EN by Dima Pereli

PERSON-CENTRED CARE TO IMPROVE MDR/RR-TB TREATMENT: ASSESSING A MULTIDISCIPLINARY PSYCHOSOCIAL SUPPORT AND HARM REDUCTION INTERVENTION IN A COHORT OF MDR/RR-TB PATIENTS WITH HARMFUL USE OF ALCOHOL IN MINSK, BELARUS

TABLE OF CONTENTS BACKGROUND TO THE STUDY ��2 OBJECTIVES ��3 OVERALL AIM ��3 SPECIFIC OBJECTIVES: ��3 METHODS ��4 STUDY DESIGN ��4 SETTING ��4 STUDY POPULATION ��4 INCLUSION CRITERIA ��4 RECRUITMENT OF PATIENTS INTO THE PROGRAMME ��5 BRIEF DESCRIPTION OF CARE PROVIDED IN THE PROGRAMME ��5 QUANTITATIVE DATA COLLECTION ��5 QUANTITATIVE DATA ANALYSIS ��6 QUALITATIVE DATA COLLECTION ��6 QUALITATIVE DATA ANALYSIS ��8 RESULTS ��9 QUANTITATIVE RESULTS ��9 DESCRIPTION OF THE MINSK MDR/RR-TB IPD COHORT ��9 BASELINE CHARACTERISTICS OF STUDY PATIENTS �� 10 MENTAL HEALTH FOLLOW UP DATA �� 15 NUMBER AND TYPE OF SESSIONS RECEIVED BY PATIENTS �� 19 TB TREATMENT ADHERENCE �� 20 FACTORS ASSOCIATED WITH ADHERENCE �� 22 TB TREATMENT OUTCOMES �� 24 QUALITATIVE RESULTS �� 25 VULNERABILITY, DESPAIR AND MARGINALISATION: �� 25 SELF-DETERMINATION, GUIDANCE AND ACHIEVING A STATE OF HEALTH �� 29 IDENTIFIED GAPS AND SUGGESTED PROGRAMME IMPROVEMENTS �� 32 LIMITATIONS �� 37 DISCUSSION �� 37 OVERALL ADHERENCE TO TB TREATMENT �� 37 TB TREATMENT OUTCOMES �� 38

PREVALENCE OF ALCOHOL DISORDER AND OTHER VULNERABILITIES �� 38 IMPORTANCE OF CLOSE PERSONAL RELATIONSHIPS �� 38 GAPS IN TREATMENT APPROACHES �� 39 DIFFICULTY IN ASSURING PLANNED FOLLOW UP SESSIONS �� 39 METHOD AND LOCATION OF DOTS �� 40 DIAGNOSIS OF PSYCHIATRIC DISORDERS �� 40 STAFFING LEVELS �� 41 PSYCHOLOGICAL TESTS �� 41 ACCEPTABILITY OF THE PROGRAMME �� 42 SUGGESTIONS FOR IMPROVEMENTS TO THE PROGRAMME �� 42 CONCLUSION �� 44 REFERENCES �� 44 ANNEXES �� 47 ANNEX 1: RISK FACTOR ASSESSMENT �� 47 ANNEX 2 RECRUITMENT OF PARITICIPANTS FOR INTERVIEWS AND FOCUS GROUPS �� 47 ANNEX 3: IN-DEPTH INTERVIEW GUIDE FOR PATIENTS �� 49 ANNEX 4: FOCUS GROUP DISCUSSION GUIDE FOR PRACTITIONERS �� 51 ANNEX 5: ADVERSE EVENTS �� 52 ANNEX 6: PLANNED FREQUENCIES OF COUNSELLING BY BASELINE RISK LEVEL AND TREATMENT PLAN �� 53 ANNEX 7: BOX PLOTS OF ADHERENCE LEVELS BY FACTORS ASSOCIATED WITH ADHERENCE �� 56 ANNEX 8: GRAPHS OF ADHERENCE OVER TIME BY FACTORS ASSOCIATED WITH ADHERENCE �� 59

BACKGROUND TO THE STUDY Tuberculosis remains a global public health problem with an estimated 10 million people affected in 2019 and 1.4 million deaths. Despite decreases in global TB incidence, control of multidrug-resistant/rifampicin-resistant TB (MDR/RRTB) remains challenging, with approximately 206 000 new cases diagnosed in 2019, up from 187 000 in 2018, meaning there are close to half a million active cases globally (World Health Organisation, 2020). The Eastern Europe and Central Asia (EECA) region is particularly affected by high levels of drug resistant amongst those suffering with TB. The incidence of active TB in Belarus is estimated at 29/100 000 (i.e. 4,900 cases) and in 2019, 38% of new and 60 % of previously treated cases being MDR/RR-TB cases. In 2019, there were more than 1,200 (920-1,600) cases of MDR/RR-TB in the country. The incidence of MDR/RR-TB has in general been declining in Belarus over the past few years, but it remains in the list of 30 high MDR-TB burden countries (World Health Organisation, 2020). There is an evident relationship between social determinants of health and tuberculosis (Lönnroth et al., 2009). TB tends to be concentrated in groups with complex health and social issues, for example homelessness, imprisonment, high rates of alcohol and substance misuse, HIV, and lack of entitlement to welfare (Craig et al., 2017). In particular, the risk of active TB is substantially elevated in people who drink more than 40 g of alcohol per day, and/or who have an alcohol use disorder (AUD). Alcohol is considered one of five key risk factors for TB and estimates that 0.72 million cases of TB are attributable to alcohol disorders annually (World Health Organisation, 2020). Furthermore, 2.35 deaths (95% CI 2.05–4.79) per 100 000 people were estimated to be attributable to alcohol and TB in 2014 (Imtiaz et al., 2017). This may be due to both increased risk of infection related to specific social mixing patterns associated with alcohol use, as well as influ-

ence on the immune system of alcohol itself and of alcohol related conditions (Lönnroth et al., 2008). Moreover, AUD increases the risk of poor TB treatment outcomes three-fold. (Alena Skrahina and Hurevich H, Zalutskaya A, Sahalchyk E, Astrauko A, Hoffner S, et al., 2013) In Minsk, risky alcohol intake is identified as one of the main barriers to adherence to TB treatment and is also the main reason for patients being sent to hospital for enforced hospitalisation. Belarus has one of the highest alcohol consumption rates in the world, (World Health Organization, 2018) and in Minsk approximately 40% of MDR/ RR-TB patients present with AUD. The legal system in Belarus can accept the request of the hospital medical council for the involuntary isolation and treatment of patients with poor treatment adherence, which is particularly common in patients with AUD. According to the decree of the Ministry of Health, «Active tuberculosis of the respiratory system with bacterial excretion” is a socially dangerous disease., so if a patient evades his treatment, for example, does not appear for an appointment, misses medications, refuses treatment, then the commission of doctors of the institution sends an application to the court. The court examines the application and the prosecutor gives a sanction for compulsory isolation, which is passed to the police who summon the patient to court, and if they do not appear, then they forcibly bring them to the forced hospitalisation centre. Although involuntary TB treatment in Belarus reduced from 341 patients in 2016 to 188 in 2019 which constitutes an almost 2-fold decrease, involuntarily isolation and treatment remains a concern and better approaches to support adherence to MDR/RR-TB treatment are needed. (Mathew et al., 2006) The literature about psychosocial and harm reduction interventions in MDR/RR-TB patients living with AUD in a complex and difficult social environment in the former Soviet Union

countries is scarce. The experience of integration of alcohol use disorder identification and management in the tuberculosis programme in Tomsk Oblast, Russia, offered a successful model in the management of co-occurring AUDs among patients with chronic medical problems (Mathew et al., 2009). Therefore, the generation of scientific evidence in order to promote changes in TB programmes in this region is needed. This study aims to contribute to the body of literature, by describing MSF’s intervention and by demonstrating its feasibility, acceptability and benefits. In 2014, MSF-OCA opened a project to support the Belarus Ministry of Health to improve MDR/RR-TB treatment adherence and outcomes by providing a psychosocial support and harm reduction intervention for MDR/ RR-TB patients with a person-centred care approach. Person-centred care is a way of delivering safe, evidence-based, holistic, respectful healthcare tailored to the changing needs of different people and communities. It respects an individual’s autonomy to manage their own healthcare choices, based on advice from healthcare professionals (Rogers, 1946). It allows for the delivery of high-quality healthcare through informed decision-making, and is applicable to any healthcare service, whether vertical or integrated into programme design. (Medecins Sans Frontieres, Operational Centre Amsterdam., 2020) The MSF psychosocial support and harm reduction intervention in MDR/RRTB patients in Minsk is a programme that aims to bring the person with MDR/ RR-TB to the centre of the activities. The focus of the project is to improve adherence to MDR/RR-TB treatment and, thereby, to improve TB treatment outcomes. In Minsk, risky alcohol intake is identified as one of the main barriers to adherence in TB treatment and is also the main reason for patients being sent to forced hospitalisation centres. The MSF intervention aims at incorporating multidisciplinary psychosocial support

or discouraging drinking before taking medication. Medical detoxification and anti-craving medications are also prescribed where necessary, but the programme is not designed for treatment/ rehabilitation of alcohol dependence as there are other programmes designed for this in the country.

This mid-term evaluation report describes the interim results of the study, which analyses whether a person-centred multidisciplinary psychosocial support and harm reduction intervention can contribute to better outcomes for MDR/RR-TB patients with a disorder due to the use of alcohol in Minsk.

2. To describe the activities and processes of the comprehensive psychosocial and harm reduction intervention programme from the perspective of the implementation feasibility (this will be done in detail later in 2021) 3. To describe sociodemographic cohort characteristics (age, gender, marital status, alcohol use, recreational drug use, homeless status, unemployment status). 4. To describe clinical characteristics or medical history including co-morbidities, TB treatment details and outcomes. 5. To explore which socio-demographic and clinical characteristics are associated with a better outcome

(outcome measured as adherence). 6. To describe feasibility from the acceptance perspective by exploring patients’ perceptions, acceptance and general experiences of this intervention. 7. To describe TB treatment outcomes at 6, 12 and 18 months according to WHO defined criteria. 8. To determine the prevalence of AUD. 9. To measure the prevalence of mental health disorders amongst patients diagnosed and treated for MDR/RR-TB. 10. To assess changes from baseline to completion of treatment in alcohol use disorders and changes of mental health scores.

STUDY DESIGN

SETTING

The study is a prospective cohort study that uses mixed methods to describe the intervention and the patients within it. The cohort study uses routinely collected data on patients admitted to the cohort from January 2019 to December 2021, though for this interim analysis only patients admitted between 1 January 2019 and 31 May 2020 have been included. The qualitative component includes in-depth interviews with selected patients and health care providers to describe the retrospective (i.e. experienced) acceptability of this intervention from the patient and health provider perspective.

This study took place in Minsk, Belarus, where the estimated population is 2,169,529, and the estimated number of MDR/RR-TB patients starting treatment every year is around 200.

between January 2019 and December 2021 who have a confirmed or suspected disorder due to the use of alcohol (either through diagnosis by a psychiatrist, through the AUDIT or ASSIST tools, or in some cases simply a suspicion from health care workers), though as stated above, for this preliminary analysis we have only included patients admitted up to May 31, 2020.

for AUD as alternatives to forced hospitalisation and other punitive measures for MDR/RR-TB patients. These activities and processes give recommendations to MDR/RR-TB patients with AUD, or at risk of AUD, to support and promote the reduction of their risky alcohol intake, by promoting alcohol free days

OBJECTIVES OVERALL AIM To assess feasibility, acceptance and the potential benefits (secular improvements in treatment adherence, TB treatment outcomes and patient wellbeing) of a person-centred multidisciplinary psychosocial support and harm reduction intervention in patients with AUD and MDR/RR-TB patients in Minsk.

