/public health

Wednesday

Grant, Liz

4F.3

Palliative care

Friday

Greaves, Colin

EP-1E.03 Diabetes/managing long term conditions/clinical consultation

Greaves, Colin

3E.3

Cardiovascular

Thursday pm

Green, Dan

2C.4

Musculoskeletal

Thursday am

Green, Trish

2F.6

Cancer

Thursday am

Palliative care Organisation and delivery of primary health care

Wednesday

Friday Wednesday Wednesday Wednesday

Wednesday

PRESENTER INDEX

Griffin, Simon

4B.4

Managing long term conditions 2

Friday

Hall, Nicola

2F.4

Cancer

Thursday am

Hamilton, Fergus

2F.2

Cancer

Thursday am

Hanley, Janet

3A.5

Managing long term conditions

Thursday pm

Harding, Alexander

EP-2B.06 Education/ethics/governance

Harding, Alexander

4C.2

Education 2

Friday

Harries, Timothy

EP-2G.03 Pot pourri 2

Friday

Harris, Tess

1E.3

Older people

Wednesday

Hay, Alastair D.

3C.1

Infection

Thursday pm

Hearn, Russell

4C.1

Education 2

Friday

Heron, Neil

1C.4

Health promotion

Wednesday

Hill, Kate

EP-1G.08 Pot pourri

Hill, Trevor

EP-2F.03

Hinton, Lisa

EP-2A.08 Children/young people/womenâ&#x20AC;&#x2122;s health

Hoddinott, Pat

1C.1

Health promotion

Wednesday

Hoddinott, Pat

4E.1

Children and young people

Friday

Horwood, Jeremy

EP-2G.04 Pot pourri 2

Howe, Amanda

1A.4

Hyde, Catherine

EP-1G.09 Pot pourri

Jani, Bhautesh

EP-2F.02

Mental health

Friday

Jani, Bhautesh

EP-2F.07

Mental health

Friday

Johnson, Maxine

EP-1D.11 Prescribing and patient safety

Jolly, Kate

4D.2

Self-management

Friday

Jones, Melvyn

1F.4

Education

Wednesday

Kellezi, Blerina

EP-1E.01 Diabetes/managing long term conditions/clinical consultation

Kellezi, Blerina

1D.3

Organisation and delivery of primary health care

Wednesday

Kendall, Marilyn

4F.4

Palliative care

Friday

Kenning, Cassandra

PP2.2

PRIZE PLENARY

Wednesday

Kerry, Sally

EP-1F.05

Research methods/epidemiology/public health

Wednesday

Keynejad, Roxanne

EP-1G.01 Pot pourri

Kilgarriff-Foster, Alexis

EP-2F.08

Mental health

Friday

Kingston, Mark-Rhys

3A.2

Managing long term conditions

Thursday pm

Knox, Richard

EP-2B.02 Education/ethics/governance

Friday

Knox, Richard

EP-2B.07 Education/ethics/governance

Friday

Knox, Richard

EP-1G.02 Pot pourri

Wednesday

Kokab, Farina

EP-1C.03 Cardiovascular

Wednesday

Kokab, Farina

EP-1C.08 Cardiovascular

Wednesday

Kordowicz, Maria

EP-1B.05 Organisation and delivery of primary health care

Wednesday

Kotecha, Jyoti

2A.3

Quality and patient safety

Thursday am

Kumar, Arun

1E.1

Older people

Wednesday

Lasserson, Daniel

4B.2

Managing long term conditions 2

Friday

Leaman, Ben

EP-1E.08 Diabetes/managing long term conditions/clinical consultation

Mental health

Research methods

Friday

Wednesday Friday Friday

Friday Wednesday Wednesday

Wednesday

PRESENTER INDEX Leber, Werner

3C.5

Infection

Thursday pm

Leedham-Green, Kathleen

4C.3

Education 2

Friday

Liddle, Jenny

EP-2E.05 Multimorbidity/user experience

Friday

Liddle, Jenny

EP-1G.12 Pot pourri

Wednesday

Liddle, Jenny

EP-1G.11 Pot pourri

Wednesday

Little, Iain

EP-1F.11

Research methods/epidemiology/public health

Wednesday

MacFarlane, Anne

4A.1

Organisation and delivery of primary health care 2

Friday

Macleod, Una

EP-1B.07 Organisation and delivery of primary health care

Wednesday

Magin, Parker

EP-1G.05 Pot pourri

Wednesday

Magin, Parker

3E.2

Magin, Parker

EP-1D.01 Prescribing and patient safety

Mair, Frances

3B.2

Cancer

Thursday pm

Majsak-Newman, Gosia

4E.3

Children and young people

Friday

Man, Mei-See

2D.1

Multimorbidity

Thursday am

Marikar Bawa, Fathima

4B.1

Managing long term conditions 2

Friday

Marry, Joseph

4F.1

Palliative care

Friday

Marsden, Kate

EP-1D.07 Prescribing and patient safety

McCahon, Deborah

3E.6

McCahon, Deborah

EP-2G.07 Pot pourri 2

McLachlan, Sarah

3B.4

Methley, Abigail

EP-1E.12 Diabetes/managing long term conditions/clinical consultation

Mhizha-Murira, Jacqueline

EP-1F.08

Mhizha-Murira, Jacqueline

EP-2E.08 Multimorbidity/user experience

Friday

Middlemass, Jo

EP-1E.09 Diabetes/managing long term conditions/clinical consultation

Wednesday

Montana, Carolyn

1B.4

Moore, Grace

EP-1A.10 Cancer and palliative care

Moore, Michael

3C.4

Morales, Daniel

EP-1D.05 Prescribing and patient safety

Wednesday

Moran, Grace

EP-2G.02 Pot pourri 2

Friday

Morden, Andrew

2C.1

Musculoskeletal

Thursday am

Morgan, Katy

EP-1F.01

Research methods/epidemiology/public health

Wednesday

Morkem, Rachael

EP-1D.06 Prescribing and patient safety

Morkem, Rachael

3D.1

Population-based practice research

Thursday pm

Morrison, Jill

2E.3

Mental health

Thursday am

Morrison, Jill

EP-2F.05

Mental health

Friday

Mounce, Luke

4B.3

Managing long term conditions 2

Friday

Murphy, Mairead

EP-2C.07 Patient-GP/person centred care/older people

Friday

Murphy, Mark

EP-1E.11 Diabetes/managing long term conditions/clinical consultation

Wednesday

Musbahi, Aya

EP-1G.03 Pot pourri

Wednesday

Nalabanda, Ananth

1E.4

Older people

Wednesday

Nash, Vanessa

2B.6

Generic issues

Thursday am

Neal, Richard

EP-2G.05 Pot pourri 2

Cardiovascular

Cardiovascular Cancer Research methods/epidemiology/public health

Prescribing Infection

Thursday pm Wednesday

Wednesday Thursday pm Friday Thursday pm Wednesday Wednesday

Wednesday Wednesday Thursday pm

Wednesday

Friday

PRESENTER INDEX

Neal, Richard

2F.5

Cancer

Thursday am

Nicholson, Brian

2F.1

Cancer

Thursday am

Nugent, Claire

1C.3

Health promotion

Wednesday

Oakeshott, Pippa

EP-2A.09 Children/young people/women’s health

Friday

O’Brien, Kirsty

EP-1D.02 Prescribing and patient safety

Wednesday

O’Brien, Kirsty

3B.3

Cancer

Thursday pm

O’Brien, Kirsty

2B.5

Generic issues

Thursday am

O’Cathain, Alicia

2B.4

Generic issues

Thursday am

O’Nions, Harriet

EP-2B.04 Education/ethics/governance

Orton, Elizabeth

EP-1F.07

Owen-Smith, Amanda

EP-2C.03 Patient-GP/person centred care/older people

Park, Sophie

1F.3

Perera, Kosala

EP-1A.04 Cancer and palliative care

Petersen, Irene

2E.2

Pinnock, Hilary

EP-1E.05 Diabetes/managing long term conditions/clinical consultation

Pinnock, Hilary

4D.3

Prasad, Vibhore

EP-2A.06 Children/young people/women’s health

Rees, Eliot L

PP1.1

Rhys, Gwydion

EP-2G.08 Pot pourri 2

Friday

Richards, Suzanne

EP-1B.06 Organisation and delivery of primary health care

Wednesday

Richards, Suzanne

EP-1D.03 Prescribing and patient safety

Wednesday

Riches, Nicholas

EP-2C.01 Patient-GP/person centred care/older people

Friday

Rimmer, Melanie

EP-1F.06

Round, Thomas

EP-2C.02 Patient-GP/person centred care/older people

Friday

Rubio, Noah

EP-1E.07 Diabetes/managing long term conditions/clinical consultation

Wednesday

Rushforth, Bruno

EP-1B.01 Organisation and delivery of primary health care

Wednesday

Rushforth, Bruno

EP-2E.04 Multimorbidity/user experience

Friday

Santer, Miriam

EP-2D.04 Information technology/infection/sexual health

Friday

Santos, Andreia C

3C.6

Saunderson, Kathryn

EP-2D.01 Information technology/infection/sexual health

Friday

Seraukina, Tatsiana

EP-1B.02 Organisation and delivery of primary health care

Wednesday

Sheppard, James

EP-1C.01 Cardiovascular

Wednesday

Sheppard, James

3E.4

Shiner, Alice

EP-2E.02 Multimorbidity/user experience

Friday

Shires, Kirsty

EP-2B.05 Education/ethics/governance

Friday

Simmonds, Rosemary

3E.5

Sinnott, Carol

EP-2G.09 Pot pourri 2

Sinnott, Carol

2D.2

Multimorbidity

Thursday am

Smith, Blair

2D.5

Multimorbidity

Thursday am

Smith, Susan

2D.3

Multimorbidity

Thursday am

Spooner, Sharon

EP-2B.09 Education/ethics/governance

Friday

Stanciu, Andrei

EP-1A.01 Cancer and palliative care

Wednesday

Steven, Kathryn

1D.4

Research methods/epidemiology/public health Education Mental health Self-management PRIZE PLENARY