SPECIFIC OBJECTIVES: 1. To describe tuberculosis treatment adherence in a cohort of MDR/RRTB patients with AUD in Minsk city within the comprehensive psychosocial and harm reduction intervention programme

METHODS

STUDY POPULATION Patients admitted to the inpatient department of the Republican Scientific and Practical Centre for Pulmonology and TB (RSPCPT) were eligible to be recruited to the study under the following criteria: all newly diagnosed MDR/ RR-TB adult patients who are residents of Minsk city and starting treatment

INCLUSION CRITERIA Patients were included into the study under the following conditions: • Starting treatment for MDR/RR-TB (pulmonary TB) between January

• • • •

01 2019 and May 31, 2020 Aged over 18 years of age at treatment commencement Living in Minsk city Confirmed or suspected disorder due to the use of alcohol Provision of written informed consent to participate in the study, amongst patients still in the cohort when the study received approval to begin (January 2020)

RECRUITMENT OF PATIENTS INTO THE PROGRAMME All newly diagnosed MDR/RR-TB patients from Minsk are admitted to the inpatient department and they were contacted by the MSF team within eight working days after admission to do an initial assessment. The purpose of the first session of psychosocial assessment was to evaluate the need for psychosocial support among MDR/RR-TB patients. Basic information of the patient including socioeconomic situation (income, employment, living condition) was collected and level of anxiety and depression are screened by using the Generalised Anxiety Disorder Assessment (GAD-7) and the Patient Health Questionnaire (PHQ-9) scoring systems. Suspected alcohol use disorder (AUD) was assessed by well-trained and experienced counsellors. During this first screening the counsellor used GAD-7 and PHQ-9, but she or he did not use any specific tool or scoring system to identify the suspected AUD. They based their assessment on conversations held during the first psychosocial sessions and their expert experience. If the counsellor suspected the patient has harmful use of alcohol or an alcohol disorder according to the ICD10 criteria, the second session of counselling (risk assessment) is scheduled within eight working days of the first session. Pre-screen and audit questionnaires were used in this second session. If the patient answered “Yes” to all three pre-screen questions, the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) tool was then used for

multiple substance use assessment. If the patient drinks only alcohol within one year, AUDIT C (the first 3 questions of AUDIT) was continued. If AUDIT C was ≥3 in women or ≥4 in men, patient was interviewed by asking the remaining questions of AUDIT tool.

BRIEF DESCRIPTION OF CARE PROVIDED IN THE PROGRAMME The risk to poor adherence is ranked by scores of AUDIT or ASSIST tool, combined with other health and socioeconomic risk factors. Low risk patients receive the baseline intervention and moderate and high-risk patients are provided with a briefing and harm reduction packages respectively. The scoring used is provided in Annex 1. Furthermore, some patients may need referral for medical detoxification where clinically indicated. Anti-craving medication prescriptions of Naltrexone was approved by the US Food and Drug Administration (FDA) in 1994 (Bello, 2019), and has since been approved in 30 countries worldwide. The use of naltrexone for patients with moderate and severe alcohol use disorder (together with medical management or psychological interventions) is, among others, recommended by the National Institute of Health (USA), the United Kingdom’s National Institute for Health and Care Excellence (NICE, 2011) the British Association for Psychopharmacology (Lingford-Hughes et al., 2012) , the American Psychiatrists Association (Vi et al., 2018), and the World Health Organization (Dua et al., 2011). Guidance from these authorities is that the use of naltrexone as an adjunct in the management of AUD can be safely done at a primary health care level. Referral and management of other co-morbidities/ co-infections are also included in this intervention. Patients with these conditions are referred to relevant MoH departments.

QUANTITATIVE DATA COLLECTION The data used for this study are currently collected as part of the monitoring activities of the MSF interven-

tion. Patients enrolled since January 2020 were invited to give consent to take part in the study. Three patients refused for their information to be included in the quantitative component. There are two main sources of data: the psychosocial intervention monitoring tool in Excel and the MDR/RRTB treatment and follow up data base (Bahmni data base). Forms for the Bahmni register are completed everyday by clinicians and collated weekly. The psychosocial database is completed by counsellor-educators, a psychiatrist, a psychologist, and a social worker. When meeting with patients, the Mental Health workers completes a patient assessment at the first meeting and then at subsequent visits follow up forms are filled at 6 month intervals. Responses are written on a hard copy of the corresponding questionnaire. After review in the field, hard copies of the questionnaires are entered into the psychosocial database in the MSF Minsk office. For clinical information on TB treatment medical doctors and nurses complete clinical follow up forms and the data are entered in the Bahmni data base. Data verification and cleaning programmes are run each month to identify errors in the data collection. When errors are found, the forms will be sent back to the medical staff for correction of errors. Data was then further cleaned in Excel and R Studio for the purposes of this study and errors were verified with the data entry and clinical teams. The Excel based psychosocial database includes sociodemographic information, and baseline and follow up scores for the screening tests (PHQ9, GAD-7, AUDIT, ASSIST and Patients self-motivation score) and patient outcomes. The follow up tests are collected at 6, 12, 18 and 24 months. The Bahmni database included information on treatment adherence, co-morbidities, type of TB treatment, adverse events, previous TB treatment, mode of DOTS, drug susceptibility testing (DST), presence of cavities on x-Ray, baseline culture, sputum smear status at baseline (and monthly thereafter) and TB treatment outcome.

QUANTITATIVE DATA ANALYSIS Descriptive analysis of all variables (socio demographic, baselines and outcome variables) was carried out. Patients were also described in terms of being adherent or non-adherent. Adherent was defined as having taken >90% of the prescribed doses, ascertained through visits with a nurse or through video observed treatment, over the whole course of treatment to date, though in this analysis it should be noted that not all patients have finished treatment. Some patients filled out an AUDIT form, whilst others filled out an ASSIST form, dependent on whether they were using drugs other than alcohol. In the analysis the AUDIT and ASSIST results are presented in some cases separately, and in some cases, they are presented together. Proportions, means, standard deviations, and their respective 95% confidence intervals were provided. For non-normal continuous variables medians and interquartile ranges were provided. Chi squared tests, t-tests and Kruskal Wallis tests were calculated as appropriate to test for statistical significance at the 0.05 level. Linear regression and least squares regression were used to plot lines of best fit on graphs of adherence. For the analysis of mental health scores over time, box plots and kernel distributions are displayed at baseline, 6 months, 12 months and 18 months. At this point in time, given the small amount of data, and small differences in patient groups, we present the risk of adherence <90% by key variables, their univariate odds ratios, and their confidence intervals calculated using logistic regression. Multivariate analyses to disentangle associations with outcomes have not been carried out. Lost to follow up (LTFU) was defined according to the WHO definitions whereby patients are considered LFTU or if their treatment was interrupted for 2 consecutive months or more i.e. if they do not attend counselling or DOT sessions after two months of attempts to organise a session by the team. For adherence any patient who was

TABLE 1: PARTICIPANT CHARACTERISTICS FOR THE QUALITATIVE COMPONENT Overall (N=12) N (%)

In-depth interview with patients Gender - Female

3 (25.0%)

- Male

9 (75.0%)

- In union

4 (33.3%)

- Single

5 (41.7%)

- Widowed/divorced/Separated

3 (25.0%)

- Employed

3 (25.0%)

- Unemployed

9 (75.0%)

Marital Status

Employment status

Опыт пребывания в МЛС - Yes

9 (75.0)

- No

3 (25.0)

- Confirmed MDR

3 (25.0)

- Confirmed pre-XDR (FQ)

5 (41.7)

- Confirmed pre-XDR (Inj)

2 (16.7)

- Confirmed XDR

2 (16.7)

- Yes

4 (33.3)

- No

8 (66.7)

- >=90% Adherence

9 (75.0)

- <90% adherent

3 (25.0)

- Yes

3 (25.0)

- No

9 (75.0)

- Alcohol dependence

6 (50.0)

- Opioid dependence

1 (8.3)

- Acute alcohol intoxication

2 (16.7)

- Personality disorder

1 (8.3)

- Mild intellectual disabilities

1 (8.3)

TB sub-classification

Hepatitis C status

Percent with adherence above 90%

Previous history of TB

Psychiatric diagnosis

Baseline ASSIST or AUDIT score - Low

2 (18.2)

- Moderate

3 (27.3)

- High

6 (54.5)

- Moderate risk

5 (41.7)

- High risk

7 (58.3)

Baseline Risk for non-adherence

LTFU were automatically classified as non-adherent, no matter the level of adherence they achieved before they were LTFU. Patients who move out of Minsk and continue TB treatment elsewhere and report so to the team are classified as “not evaluated” by the intervention team, but are included in this analysis and their outcomes were confirmed to be either completed treatment or still on treatment at the time of analysis. For many of these patients, they are transferred to Sosnovka hospital.

QUALITATIVE DATA COLLECTION In-depth interviews were conducted with patients, and focus group discussions were then conducted with practitioners in order to understand the retrospective acceptability of the intervention. Using the preliminary results of the quantitative analysis, a subgroup of participants was purposively selected, to explore acceptance of the programme and experiences of different groups of patients. Patients were selected to ensure that the sample included the full range of patients - based on their gender, level of adherence, psychiatric diagnosis, reported test scores in the GAD7, PHQ9, ASSIST and AUDIT scores, current treatment status (IPD, OPD or completed) and willingness to participate in an interview. Further details can be seen in Annex 2. Initial candidates who were likely to be willing and open to talk were suggested by counsellor-educators, then further patients were selected to ensure relevant patient characteristics were represented. In some cases, patients too unwell on medical advice were not included. Counsellor-educators then offered the opportunity for patients to volunteer to consent to talk to the interviewer and translator, either in person or over the phone. Furthermore, individual counselling/support with counsellor educators were offered to patients after the interview, if they wanted emotional support. Verbal consent was given via counsellor educators. Due to COVID restrictions, a minority of patients only gave verbal consent over the phone, in which case, the counsellor has written

Overall (N=12) N (%)

In-depth interview with patients Treatment location at time of interview - Forced Hospitalisation

2 (18.2)

- IPD

2 (18.2)

- Completed treatment

1 (18.2)

- OPD

6 (58.3)

- LTFU

1 (8.3) Overall (N=20) N (%)

Focus groups with practitioners Gender - Female

13 (65.0)

- Male

7 (35.0)

- MSF

11 (47.8)

- RSPCPT

6 (26.1)

- TB3

3 (13.0)

- Counsellor-educator

5 (21.7)

- Nurse

4 (17.4)

- Psychiatrist

2 (8.7)

- Psychologist

1 (4.4)

- Social Worker

1 (4.4)

Employer

Employment

- TB Doctor this in the consent form until it may be possible to get a written signature when COVID-19 restrictions are reduced. Due to the COVID-19 restrictions 9 of the 12 interviews were done over the phone rather than in person. For the 3 in person interviews, these were conducted in the private space in the IPD or in the patient’s own room, using PPE measures for MDR/RR TB. Patients experience of MDR-TB treatment and their time in the programme were explored using participant led open question techniques. See topic guide, annex 3. A second subgroup of participants was purposively selected to explore experience and attitudes based on a collectivist, multi-vocal method of focus group discussion. This gave additional insight that the group interaction yields (Stewart, David W,

10 (43.5)

Shamdasani, Prem N., and Rook, Dennis:, 2015). Feedback and follow up questions were asked to participants of focus groups, by the investigators in an iterative approach to refine and improve upon the data collected. An information and consent paper outlining the purpose of the research was shared with all participants. Voluntary consent was completed individually before the discussion. All focus group participants volunteered to consent to audio recording the discussions and agreed to respect confidentiality. For both in-depth interviews and focus group discussions, English translated verbatim transcripts of the audio recording were prepared for analysis. The focus group discussions with practitioners were conducted online (via skype) due to COVID pandemic

restrictions as per independent ethical evaluation. In some cases, participants were together in the same room, wearing masks, and socially distanced by 2 metres. See the topic guide in annex 4

QUALITATIVE DATA ANALYSIS For qualitative data, transcripts were analysed thematically by two co investigators, using an approach that aims to identify and explain patterns in the data. Negative cases (i.e. data that challenges the emerging analysis) were examined in order to test emerging themes and to explain why these cases

are different (Bradley, Curry and Devers, 2007). Field notes made throughout the fieldwork period were used to guide data analysis. This included a coding dictionary and analytic memos. The two separate analyses were discussed and scrutinised until a final analytical framework and coding structure was created. Results and analysis were also presented to and scrutinised by focus group participants as part of the iterative approach to data collection and analysis. This process was not possible for patient participants due to fears of accidental breach of confidentiality, that could lead to further feelings of isolation and prejudice by patients. Selected

anonymised interview excerpts were drawn out to ensure the individual ‘stories’ are not lost and to explore how the themes interrelate in particular cases (Bradley, Curry and Devers, 2007). The qualitative data was sorted and coded manually using Excel and Word. An additional level of verification will be made through sharing preliminary analysis with participants for commentary and their views will be included in final write up.

ETHICS Ethical approval for this study was received from the MSF ERB and from the ethics committee at the RSPCPCT.

RESULTS: QUANTITATIVE RESULTS

2000 1500 1000

500 0

2342

1955 68%70% 64% 62% 60% 1568 1486 55% 50%

1979

55%

80%

40% 30% 20% 10% 0%

2013 2014 2015 2016 2017 Year of treatment commencement Figure 1a: Yearly outcomes of patients in the Minsk MOH MDR/RR-TB cohort

% of patients

At the time this data analysis (October 2020) was conducted there were 50 patients in the Minsk In Patient Department (IPD) with MDR/ RR-TB, of which 39 (78%) were male and 11 (22%) were female. Amongst males 20 (51%) had a substance use disorder and amongst females this was 1 (10%), so overall 42% (n=21) were estimated to have a substance use disorder. Of these 21 patients, 9 are in the MSF person-centred multidisciplinary programme, the other patients with alcohol problems were not in the MSF programme as they are not registered as living in Minsk so could be followed once they leave the IPD. During the study period (1st January 2019 to 30th November 2020) there were 143 newly diagnosed MDR/RRTB patients in Minsk suggested by MOH staff for assessment for entry into the programme, and of these, 122 (83%) went for assessment. The median number of days between assessment and first session in the cohort was 3 days (IQR 2-5).

2500

Number of patients

DESCRIPTION OF THE MINSK MDR/RR-TB IPD COHORT

Total cases % Success % Failure % LTFU % Died

Figure 1a below shows the patient outcomes in the overall MDR/RR-TB cohort in Minsk for the last years. From 2013 to 2018 there is a decrease in the number of patients enrolled into care and an increase in success rates - which is accompanied by a decrease in the treatment failure rate. LTFU and death rates remain fairly steady at around 10%.

BASELINE CHARACTERISTICS OF STUDY PATIENTS There were 59 patients included from the Minsk PS MDR/RR-TB intervention in this analysis, of which 8 (13.6%) were female and 51 (86.4%) were male. The majority of patients had a history of either past incarceration 41 (69.5%), homelessness 8 (13.6%) or unemploy-

ment 37 (62.7%). Amongst males, 21 (41.2%) were either married or living with a partner and the same figure for females was 5 (62.5%). Amongst patients for whom data was available, 24 i.e. 100% were recorded as being smokers. The full baseline social characteristics, stratified by whether the patient was adherent to tuberculosis treatment or not can be seen in Table 2.

TABLE 2: BASELINE SOCIAL CHARACTERISTICS BY ADHERENCE STATUS >=90% adherence (N=38) N (%)

<90% adherent (N=21) N (%)

Overall (N=59) N (%)

Gender

p value 0.142 (1)

- Female

7 (87.5)

1 (12.5)

8 (13.6)

- Male

31 (60.8)

20 (39.2)

51 (86.4)

Age group

0.106 (1) - <35

5 (62.5)

3 (37.5)

8 (13.6)

- 35-55

26 (74.3)

9 (25.7)

35 (59.3)

- >55

7 (43.8)

9 (56.2)

16 (27.1)

Marital Status

0.084 (1) - In union

16 (61.5)

10 (38.5)

26 (44.1)

- Single

13 (86.7)

2 (13.3)

15 (25.4)

Widowed/divorced/Separated

9 (50.0)

18 (30.5)

Education level

0.739 (1) - Did not finish school

1 (50.0)

2 (3.4)

- Secondary

35 (66.0)

18 (34.0)

53 (89.8)

- University

2 (50.0)

4 (6.8)

Employment status

0.924 (1) - Employed

14 (63.6)

8 (36.4)

22 (37.3)

- Unemployed

24 (64.9)

13 (35.1)

37 (62.7)

History of Incarceration

0.810 (1) - Yes

26 (63.4)

15 (36.6)

41 (69.5)

- No

12 (66.7)

6 (33.3)

18 (30.5)

Homeless status

0.501 (1) - Homeless

6 (75.0)

2 (25.0)

8 (13.6)

- Not homeless

32 (62.7)

19 (37.3)

51 (86.4)

- Yes

14 (58.3)

10 (41.7)

24 (100.0)

- No

0 (0)

Smoker

IV illicit drug use

0.416 (1) - Yes

1 (33.3)

2 (66.7)

3 (23.1)

- No

6 (60.0)

4 (40.0)

10 (76.9)

>=90% adherence (N=38) N (%)

<90% adherent (N=21) N (%)

Overall (N=59) N (%)

Non-IV illicit drug use

p value 0.505 (1)

- Yes

1 (33.3)

2 (66.7)

3 (25.0)

- No

5 (55.6)

4 (44.4)

9 (75.0)

Discharge from Orsha Prison

0.191 (1) - Yes

9 (81.8)

2 (18.2)

11 (19.3)

- No

28 (60.9)

18 (39.1)

46 (80.7)

Discharge from Volkovichi Forced Hospitalisation Centre (FHC)

0.646 (1) - Yes

4 (57.1)

3 (42.9)

7 (12.3)

- No

33 (66.0)

17 (34.0)

50 (87.7)

(1) Pearson’s Chi-squared test Table 3 shows the baseline medical characteristics stratified by whether or not they were adherent to TB treatment. Amongst the MSF patients, adverse events (AE) were relatively common with 6 (23.1%) having one or two adverse events and 16 (61.5%) having 3 or more adverse events. One patient had a total of 9 adverse events and 5 others had serious adverse events recorded. Annex 5 describes the type of adverse events experienced. In the cohort as a whole the most commonly recorded co-morbidity was Hepatitis C (n = 25, 42.4%), followed by HIV (n = 18, 30.5%) then heart disease (n = 12, 20.3%). TABLE 3: BASELINE MEDICAL CHARACTERISTICS BY ADHERENCE STATUS >=90% Adherence (N=38) N (%)

<90% adherent (N=21) N (%)

Overall (N=59) N (%)

Drug regimen

p value 0.968 (1)

- BDQ

21 (65.6)

11 (34.4)

32 (54.2)

- DLM and BDQ

15 (62.5)

9 (37.5)

24 (40.7)

- NA

2 (66.7)

1 (33.3)

3 (5.1)

Regimen length

0.852 (1) - Long (18-20m) regimen

28 (65.1)

15 (34.9)

43 (100.0)

- Short regimen

10 (62.5)

6 (37.5)

16 (100.0)

WHO Treatment group

0.437 (1) - New

19 (55.9)

15 (44.1)

34 (58.6)

- Other previously treated patients

3 (100.0)

0 (0.0)

3 (5.2)

- Relapse

10 (76.9)

3 (23.1)

13 (22.4)

- Treatment After Failure

1 (50.0)

2 (3.4)

- Treatment after loss to follow up

4 (66.7)

2 (33.3)

6 (10.3)

TB sub-classification

0.699 (1) - Confirmed MDR

11 (78.6)

3 (21.4)

14 (23.7)

- Confirmed pre-XDR (FQ)

7 (53.8)

6 (46.2)

13 (22.0)

- Confirmed pre-XDR (Inj)

10 (66.7)

5 (33.3)

15 (25.4)

- Confirmed XDR

9 (60.0)

6 (40.0)

15 (25.4)

- R resistance with H susceptibility

1 (50.0)

2 (3.4)

>=90% Adherence (N=38) N (%)

<90% adherent (N=21) N (%)

Overall (N=59) N (%)

Previous history of TB

0.111 (1) - Yes

19 (76.0)

6 (24.0)

25 (42.4)

- No

19 (55.9)

15 (44.1)

34 (57.6)

HIV status

0.902 (1) - Positive

12 (66.7)

6 (33.3)

18 (31.0)

- Negative

26 (65.0)

14 (35.0)

40 (69.0)

Hepatitis C status

0.247 (1) - Yes

14 (56.0)

11 (44.0)

25 (42.4)

- No

24 (70.6)

10 (29.4)

34 (57.6)

Hepatitis B status

0.665 (1) - Yes

1 (50.0)

2 (3.4)

- No

37 (64.9)

20 (35.1)

57 (96.6)

Diabetes status

0.665 (1) - Yes

1 (50.0)

2 (3.4)

- No

37 (64.9)

20 (35.1)

57 (96.6)

COPD

0.175 (1) - Yes

0 (0.0)

1 (100.0)

1 (1.7)

- No

38 (65.5)

20 (34.5)

58 (98.3)

Cirrhosis

0.624 (1) - Yes

3 (75.0)

1 (25.0)

4 (7.3)

- No

32 (62.7)

19 (37.3)

51 (92.7)

Renal illness

0.533 (1) - Yes

2 (50.0)

4 (6.8)

- No

36 (65.5)

19 (34.5)

55 (93.2)

Heart disease

0.925 (1) - Yes

8 (66.7)

4 (33.3)

12 (20.7)

- No

30 (65.2)

16 (34.8)

46 (79.3)

BMI

0.498 (1) - Underweight (<18.5)

1 (33.3)

2 (66.7)

3 (12.0)

- Normal weight (18.5 to 25)

13 (68.4)

6 (31.6)

19 (76.0)

- Overweight (25 to 30)

2 (66.7)

1 (33.3)

3 (12.0)

No. Adverse events (AE)

0.044 (1) - 1-2 AE

4 (66.7)

2 (33.3)

6 (10.2)

- 3+ AE

12 (75.0)

4 (25.0)

16 (27.1)

- No AE

0 (0.0)

4 (100.0)

4 (6.8)

22 (66.7)

11 (33.3)

33 (55.9)

- Unknown Serious adverse events

(1) Pearson’s Chi-squared test

p value

0.090 (1) - 1-2 SAE

1 (20.0)

4 (80.0)

5 (8.5)

- No SAE

15 (71.4)

6 (28.6)

21 (35.6)

- Unknown

22 (66.7)