Research methods/epidemiology/public health

Infection

Cardiovascular

Organisation and delivery of primary health care

Friday Wednesday Friday Wednesday Wednesday Thursday am Wednesday Friday Friday Wednesday

Wednesday

Thursday pm

Thursday pm Friday

Wednesday

PRESENTER INDEX Stevens, Sarah

3A.6

Managing long term conditions

Thursday pm

Stewart, Jamie

2A.5

Quality and patient safety

Thursday am

Stirling, Nicola

EP-1A.09 Cancer and palliative care

Wednesday

Stocks, S Jill

EP-1G.10 Pot pourri

Wednesday

Stocks, S Jill

1B.3

Prescribing

Wednesday

Sundvall, Pär-Daniel

3C.3

Infection

Thursday pm

Sundvall, Pär-Daniel

3C.2

Infection

Thursday pm

Taggar, Jaspal

EP-1C.04 Cardiovascular

Wednesday

Taggar, Jaspal

EP-1C.09 Cardiovascular

Wednesday

Tang, Eugene

2E.5

Mental health

Thursday am

Tarrant, Carolyn

1D.1

Organisation and delivery of primary health care

Wednesday

Taylor, Clare J

EP-1C.05 Cardiovascular

Taylor, Steph

2C.3

Musculoskeletal

Thursday am

Taylor, Steph

4D.1

Self-management

Friday

Thomas, Gwilym

3D.3

Population-based practice research

Thursday pm

Thornton, Hannah

EP-2D.06 Information technology/infection/sexual health

Friday

Timblin, Clare (or Bryan)

EP-2A.02 Children/young people/women’s health

Friday

Toleman, Michelle S.

EP-2D.07 Information technology/infection/sexual health

Friday

Tonkin-Crine, Sarah

3A.3

Managing long term conditions

Thursday pm

Torrance, Nicola

1B.2

Prescribing

Wednesday

Tzortziou Brown, Victoria

EP-1G.06 Pot pourri

Valderas, Jose

2A.2

Valderas, Jose

EP-1B.03 Organisation and delivery of primary health care

Valderas, Jose

2A.6

Quality and patient safety

Thursday am

Valderas, Jose

2A.1

Quality and patient safety

Thursday am

Vaz, Luis

3D.4

Population-based practice research

Thursday pm

Vellinga, Akke

2B.1

Generic issues

Thursday am

Vinogradova, Yana

3D.5

Population-based practice research

Thursday pm

Visram, Shelina

EP-1E.06 Diabetes/managing long term conditions/clinical consultation

Wednesday

Wallace, Emma

EP-1F.09

Research methods/epidemiology/public health

Wednesday

Walter, Fiona

EP-1A.02 Cancer and palliative care

Wednesday

Warren, Fiona

EP-1B.04 Organisation and delivery of primary health care

Wednesday

Weston, Charlotte

EP-2E.07 Multimorbidity/user experience

Friday

Williams, Nefyn

EP-2C.09 Patient-GP/person centred care/older people

Friday

Williams, Pauline

EP-1A.07 Cancer and palliative care

Wednesday

Wilson, Patricia

EP-1E.10 Diabetes/managing long term conditions/clinical consultation

Wednesday

Windle, Karen

EP-2C.08 Patient-GP/person centred care/older people

Friday

Wong, Samuel

EP-2F.01

Worrall, Samantha

EP-1A.06 Cancer and palliative care

Wednesday

Wright, Christine

EP-1D.08 Prescribing and patient safety

Wednesday

Wright, Christine

EP-2B.08 Education/ethics/governance

Friday

Wright, Daniel

EP-1E.04 Diabetes/managing long term conditions/clinical consultation

Wednesday

Young, Ben

3B.6

Quality and patient safety

Mental health

Cancer

Wednesday

Wednesday Thursday am Wednesday

Friday

Thursday pm 81

NOTES

MOU

C EU CL HS

IL ME LV

ADO W LN ME

C BU EUCH

L BUCC

A700

O RR TE

1 A7 R LE

E DR

PO T

ID BR

MEL VILL

ON S ICHT

A CH

S ER MB

ATE

COWG

JEFFREY ST

ANC

EDINBURGH

GATE

COW

HIG

H ST

COCKBURN

T ST ARKE

A7 S

N PL LAURISTO

K AR

S AS GR

S NCE PRI

THE

G BRID

MIL

YR OL

YA L

RO THE

CALTON RD

DRIVE

’S ST

OR GE

PLE AS

QUEEN’S

Y AR RDS NA

O LE ST

KS T

A7 00

Dynamic Earth

Playfair Library, Old College

George Square Lecture Theatre

Appleton Tower

SAPC annual conference 2014 venues

Room 2.07

Room 2.11

Room Room Room 2.06 2.05 2.04

Room 2.12

Room 2.14

Lifts

LEVEL 2 FLOOR PLAN

Room M1

Room M3

Speaker Preview

Lifts

LT4

LT5

LEVEL 1 FLOOR PLAN

Main Entrance

Room G.05

Lunch & Refreshments

Registration & Information Lifts

LT1

LT2

LT3

GROUND FLOOR PLAN