11 (33.3)

33 (55.9)

Table 4 shows the baseline mental health characteristics. There were 24 (40.7%) patients who had been registered as having a disorder due to the use of drugs or alcohol and 5 (8.5%) who had been previously diagnosed with mental illness before cohort entry. This meant just over half had no formal psychiatric diagnosis before entry into the psychosocial support programme. At cohort entry all patients had a psychiatric diagnosis of some sort recorded. The most common type of diagnosis was a disorder due to the use of alcohol with 45 (76%) patients having a confirmed or suspected diagnosis of this at admission. Among the remaining 14 with no formal diagnosis there was only 1 who had a low score according to the AUDIT or ASSIST tools, but the patient was suspected by their counsellor to have problematic alcohol consumption. There were also a further 5 patients with opioid dependence, and amongst those with alcohol dependence 6 patients had a combination of alcohol and

other drug dependence. The full breakdown of types of psychiatric disorder can be seen in Table 4. Only 3 patients are recorded as having a personality disorder diagnosis, though there are thought to be other patients who would meet the criteria for this diagnosis if the psychiatrist had gone through the full process of diagnosing it. It should also be noted that any patient with a more severe psychiatric disorder such as psychosis would be referred to another facility rather than treated in the cohort. When combining the results for the ASSIST and AUDIT scores 30 (50.8%) had moderate alcohol or other substance use, 12 (20.3%) had high levels of use and 15 (25.4%) self-report low use of substances. Self-reported alcohol use amongst women was considerably lower with 2 out of 8 women (25%) having moderate or high levels of alcohol use, compared to 40 out of 51 (78.4%) among males. Baseline reported IV drug use, for patients where data was available, was 1 out of 3 (33.3%)

females and 2 out of 23 (8.7%) males. Males however were more likely to have hepatitis C with 24 out of 51 (47.1%) compared to 1 out of 8 females (12.5%) which may indicate higher levels of past drug use. Levels of anxiety at baseline were generally low, with 48 patients (81.4%) reporting minimal levels of anxiety. Levels of anxiety among women were higher with 2 of 8 (25%) having moderate or severe anxiety, compared to 3 of 51 (5.9%) males. The PHQ9 showed that 8 patients (13.6%) had moderate depression and 4 (6.8%) had moderately severe or severe depression. Levels of depression among women were similar to men with 2 of 8 (25%) women having moderate or severe depression, compared to 10 of 51 (19.6%) males. Patient’s baseline motivation to complete treatment was variable with 25 (42.4%) reporting a high level of motivation and 5 (8.5%) a low level of motivation.

TABLE 4: BASELINE MENTAL HEALTH CHARACTERISTICS >=90% Adherence (N=38) N (%)

<90% adherent (N=21) N (%)

Overall (N=59) N (%)

History of MH Issues

p value 0.233 (1)

- Yes

2 (40.0)

3 (60.0)

5 (8.5)

- No

36 (66.7)

18 (33.3)

54 (91.5)

MH Registered at narcological dispensary

0.349 (1)

- Yes

20 (58.8)

14 (41.2)

34 (58.6)

- No

17 (70.8)

7 (29.2)

24 (41.4)

Other psychological illness

0.360 (1) - Yes

34 (66.7)

17 (33.3)

51 (86.4)

- No

4 (50.0)

8 (13.6)

Substance abuse diagnosis

0.271 (2)

- Alcohol dependence

23 (67.6)

11 (32.4)

34 (57.6)

- Alcohol and other substance dependence

2 (33.3)

4 (66.7)

6 (10.2)

- Opioid dependence

5 (100.0)

0 (0.0)

5 (8.5)

- Acute alcohol intoxication

2 (66.7)

1 (33.3)

3 (5.1)

- Alcohol amnesic syndrome and alcohol dependence

1 (50.0)

2 (3.4)

- None recorded

5 (55.6)

4 (44.4)

9 (15.3)

>=90% Adherence (N=38) N (%)

<90% adherent (N=21) N (%)

Overall (N=59) N (%)

Other psychiatric diagnosis

0.482 (2)

- Borderline personality disorder

2 (100.0)

0 (0.0)

2 (3.4)

- Other personality/behavioural disorder due to physical cond.

1 (100.0)

0 (0.0)

1 (1.7)

0 (0.0)

1 (100.0)

1 (1.7)

- Toximetabolic encephalopathy

1 (100.0)

0 (0.0)

1 (1.7)

- Reaction to stress

1 (100.0)

0 (0.0)

1 (1.7)

0 (0.0)

1 (100.0)

1 (1.7)

33 (63.5)

19 (36.5)

52 (88.1)

- Other MH disorder due to phys cond.

- Mild intellectual disabilities - None recorded Baseline PHQ9

0.825 (1) - <4 Minimal

24 (66.7)

12 (33.3)

36 (62.1)

- 5-9 Mild

6 (60.0)

4 (40.0)

10 (17.2)

- 10-14 Moderate

6 (75.0)

2 (25.0)

8 (13.8)

- >=15 Moderately severe/severe

2 (50.0)

4 (6.9)

Baseline GAD7

0.831 (1) - Minimal

32 (66.7)

16 (33.3)

48 (85.7)

- Mild

2 (66.7)

1 (33.3)

3 (5.4)

- Moderate or severe

4 (80.0)

1 (20.0)

5 (8.9)

Baseline ASSIST or AUDIT score

0.866 (1) - Low

10 (66.7)

5 (33.3)

15 (26.3)

- Moderate

20 (66.7)

10 (33.3)

30 (52.6)

- High

7 (58.3)

5 (41.7)

12 (21.1)

Baseline risk for adherence issues

0.850 (1)

- Low risk

2 (66.7)

1 (33.3)

3 (5.1)

- Moderate risk

26 (66.7)

13 (33.3)

39 (66.1)

- High risk

10 (58.8)

7 (41.2)

17 (28.8)

Baseline Self-Motivation

0.904 (1) - Low

3 (60.0)

2 (40.0)

5 (8.8)

- Moderate

17 (63.0)

10 (37.0)

27 (47.4)

- High

17 (68.0)

8 (32.0)

25 (43.9)

(1) Pearson’s Chi-squared test

MENTAL HEALTH FOLLOW UP DATA Figures 2 shows the baseline mental health scores for each test, and 2a to 2f shows the patients screening scores over time. The red box displays the median and interquartile range, and the grey shapes display a kernel distribution of the data. AUDIT and ASSIST scores stayed fairly steady, with a

p value

(2) Fisher’s exact test median of 9 and 9 at baseline then a slightly higher median of 9.5 and 13.5 at six months and then 4.5 and 15 at 12 months. There is very little data at 18 months thus far. There was a considerable amount of missing data, with only 25 (56.8 %) people having AUDIT or ASSIST scores recorded at 6 months amongst those on treatment at 6 months, and 12(63.2 %) at 12 months

of those on treatment at 12 months. Of the 12 with high AUDIT or ASSIST scores at baseline, only 3 have a follow up score at 12 months. Anxiety and depression scores measured with the PHQ9 and GAD7 tools had a median of 2 and 3 at baseline then 2 and 4.5 at six months and 1 and 2 at 12 months. Similarly, there was a considerable amount of missing data.

NUMBER AND TYPE OF SESSIONS RECEIVED BY PATIENTS Table 5 shows the median and interquartile range of the number of follow up sessions per week, stratified by baseline risk group. The median total number of sessions per patient with either a social worker, psychiatrist, or counsellor-educator was 49.5 sessions (IQR: 26.5 - 64.8) which equates to 1.2 (IQR: 0.9 - 2.1) per week. The 3 patients with low risk had 1.2 (IQR: 0.9 - 1.4) per week whereas the 39 patients at moderate risk had 1.2 (IQR: 0.9 - 1.8) per week. The 17 patients at high risk had 1.3 (IQR: 0.7 - 2.4) per week.

The majority of sessions were with counsellor-educators. Patients had a median of 10 (IQR: 2.5 - 16.5) follow up counselling sessions and 22 (IQR: 12 - 29) short contact sessions which equates to 0.3 (IQR: 0.1 - 0.5) and 0.5 (IQR: 0.4 - 0.8) sessions with a counsellor-educator per week, respectively. See Annex 6 for full details on the number of patient sessions scheduled per week according to MSF protocols to be able to compare in full, but for high and moderate risk groups the realised sessions were considerably fewer than planned sessions. Furthermore, 14 (23.7%) patients attended group counselling, 14 (23.7%) patients received home visits and 13 (22%) received at least one family intervention.

Fourteen (23.7%) were seen by the social worker, and the median number of sessions with the social worker, amongst those referred was 11 (IQR: 8.2- 13.8) which equates to a total of 0.3 (IQR: 0.2- 0.4) per week. Twenty-four (40.7%) were referred to a psychiatrist, of which 15 (25.4%) were prescribed psychiatric drugs (for alcohol withdrawal, sleeping tablets, anti-convulsant, anti-craving medication or vitamins) and amongst these the median number of sessions with the psychiatrist was 4 (IQR: 1.8- 19.2) . The three patients in the low risk group had a median 10 (IQR: 5.0, 18.0) sessions with the psychiatrist, more than the high and moderate risk groups.

TABLE 5: NUMBER OF FOLLOW UP VISITS BY BASELINE RISK GROUP Low risk (N=3)

Moderate risk (N=39)

High risk (N=17)

Total (N=59)

Individual FU sessions per week

0,663 (1) - Median

0,3

0,2

0,3

- Q1, Q3

0,1 0,3

0,2, 0,4

0,1, 0,6

0,1, 0,5

Total individual FU sessions

p value

0,528 (1) - Median

14,0

11,0

6,0

10,0

- Q1, Q3

7,0, 15,0

3,0, 20,5

2,0, 13,0

2,5, 16,5

Low risk (N=3)

Moderate risk (N=39)

High risk (N=17)

Total (N=59)

Patient contacts per week

p value 0,811 (1)

- Median

0,5

0,6

0,5

- Q1, Q3

0,4, 0,6

0,5, 0,7

0,4, 0,8

Total patient contacts

0,784 (1) - Median

24,0

20,0

22,0

- Q1, Q3

17,5, 28,0

12,5, 33,0

12,0, 26,0

12,0, 29,0

Total psychiatrist sessions

0,295 (1) - Median

10,0

1,0

2,5

1,0

- Q1, Q3

5,0, 18,0

0,0, 3,0

0,0, 12,2

0,0, 5,8

Total social worker sessions

0,151 (1) - Median

0,0

7,0

1,0

3,0

- Q1, Q3

0,0, 6,5

0,0, 11,5

0,0, 4,0

0,0, 11,0

Total MH sessions per week

0,854 (1) - Median

1,0

1,2

1,3

1,2

- Q1, Q3

0,9, 1,4

0,9, 1,8

0,7, 2,4

0,9, 2,1

Total MH sessions

0,529 (1) - Median

59,0

52,0

37,0

49,5

- Q1, Q3

42,5, 64,0

29,0, 66,5

25,8, 50,8

26,5, 64,8

(1) Kruskal-Wallis rank sum test

TB TREATMENT ADHERENCE Overall, 38 (64.4%) of patients took at least 90% of their treatment in the programme to date. If the 12 patients who are LTFU are excluded from this then 38 out of 47 (80.9%) are at least 90% adherent; 44 (77.6%) of patients

in total had adherence above 80%, and excluding those lost to follow up this becomes 44 out of 47 (93.6%). The median overall adherence for all patients in the programme was 94.9 (IQR: 8198.1). For patients who were not LTFU this was 96.4 (IQR: 93.2- 98.8) and was 64.1 (IQR: 44.9- 73.2) for patients who

are LTFU. It should be noted that this is not the final adherence as 19 (32.2%) of the patients have not yet had a treatment outcome. Adherence by trimester can be seen in table 6 and figures 3a to 3c show graphical representations of adherence by treatment month, semester and also by calendar month.

TABLE 6: ADHERENCE BY TREATMENT TRIMESTER Time since start of treatment

Number patients

Median adherence

Lower quartile

Upper quartile

Maximum adherence

Minimum adherence

Month 1-3

100,0

98,7

100

Month 4-6

97,1

88,2

100

Month 7-9

97,7

91,7

100

Month 10-12

100,0

96,7

100

Month 13-15

98,8

91,4

100

Month 16-18

100,0

89,3

100

Month 19-21

100,0

96,2

100

Figures 3a to 3d show the % adherence to treatment of patients on the y axis, and month of treatment on the x axis. In graph 3a, 3c and

3d the x axis shows month since start of TB treatment. For graph 3b it shows average adherence of all patients in the study by calendar

month of treatment. The lines on the graphs show the line of best fit, calculated with a least squares’ method in most cases, but a linear

trend is also shown in Figure 3a. For figure 3b, there is a slight dip

towards the fourth quarter of 2019 and again in the second half of

2020, the latter is probably related to COVID-19.

Patients received DOTS in a number of different locations or means, and this was noted for each day of treatment. This could be in-patient, out-patient facility, home-visit, or video DOT. The above graph shows the average adherence per month by type of DOT. Most of the treatment received by patients was at an in-patient facility. To date, of the 624 patient/months included in this analysis 355 (56.9%) were at an in-patient facility. This means that between the 59 patients they have passed in total 624 months of treatment, and between them just over half was spent in an in-patient facility. Sixteen (27%) patients have only been treated at an in-patient facility to date and 7 (12%) patients have been in forced hospitalisation. Median adherence during the in-patient period was extremely good, at 100% (IQR: 100-100%). After discharge from IPD, patients in the study received a total of 81 (13%) months of their treatment via DOT administered by video observed treatment. The median adherence at video DOT sessions was 100% (IQR: 93.3-100%) compared to 92.9% (IQR: 63.1-100%, n= 155) for patients visiting an out-patient facility for DOT and 100% (IQR: 100-100%, n= 32) for patients receiving home-based DOT. All adherence months where patients

took 0% of their treatment were recorded at out-patient facility-based DOT sessions. Finally, graph 3d above shows that poor adherence is a good indicator of a LTFU outcome. Furthermore, for some patients, death was preceded by a period of poor adherence, which was usually related to a severe adverse event.

FACTORS ASSOCIATED WITH ADHERENCE Tables 1 to 3 show the proportion of patients that adhered to treatment (defined as above 90%) for different patient characteristics. Annexes 7 and 8 show graphical representations of adherence for variables that show some variation in adherence (when stratified by these variable’s values). Table 6 shows the unadjusted odds ratio of poor adherence amongst selected characteristics. The only factor which shows a significant difference in levels of adherence is the number of MH sessions attended. Those with higher attendance tend to have higher risk of low adherence, which is likely because high risk patients have more sessions scheduled. Patient with alcohol and other substance abuse disorder were 7.67 times more likely to not be adherent than patients with alcohol disorder only (p=0.083) At this

early stage in the study there are few concrete patterns in adherence, and multivariate analysis therefore was not conducted. Patients who had at least one serious adverse event tend to have worse adherence than others, as can be seen in table 2 and in the graph in Annex 7. The few patients who had no adverse events recorded has worse adherence - though, the graph in Annex 7 suggests this is presumably due to bias in that the patient’s adverse events were not recorded due to being LTFU early in treatment. The median adherence for patients with Hepatitis C is lower than that of patients without Hepatitis C (95.7%; IQR: 89.4- 98.9 vs 92.2%; IQR: 77.9- 97.2 p=0.063). Patients who were single also tended to have better adherence than those in a union compared to those that are divorced, widowed or separated (96.8%; IQR: 94.299.5 vs 95%; IQR: 85.7- 97.5 vs 86.7%; IQR: 70.3- 97.1, p=0.073). The graphs in annex 7 suggest that, whilst there is no statistically significant overall in differences in adherence, there is a noticeable dip in adherence in the first few months of treatment for certain groups of patients: those who are employed, those who have moderate or severe depression, those who have hepatitis B and those who use drugs other than alcohol.

TABLE 6: ODDS OF ADHERENCE BY SELECTED CHARACTERISTICS >=90% Adherence N (%)

<90% adherent N (%)

OR (univariable) (95% CI, and p value)

Female

7 (87.5)

1 (12.5)

Ref

Male

31 (60.8)

20 (39.2)

4.52 (0.73-87.77, p=0.173)

In union

16 (61.5)

10 (38.5)

Ref

Single

13 (86.7)

2 (13.3)

0.25 (0.03-1.15, p=0.103)

Widowed/divorced/Separated

9 (50.0)

1.60 (0.47-5.51, p=0.449)

<35

5 (62.5)

3 (37.5)

Ref

35-55

26 (74.3)

9 (25.7)

0.58 (0.12-3.26, p=0.506)

>55

7 (43.8)

9 (56.2)

2.14 (0.39-13.61, p=0.390)

Employed

14 (63.6)

8 (36.4)

Ref

Unemployed

24 (64.9)

13 (35.1)

0.95 (0.32-2.92, p=0.924)

Homeless

6 (75.0)

2 (25.0)

Ref

Not homeless

32 (62.7)

19 (37.3)

1.78 (0.37-13.00, p=0.505)

Dependent: above90 Gender

Marital status

Age

Employment status Homeless

>=90% Adherence N (%)

<90% adherent N (%)

OR (univariable) (95% CI, and p value)

Yes

14 (56.0)

11 (44.0)

Ref

24 (70.6)

10 (29.4)

0.53 (0.18-1.56, p=0.250)

Yes

19 (76.0)

6 (24.0)

Ref

19 (55.9)

15 (44.1)

2.50 (0.83-8.30, p=0.115)

Alcohol disorder

23 (65.7)

12 (34.3)

Ref

Alcohol plus other substance dependence

1 (20.0)

4 (80.0)

7.67 (1.00-159.19, p=0.083)

Opioid dependence

5 (100.0)

0 (0.0)

0.00 (колич.A-I колич.f, p=0.995)

Personality disorder

3 (100.0)

0 (0.0)

0.00 (колич.A-I колич.f, p=0.996)

None recorded

6 (60.0)

4 (40.0)

1.28 (0.28-5.39, p=0.740)

Confirmed MDR

11 (78.6)

3 (21.4)

Ref

Confirmed pre-XDR (FQ)

7 (53.8)

6 (46.2)

3.14 (0.62-19.12, p=0.181)

Confirmed pre-XDR (Inj)

10 (66.7)

5 (33.3)

1.83 (0.35-10.89, p=0.476)

Confirmed XDR

9 (60.0)

6 (40.0)

2.44 (0.49-14.32, p=0.286)

R resistance with H susceptibility

1 (50.0)

3.67 (0.12-114.48, p=0.404)

Mean (SD

1.2 (0.7)

2.4 (1.6)

3.30 (1.69-7.99, p=0.002)

Dependent: above90 Hepatitis C Previous TB diagnosis

Psychiatric diagnosis

TB type

MH sessions per week

TB TREATMENT OUTCOMES Patient outcomes were reported at 6, 12 and 18 months since treatment debut. At 6 months, 44 (74.6 %) were still on treatment whilst 14 (23.7%) had either died, or were LTFU. All patients who were LTFU had been followed up at least twice by their counsellors before being

declared LTFU. Outcomes at 12 and 18 months can be seen in the figures 4b to 4c below. There were 9 patients who were transferred to other facilities and 6 (66.7%) were recorded as being transferred to Sosnovska hospital. Of these, at the time of analysis, 7 were still on treatment, 1 had completed treatment and the other was listed as cured. Figure

5 shows a Kaplan Meier survival curve where the outcome is set as died, failed or LTFU. It shows that at 6 months 76% (95%CI: 66% - 88%) were in care therefore 24% (95%CI: 12%-34%) had had a negative outcome by 6 months. By 12 months 33% (95% CI: 19%-45%) had a negative outcome and at 18 months 36% (95%CI: 21% - 49%) had a negative outcome.

QUALITATIVE RESULTS The findings are presented from both in-depth interviews with patients and focus group discussions with practitioners. Two main themes emerged from the data, in the first instance the theme that emerged was a description of the situation that patients are living in, which is characterised by a state of vulnerability, despair and experience of prejudice and marginalisation due to their medical conditions, prison history and socio-economic situation. The loneliness and difficulties in life experienced by many patients have potential to lead to poor health and treatment outcomes for the patients. The second theme relates to the patients’ path out of this vulnerable state. Patients described their vision of what it means to be healthy, and described the fierce self-determination they all possess which, if met with a desire to be well and a trusting health practitioner to be their guide, then this can lead to positive health outcomes. Patients described having trusted loved ones was also important for positive health outcomes, and for some the programme was able to help them cultivate positive relationships with their loved ones. The combination of the two themes indicated that the programme was both acceptable and feasible for most participants in most cases, but there are some areas where the programme could be improved or adapted to suit the context.

VULNERABILITY, DESPAIR AND MARGINALISATION: Patients that participated in interviews expressed vulnerabilities and despair in their personal lives and socio-economic situations that are exacerbated by having MDR/RR TB, alcohol or substance abuse and by being hospitalised. SICKNESS AND DEATH Patients describe very powerfully both a fear of death and of sickness, and an experience of sickness and fragility caused both by TB and by the side effects of TB treatment. Practitioners described how patients experience ‘rather high anxiety levels at the admission stage because the patients do not understand their diagnosis at all.” [Fgd1] or as one patient described: “I want to put it behind me as a nightmare.” [IDI8] For many the anxiety experienced at the beginning of treatment can be exacerbated by the fear of the effects of the disease on their loved ones around them: “Well, I was shocked that could happen to me. Firstly, I didn’t know how to survive it, how to break it to my family. Secondly, I worried so much about my kids, so that they didn’t get it. I worried about my husband… Psychologists were treating me at the hospital, as I couldn’t get over it, bad stuff kept coming to my mind.” [IDI8] For some patients they were not just coping with TB: “If you take tuber-

culosis, HIV, Hepatitis C and social issues all intertwined, it is a heavy load altogether.” [FGD1] For some they discovered they had all three at the same moment: “On the 15th I took a fluorography, and on the 23rd I was in hospital already. They found three diseases. And all of them are fatal. Well…that’s it, I am on treatment” [IDI11] The above patient then also contracted COVID-19 and became suicidal once in forced isolation. Some patients were extremely sick from the TB, whilst others felt healthy despite the TB infection. One lucky patient said “I don’t notice challenges at all. I’m surprised myself. The drugs are easy to take. Very easy.” [IDI3] But patients in the IPD were struggling with weakened legs, and others described coughing up sputum and blood. It was similar with the pills, some had no side effects whilst others had strong side effects from the pills: “I feel tired, weak, and I don’t want anything. It’s hard to move. Everything gets hard [after taking the pills]” [IDI10] One patient worries so much about the side effects, and his situation of forced hospitalisation he cannot believe the treatment can be approved for use, despite staff indicating that he had repeatedly been shown evidence the treatment is approved: “are they legal, not legal? Until now, I have not had a chance to read or see any such certificates, documents, so to speak.” [IDI6]

In the current climate, one patient expressed further fears that his vulnerable state of health could be worsened by the COVID-19 situation, and furthermore he felt he was not being taken care of correctly: “they say on TV that those having chronic diseases shouldn’t go out, but they have us take public transport. We’d better not watch TV. Well, not meaning at all… I did have complaints – and they had the doctors to see me, but what’s the use [of seeing a doctor]? No use whatsoever” IDI10 Those forced into isolation and treatment described even stronger anguish, for example a fear of dying, fear of mistreatment and being alone and left with nothing: “Is it that I’ll sit here for two years with such super-duper doctors. And I will also come home, and in a year, I’ll kick the bucket at the age of at thirty-three or thirty-four? What kind of treatment is this? I have debts there for the apartment, here – I don’t care about you [I.e. me]. Then they will release me, bugger me, I will be there without an apartment, without anything, I’ll lie down and die. So what good am I, why treat me? Let me out, and I’ll die there, in the forest.” [IDI6] “Why do the police have such control over the patients? … So little space here, very little. They need communication. And distraction. They need to be distracted. That’s where the suicides come from. You see how it can be. It happened twice during my stay here.,” [IDI5] For some of the patients in forced hospitalisation, it was noted that the experience had been worse than it should be if the set process had been followed. For example, they did not receive notification of their legal summons to court, before they were forcibly taken into hospital by the police, so the experience was quite a shock. FRAGILE SOCIAL SITUATIONS Economic worries including joblessness or homelessness were also an issue for the majority:

“I can’t get a regular job. I can’t get a pass from the medical commission to get a job if it can harm me in any way. But then, harm can happen at any job. Well, I don’t understand why they don’t give me a disability group for the time being. You’re left to pull through it, the best you know.” [IDI10] And again, economic issues were worsened by forced hospitalisation: “I don’t pay [for my apartment] now. I’ll get out of here and I don’t understand how I’ll live there, how I’ll find a job. ….. When I am discharged… how I will live in general. I just don’t understand this.” [IDI6] Amongst those not in isolation however, some were now working part time or casual jobs, though even the interview candidate who seemed in the strongest and healthiest state still expressed a level of stress: “I want less stress. Nerves, of course, ruin health. Well, all is fine. It would be great if someone else helped financially.” IDI3 USE OF ALCOHOL AND OTHER DRUGS Many patients spoke articulately about the reasons for and the experience of abuse of alcohol or other drugs: “Before, when I was taking drugs every day was like Ground-hog Day. Each day was the same. Wake up, look for money, steal something” IDI3 “You want to give up drinking (slang: literally ‘climb off the glass’)… and you can’t. I mean, it seems like you gonna just have a hangover drink, just a little and that’s it. But in fact this ‘just a little’ is too little, not enough, … and back again… you wake up in the morning and feel sick again. Have a hangover drink again and then you walk like this… Shsh-sh-sh (makes circle movements with his hands). And it goes on…” IDI2 Alcohol was reported to be very acceptable in Belarus, even taboo to not drink at all and there is a lack of sufficient government policies to encourage safe or reduced use of alcohol. Practitioners also suggested that patients lacked methods of dealing with stress,

other than alcohol. Patients described how alcohol could be used to drown out feelings of ill-health: “Well, I was under a local anaesthesia” [IDI1] or “I washed it down with a drink. I didn’t want to believe I had TB.” [IDI5] The topic of substance abuse wasn’t easy for some participants to discuss. This was especially the case for women, who would not admit to abusing alcohol or not looking after their own health in interviews. For example, one woman said: “I know such people, but I’ve got no such challenges. My family comes first for me.” [IDI8]. Whilst none said this explicitly, it might be hypothesised that women were scared to open up about alcohol abuse because of the fear of what would happen to their families if they were registered as having an alcohol disorder. The patients that had recovered from their substance abuse disorder tended to mention that meeting a certain girl, or the close bonds of family were part of the process or recovery: “Maybe I’d keep drinking if not meeting my girlfriend who provided some support. I don’t know. It happened…. Of course, I have others. I have my mother. I have a wonderful mother. She always supports and helps me” [IDI3] PREJUDICE AND MARGINALISATION For many patients, an experience of abandonment or marginalisation from their loved ones was common, and in some cases this same treatment was felt from health practitioners working with them. One patient described his own reasons for alcohol dependence as follows: “It is just that I had a psychological trauma. Everyone wants to be healthy, but the thing is that I faced the situation when the closest people let me down…I was in such a mental state that I felt that no one needed me and they did not care about me. However, whatever happens, we have gone through thick and thin together with my wife (deep breath). I’m sorry. Well … It’s a long story. It’s hard to understand it all at once. One has to live this life fully. Many people betrayed me, but all is fine. This is а sore point. Better not touch” [IDI1]

Some MOH practitioners talked about their patients using the following language: “We have problems only with the patients from the risk group – drug addicts, alcoholics, and individuals with an imprisonment history… but well, trust me, not just healthcare practitioners are challenged by those patients. They are a burden to other healthcare facilities as well, I mean the policlinics, as well as to their families and, in fact, to the whole world.” [FGD2 – MOH staff] “They want to have it their way. (Indignantly) They want to live as they like, as they have lived before, not having to change anything and in terms of alcohol – yes, they all want to have fun, they want to enjoy themselves, they don’t want to work… Well, it looks so elementary to us all to get it, doesn’t it?” [FGD2- MOH staff] It should be noted however, that professional burn out was a common problem for MoH staff due to their heavy and difficult workloads, so these comments should perhaps be taken with that in mind. Some patients clearly described they felt that they were being treated in a way that was less worthy than other people. For example, one patient, who after a year and a half of forced hospitalisation asked “It’s not clear at all whether this is a treatment, or I’m in a madhouse, It’s not clear at all.” One patient also described that a staff member had revealed sensitive information about the precarity of his situation to another patient, thus betraying his confidentiality. The patient then described the counsellor as follows: “he likes to talk… Well, what he says, for example, confidentiality, or maybe it was just out of inexperience, it happens, that he blurted out. At first, I thought it was a joke, that the other patient knew what we were talking about. I thought it was a joke.” [IDI6] The counsellor, admittedly revealed this information to another patient in the context of asking if the same help was needed by this other patient for a similar vulnerability, but the damage was done. The patient’s confidentiality was

betrayed and he felt judged and teased by others for his situation, even if this was not the counsellor’s intention. ISOLATION AND LONELINESS Patients described feelings of physical and emotional isolation, especially in the case of forced hospitalisation. Patients who were forced into isolation described the negative impact this had on them: “He’s back on the block, he called a couple of times and that’s it. Now he hasn’t called for three or four months.” [IDI6]. One patient who had been in an out of prison for many years described forced isolation and treatment as “a terrible mess. They don’t take patients for humans…. it’s very scary out there. Worse than in prison.” [IDI7]. Another patient who had never before been in prison, was very disturbed by this: “their duty is to guard us, I do not flee, do I? I do not trespass the perimeter.... I do not climb up the fence” [IDI5]. But even outside of forced hospitalisation, many patients also described living a lonely and isolated life, which was sometimes exacerbated by having TB or alcohol abuse: “Well, if you want my honest opinion … (laughs) I don’t want to say that I have no friends. All are only sympathizing. Many are afraid.” [IDI1] “Well, no, I had a family once, but we aren’t close ‘cause… whatever” [IDI 10] “You know, I cannot say anything [about loved ones], unfortunately. They were buried long ago – my mother, father, grandparents. I’m left here alone.” IDI5 Some patients described close and important bonds with only a few remaining people, or for some simply their dog or cat and had very determined views about trust and the importance of it: “I find support in those who are close to me. Those whom I allowed to get close to me.” and “If it wasn’t for her, why would I need my life then? [IDI11]

“Now you cannot even trust your family. That’s it…. As strange as it sounds. Iin my little, so to speak, life, I have lived only 32 years, and to be honest, I have seen very little good. … With friends, I have started to communicate with them less often now. For me now my dog is more like a friend and companion. Who will not betray me or bugger me. And the rest… I say, I have a little disappointment in people.” [IDI6] Some interviewees described their lives in a positive and healthy light, which was sometimes contradicted by the rather sad tone of their voice or descriptions that indicated their isolation and loneliness: “If you don’t love yourself, you’ll dig a hole for yourself, right? If you treat yourself like an animal. You should love and care about yourself, it’s only once that we live here.” [IDI5]

SELF-DETERMINATION, GUIDANCE AND ACHIEVING A STATE OF HEALTH SELF-DETERMINATION Despite the tough situations characterised by isolation, despair and prejudice from those around them, most patients described real strength in themselves and a determination to get through: “I’ve made up my mind. I’ve got a will power. I want to be cured. I want it to be all right, to go on living” [IDI5] This resilience and determination was for some learnt as part of a tough life, often with one or more stints in prison. Though for some this strength of character and determination can have another more negative side to it: “We, Belarusians, are like this: if we serve [in the army], we do serve. If we have fun, we do have fun” or according to the Belarusian proverb: “A quick mare is in time everywhere” [IDI1]. Meaning that the same determination to carry out difficult tasks like being in the military, or getting through prison might on the other hand be the same determination that can keep someone in a cycle of abuse of alcohol or other substances. For example, one patient in a Forced Hospitalisation Centre refused to take

his treatment until his partner, who had escaped, was returned to be with him. He would only take alcohol during that time, but as soon as she was back, he continued his treatment and stopped drinking, with the advice of a trusted doctor. Many patients went further to say that there was nothing to be done for patients who were stuck in a cycle of drinking, until that self-determination could be directed to getting better: “I would say if someone does not need that support, drinks throughout the treatment. You’ll not help them until they start wanting it.” [IDI3] GUIDANCE THROUGH THE PROGRAMME The guidance and care from practitioners in the programme were essential to reinforce any self-determination that a patient had to get through treatment. One counsellor-educator described the programme as follows:

“What is good in the system, as I see it, where it works well and efficiently… When I establish contact with a patient and he trusts me, he may not have good contact with the attending doctor. So, he tells me about his insomnia or mental issues. I can resort to the psychiatrist. Then the psychiatrist is involved from her side. If there is a medical issue, I will address it to the doctor, to help with the medical aspect. For social issues, I’ll contact the social worker. This is how the multidisciplinary approach works. The patient can get help from all sides and he has people he trusts. He realizes he is not alone in certain processes.” {FGD3]. Counsellor-educators were referred to as inspirational, trustable and professional by both patients and practitioners working outside of the programme. The symbiosis of the relationship between positive practitioner approaches to care and the patient’s

own efforts to keep healthy, and follow advice make for a strong potential for the programme to work. There were numerous descriptions of how counsellors really help patients, which can be seen in box 1. Patients described that counselling would help them deal with the stress of hospitalisation and sickness, also personal issues such as conflict with their partner, or would help relieve their financial or administrative stresses in life or help patients deal with their fears about TB and TB treatment. One MOH doctor described that for some patients, this counselling would be one of the only times a patient was treated as a human being. Some patients appreciated the counsellors distracting them from the boredom and horror of in-patient life, either by telling them about the outside world or providing activities such as games or painting. The counsellors were particularly appreciated by those in forced isolation and treatment.

BOX 1: QUOTES DESCRIBING THE BENEFITS OF COUNSELLING IN THE PROGRAMME “When I had some issues with my wife and nuances, I also contacted [the counsellor]. Well, it is he who helped me. I am glad that I have someone to contact, to talk, to seek advice. Sometimes even not about the disease, but about real life, as they say. So, I’ve got his phone number, we are always connected. Well, it means I am glad that there is such an organization and I have someone to address.” [IDI1] “Sometimes it happens in the case of patients with a drug addiction that they want to talk, and one has to, dig deep into their souls to try and stabilize them from a psychological point of view, and the MSF team helps a lot along these lines by assigning counsellors, psychiatrists who come and talk with the patients, sometimes for hours, trying to uncover their deepest fears. And in fact, they often manage to do it” [FGD2] “As I see it, many patients stay more-or-less adherent to treatment just because of this support, and they feel, well, a kind of gratitude for being treated with such care as if they were small children. But at the same time, this cannot be said about everyone, but for many – yes, it is true for many patients. It is great and really nice, to my mind.” [FGD2] “She somehow inspires, confidence. Because it's somehow easy to communicate with her, and she always helps so much. If you have any questions, you can call her, or when she arrives, you can discuss with her. She always finds a solution, an approach. Well, she's such a good person. That’s it. For me she (chuckles) is such a good person, quite educated, and a specialist, an assistant. I trust her more than the others. How do I trust? She has more of an approach or something. She is somehow without any aggression.” [IDI6] “Many young girls who faced TB for the first time were meeting your psychologists very often. I saw that it was helping quite well.” [IDI3] “I mean, like the counsellors visiting us. They come and support, distract us, communicate with us.” [IDI5] “When I was here it was natural. I had a chat with a psychologist. Helped me to relieve my stress. Of course, I benefited from this. Though, I did not need a psychologist a lot. I’ve understood everything long ago.” [IDI3] “We help him understand what he wants in his life. Very often a patient faces money issues, no documents or no money for the documents. Of course, his alcohol or substance abuse hugely affects his adherence. In this case, step by step, addressing it we changed his life for the better. This patient had drinking problems. Besides, as far as I remember, he had problems with documents. As a result of his drinking, he had employment difficulties. Over time, all of these were settled one by one.” [FGD3]

Counsellor-educators described that achieving these results was not as simple as just taking a text book approach:

how this would help them to get well: “The main thing is to follow the doctor’s prescriptions.” [IDI1]. Even for patients who reported very low trust in anyone:

“If you just technically follow some guidelines, I believe the efficacy will be pretty low. … you have to turn on your personality… your charm and emotions, like you smile or tackle some patient’s personal issues if he’s eager to talk about it. So, you build up trust. In fact, this is the first step, so to say. The patient then opens up and begins to trust you. Your word gets value and meaning. Then you can move on to such issues uh… concerning adherence or medical stuff.”

“I am a distrustful man. Trust nobody. Well. But I trusted him [the doctor]. And I tested him. And when you trust a person, you know…Well, he didn’t give me away to anybody. He had me on my feet – well done to him, all is not so bad. He is not a stranger to me. He treated me like a person whom he is close to. More, he treated me fondly” [IDI11]

Counsellor-educators also described how they would approach dealing with patients who were suffering in a cycle of alcohol abuse and the shame they felt from it, in front of their loved ones: “We need to work with the patient’s concept of the world, so that he could accept his disease and treat it like a disease and not like his own fault, like he is guilty for his illness. Very often it damages family relations when there’s no support or acceptance from the family. We need to work with the patient on the acceptance of the disease. Then we need to work with the family.” [FGD3] For some however, the approach was not appreciated. One particularly mistrustful patient, who was suspected to have a mixed personality disorder, and had rejected the more extensive psychiatric aid, such as mood stabilisers, that it was thought he needed, said about his counsellor “I will tell you right away…Does it help him…or me? I don’t know.” [IDI11] Another patient also described how she didn’t understand how talking to counsellors was helpful for her TB treatment, but she did appreciate being able to talk about life and family. One patient, an ex-opioid addict - appreciated the interaction, but felt that he had already dealt with his issues before entry in to the programme this time. Patients also talked specifically about their trust in doctors, many of whom were not part of the MSF programme but were MoH doctors, and

However, one patient described how the high turnover of staff was a problem when trying to build a relationship in order to receive psychiatric care: “They were coming and running away. I remember Thumbelina, then Snow White [both nicknames] came instead of her. They were changing fast. You only get used to her and the next day she disappears. A new person, whatdoyoucallit, comes. That’s why I don’t know … From the very start, I didn’t get on well with them. The Snow White comes and asks “Are you taking drugs”? That’s it. All the communication you get….” [IDI7] Patients would often be treated by different doctors or psychiatrists, depending on the shift and depending on where they were seeking care – be it forced hospitalisation centres, outpatient or in patient at the institute. The system of changeover also meant the patients switched counsellors in different treatment phases, which can’t have been optimal for patients who have experienced such betrayal and marginalisation throughout their lives. For a few patients however, which is contrary to the idea of the programme, they felt that the doctors they encountered would not treat them as human beings but would only care about their adherence to TB treatment and were not interested in other ailments they are experiencing: “But they already have such an attitude that you are not a human being, but a piece of some incomprehensible flesh... He says: - You know, I don’t effing care that your kidneys hurt, your teeth ache,

your head hurts, I don’t care. He says to me: -Do you want to live? I say: - well, yes. He says: The main thing for me is your lungs, and about the rest, I don’t care.” [IDI6 – Patient in FHC] For one of the focus group participants it was clear some doctors made the effort to engage with patients, while others had no motivations to: “Some come to work as they would come to a factory – to perform a mechanical set of actions, do a monotonous work and leave, and some get creative about it all.” [FGD1] The social interventions provided to patients were also important and often well thought out and appreciated. For many transport cards and food packages (distributed by the state) are important motivators to help patients through treatment. “They hand out those transportation cards, personal hygiene things, cleaning things for the house, and then food packages also.” [IDI10]. The work of a social worker in the project was described as follows: “We put our efforts to solve the patient’s needs, because the issues were ranging from legal to basic questions like: “Where shall I go with this piece of paper? What will I get here?” When the patient finally saw the way out, what I kept doing was motivating the patient to use his own resources.” [FGD1] This process of guidance could also lead to helping patients get a job or pay off debts. This was in fierce opposition to the experience of one patient in forced hospitalisation at FHC who described losing his job and his apartment when he was suddenly and forcibly taken for isolation and treatment. Perhaps if treated as an outpatient he would have benefited substantially from a social worker, and would not have been so worried about how he could live when he left hospital. HEALTH, THE OUTDOORS AND FREEDOM Patients described the state of health that they had experienced in the past, or that they would like to achieve

again in the future. This healthy state was described in the context of escaping, being in nature, and having a sense of future. Patients describe an idea of health as being: “not smoking, doing sports, riding a bicycle, fresh air. I saw happiness on this man’s face. I mean, he defeated the disease.” [IDI1] or “I don’t lose heart… and once a day I exercise – I go out to over there. Well… I keep my spirits high, telling myself everything’s going to be just fine.” IDI10 The experience of forced treatment would take away this important outdoors: “It’s only one thing here… there’s no freedom. Little, little freedom we have here. I’d like to have more. More space to go for a walk. We don’t have it here - just the fence and concrete.” IDI5 Patients described a fairly individualise image of healthiness, characterised by being strong and in the outdoors. But, whilst it wasn’t mentioned explicitly in the context of being healthy, many patients had also described vividly the importance of close relationships with those around them that they trust, which is clearly important in terms of mental and social health. ACCEPTABILITY OF THE PROGRAMME The two overarching themes together indicated that the programme was well designed to deal with the fragility and marginalisation experienced by patients in the cohort. But at the same time the loneliness and isolation experienced by patients, which can be worsened by long periods of hospitalisation, could potentially undermine the successful outcomes seen in most patients. Patients self-determination, coupled with the personalised guidance given by the multidisciplinary team worked well together to bring patients back to a state of healthiness which could involve feeling strong, free and in the outdoors whilst also helping patients improve their social and economic situations. One practitioner participant described that: “I believe the project is highly effective judging from my feelings, from the discussions with patients and colleagues.” [FGD3]

IDENTIFIED GAPS AND SUGGESTED PROGRAMME IMPROVEMENTS Practitioners and patients indicated a number of ways the current approach could be improved, these results were fed back to practitioner participants, and improved upon as part of an iterative approach. EXTENSION OF THE MULTIDISCIPLINARY PATIENT CENTRED APPROACH Ministry of health staff described how TB patients with drug and alcohol problems tend not to stick to treatment, and get into a vicious cycle: They are motivated only when they feel bad, when they have fever at the point of 39 degrees, hemoptysis, when their lungs fall apart – that’s when they begin to plead, to beg, to say they will follow the treatment. As soon as we have stabilized them – after they have stayed on treatment for several months, and the cavities in their lungs have closed up, and their temperature has gone down – alas, they feel well, and nothing interests them again, they abscond and the story repeats in the course of the next few years to the point when they have developed a chronic disease and an MDR form, and the whole cycle goes on and on until the point when they are completely destroyed. [FGD2] For this reason, perhaps it might be worthwhile to utilise the approach for non-drug resistant TB cases, rather than waiting for drug resistance to develop before taking this more immersive approach for patients with known adherence challenges. Furthermore, patients complained that doctors would only treat their tuberculosis, and would not take into account other aches and pains that they were experiencing in other parts of their body, or would not treat other co-morbidities such as Hepatitis or HIV at the same time as their tuberculosis. For this reason, the multidisciplinary approach could be extended to include other medical specialities.

ENHANCED ENGAGEMENT FOR PATIENTS WITH SUSPECTED PERSONALITY DISORDER Participants described how the approach could be improved for certain patients. MOH staff describe that patients with addiction to drugs other than alcohol or those with a significant criminal history are often not motivated with treatment and are prone to manipulation, and that harsher measures are needed to get them through treatment: “it gets most challenging with the druggies. Cause, as a rule, they are engaged in the drug deals and have money. Erm, and they are, in fact, completely delusional, and it is very difficult to find something that will motivate them. Most often they manipulate and use all the aid that’s provided in their personal interest. And therefore, when working with such people we need, and it is our common opinion, we need either to isolate them or, (Sighs) to find individually tailored psychological approaches, or psychiatric methods, well, whatever is there. But this is the most difficult group there is.” [FGD2] Some patients in the cohort were challenging, despite not necessarily being in the above category. For example, those who “You spend hours on health education, several times, one session after another… Well, they just don’t hear you.” [FGD3]. This desire to not hear health education was evident in several patient in interviews, for example one patient refused to believe that he had TB after 18 months of treatment or that the treatment was proven to be effective, and another who was scared to share a glass with someone who had HIV for fear of catching the disease: “There were also HIV patients. Initially, I was sharing a room with an HIV patient. How do I know if he is having AIDS? I was sharing a glass with him. This is the way they treat us.” [IDI7] According to the MSF project psychiatrist “the main problem is personality disorder” [FGD1], as it is these patients that are difficult to reach an agreement with, particularly those that

have not been diagnosed and treated for their personality disorder. Those with a diagnosed personality disorder were more likely to end up in a forced hospitalisation centre, as it was too difficult to provide a tailored psychological approach for them with the current resources. This tailored approach could be a token or incentive system, where rewards are provided in the short term for adherence or treatment compliance, rather than relying on the ‘reward’ of being cured of TB after 2 years, especially in the case that the patient feels well. This provision of incentives is not a simple concept however. One patient, who was a renowned and capable manipulator with a personality disorder who had had TB for 20 years admitted that “You know, I got a flat because of TB. We live in a 3-room flat.” [IDI7] For him the provision of a flat was motivation for him to keep having TB, though he also stated “If 20 years ago I knew that TB… I would sacrifice 1 full year instead [of taking the flat].” The idea around incentives and tokens would be to help patients with personality disorder make the right decision upfront. Behavioural work or very engaged psychotherapy on top of the standard counselling would also be important for patients with personality disorders, and work would need to start in IPD then continue in OPD. It should be noted that this approach is very resource intensive, and there are not enough staff in the project with the appropriate skillset to carry this out. In general, psychiatrists reported to be cautious in making psychiatric diagnoses such as for personality disorder, or for a formal diagnosis of substance abuse, as there are negative consequences to having a formal diagnosis for patients, and not necessarily any benefits. One recovered opioid addict described the diagnosis of drug addiction: “There was not much of an advantage. I’m saying again. I stopped only when I myself came to it. Not because of being registered …. I have been registered for… Let me count exactly… 13 years. And until the moment I took my own decision, I had been taking drugs. [IDI3]

TIMELINESS BETWEEN DIAGNOSIS AND TREATMENT DEBUT, AND ENHANCEMENT OF CONTACT TRACING/ACTIVE CASE FINDING When it comes to TB diagnosis for some patients, the process of being diagnosed and starting treatment seemed to be quite slow. One patient mentioned it was 2 and a half months before starting treatment, whilst for others it was a matter of weeks. Interviews with patients also indicated that when an MDR/RR TB case is found, more extensive case finding and health education could be beneficial as: “I visited one place and the doctors told me later that I shouldn’t have stayed at that place. There were people with TB who ignored it and never showed up to the requests sent to their registration place (place of living). They had known they were ill, but they didn’t warn us about their disease. This is where I got TB.” IDI5 If someone is positive, it is usually just their close family that are traced, but for those who are known to have alcohol or drug abuse issues, the tracing could perhaps go further. RELATIONSHIP BETWEEN MSF AND MOH STAFF Furthermore, the working relationship between MSF and MOH staff, though usually good, was not always so: “not all MSF employees establish good relations with the Institute’s staff. Whilst in some cases MoH staff would not collaborate with MSF… a [health worker] who used to work here… He had prejudice against MSF and he repeatedly mentioned it. Well, as a result, he was prejudiced against me as well. He refused to collaborate. Now he works at XXX but we don’t interact at all, he does not even pick up his phone when I call him.” FGD1 “The only drawback that I see in our interaction is MSF’s staff turnover, … by the time we get used to each other and the patients get used to him too – this time, it seems to me, is lost somewhere.” FGD1 FGD participants also complained that MSF doctors are not in control

of medical decisions for the patients which can cause issues if they reach an agreement with a patient that is then not agreed by the MoH doctors who take decisions. “on the part of MSF... we want them to cover more. They don’t, for example, keep a medical history, don’t keep drug charts, for example, or aren’t engaged in stopping adverse events in a patient “ [FGD1] This could be improved if MoH doctors were capacitated to take a more patient centred approach to treatment. In some cases, the role of the MSF doctor role was described to be “settling conflicts and bringing the patient back to treatment”. MSF doctors have more time to spend with each patient, and therefore can have more ‘patience’ for each one. Furthermore, MOH considered that “the work that MSF does now with so many staff, in fact, can be performed with fewer people” [FGD2] and this perspective was generally agreed within MSF, and actions taken to expand the number of patients. MULTI-DISCIPLINARY APPROACH Multi-disciplinary working, where the patient is included with other specialists as a decisionmaker is not yet normal in the Belarussian context: “a TB specialist may not communicate well with an infectious disease specialist. Doctors believe you have to perform your duties and you have to do that well. And that’s sufficient, in general. And you don’t need to consult a patient. Well, this is our peculiar Belarusian feature.” [FGD1] Occasionally the person-centred vs traditional medical approach do not always work well together and it ends up being “a crayfish and a swan who pulled the wagon in different directions” [FGD1] This multidisciplinary, team work approach could also be taken further. For patients who will not accept a decision or listen to health education, more group work could be beneficial so patients can learn from others who have already lived the consequences. For others, more multi-disciplinary interventions by a group of practitioners might be persuasive:

“group meetings with different specialists organized for the patients are efficient in such cases [those who will not listen to health education]. Namely, the MSF doctor, the MoH doctor, a psychiatrist, a counsellor, a unit head, everyone gives his vision on the disease so that the patient gets the joint opinion from many specialists.” [FGD3] PATIENT CENTRED APPROACH One patient, who had just been informed that doctors were taking him off treatment due to poor adherence, wondered how and why decisions would be taken about patients without including patients in the consilium decision making process. Though, some stated that whilst the person-centred concept was initially vague and not usual for their practice: “We don’t consult a patient this is not the Belarusian way” [FGD1]. However, for some MoH staff, this harsher approach is still considered superior for some patients, though some ambiguity is clearly understood regarding the benefit of a patient centred approach vs an approach that is highly effective in ensuring treatment adherence but rather inhumane: “it is my conviction, and I will stick to it, that if we succumbed less to your staff’s position and were more tough with the patients and followed the policy envisaged by our legislation proceeding with the forced treatment, then several patients who were lost to follow-up, would have completed their treatment. But it all turned out that we were indulging them (literally – stroking their hair) and they liked it a lot – in fact, it was a positive thing here and for many patients it was important to get this kind of support, they enjoyed it and certainly they felt treated as human-beings – what doesn’t normally happen often in their lives, and that probably lead to at least one of our patients, for sure, didn’t complete her treatment.” [FGD2] But MSF’s approach does reportedly try to deal with this: “Of course, we had such negative cases where a patient would come back to his environment where he lived, and that might lead to certain consequences when he skipped treatment for a few

days. We tried to control such cases, and explained the importance of this to the patient, so that, he could solve his social issue on the one hand and didn’t face problems on the other.” [FGD1] Furthermore, some patients do request forced isolation, and one ex-opioid addict partially backed up the efficacity of the threat of harsh treatment by saying that he was able to stick to treatment for the following reason: “I don’t know. Maybe I was partly afraid of being sent to a closed-type hospital. I don’t know. Sometimes I wanted to give up, but something was deterring me. [IDI3] Though the hesitation he expresses in this statement isn’t fully convincing, and perhaps this fear of punishment can be achieved through other means. SCALE UP OF VIDEO DOTS During the course of the project, video DOT was utilised more due to the COVID-19 situation. Patients and health workers did sing the praises of video DOT: “Video-observed treatment is a huge success in solving those issues. It allows a patient to live beyond the schedule of dispensary daily visits once or twice a day.” [FGD1] Though it isn’t available for all patients, notably those with poor adherence or with side effects or co-morbidities that need constant monitoring could not take part in video DOTS. However, some patients in this category did prefer the idea of video DOT: “Well, first thing, they should give the pills in one round for every ten days at least, so that you don’t have to go there every time.” [IDI10] Notably however, the same patient did admit that, when in a binge drinking session, he would not take his TB pills: “but once a month – then I drink for three-four days. Once a month that it happens. I don’t go to get my pills then, that’s why I have problems… it is my rule, you know, I don’t take pills together with vodka. Because I have problems with my liver. So you have to choose – it is either this or that.” [IDI10] But, having the pills at home with him, would undoubtedly mean any

drinking induced treatment interruption could be shorter. Video DOT isn’t popular amongst all patients however: “I fell in love with a young man, I go [to the OPD] for him. I don’t want to switch to viber [VideoDOT].” [IDI8] For such patients, the support and love they can get from meeting and spending time with people in a similar circumstance outweighs the benefits of video DOT treatment. It was not uncommon for patients to fall in love with others in the treatment programme, and often this was very beneficial for them. SCALE UP AND ROLL OUT OF THE PROGRAMME Focus group participants described that the main obstacle to roll out of the programme at scale would be “Financing. The primary challenge. … and probably, HR resources” [FGD2]. In particular “financial benefits to health care providers.” [FGD1] could be important, to ensure that health providers are incentivised, or at least are adequately resourced, to go the extra mile and spend the time with patients. Beyond this resourcing, staff members in MOH would need more training in communication, particularly how to communicate with patients who have alcohol and substance disorder or other mental health disorders. Furthermore, psychological first aid and perhaps some other basic psychiatric and counselling skills i.e. active listening could be more widely rolled out, in combination with a more substantial supportive supervision system for staff in the programme to ensure that all health workers are learning the skills needed to carry out multi-disciplinary and person-centred care whilst avoiding professional burnout. Practitioners suggested that some might be better candidates to employ this approach than others, as not all are motivated to do more than “to perform a mechanical set of actions, do a monotonous work and leave” [FGD1]. Indicating that perhaps if it were to be rolled out certain staff should be selected to implement it. In particular for roll out to the public sector, social workers would need to be employed as standard which is not currently the case.

LIMITATIONS •

•

Personality disorders are under diagnosed in this study according to the project psychiatrist, and based on reports in patients’ files where personality disorders are only ‘queried’. Treatment for this problem is not possible within the programme, and more information on a patient’s history would be needed to make this diagnosis. Patients are either treated as part of the MSF cohort or the Global Fund (GF) cohort. Both cohorts receive the psycho-social intervention, but only patients in the MSF cohort have their TB treatment within the programme. This means for patients treated for TB that is paid for under Global Fund (33 or 56%) - some baseline medical characteristics are not available in the data base. This includes variables such as whether the patient experiences adverse events from treatment, whether the patient smokes, the x-ray cavity profile. Patients responses on the PHQ9, GAD7, AUDIT and ASSIST questionnaires may not always accurately represent the patients actual state of mental health, and in some cases patients were included into the cohort despite normal scores,

•

as it was suspected by a psychiatrist, social worker, nurse, doctor or counsellor-educator that the patient was struggling with alcohol. Honesty in these forms might increase as time in the programme increases, as the counsellor-educators and other health care workers’ relationship with the patient improves. This may partially explain why some scores go up over the course of the programme, but it may also be because patients leave the IPD at around 4- 6 months so have more access to alcohol after their baseline assessment, which is done in IPD. Half (3) of all women offered the opportunity to take part in the indepth interviews refused to take part in an interview, whereas only 1 man refused to take part in interviews. This is thought to be because the subject matter of the interview, and especially drinking, is more stigmatised for women. Furthermore, the women who did agree to participate, did not talk openly about harmful or high levels of alcohol use, and in general seemed to give formal and socially acceptable answers, in comparison to the information available

•

through the patient’s files. A small number of the men also seemed to share a more formal and socially acceptable description of their current lives than would be indicated by the information in their files. It was not possible to do in-depth interview patients who were voluntarily LTFU from the programme, or with patients who are recorded as having an alcohol and other substance abuse disorder, so, there is some response bias present in interviews as patients who were not adherent to the programme were under-represented. Though we were able to speak to 7 patients with high risk of adherence issues, and 3 who had adherence below 90%, also with one patient who was taken off treatment around the time of interview due to very poor adherence, and recorded at LTFU. Counsellors recommended against inviting the two patients with alcohol and other drug problems who were eligible for interview due to the risk they would not accept, and one is now LTFU. We did try to invite another patient voluntarily LTFU, with whom it was thought it might be possible he would agree to interview, but he could not be contacted.

DISCUSSION OVERALL ADHERENCE TO TB TREATMENT Adherence to MDR/RR-TB treatment for patients enrolled in the person-centred multidisciplinary psychosocial support and harm reduction intervention is high at 95%. This initial result, whilst not yet final, is encouraging, given the complex issues faced by the patients included in the programme, and given forced hospitalisation was not required for the majority (~only

around 12%) to achieve this favourable result. Thus far, this study showed that there is little variation in adherence in most patient sub groups - which may be an early indication that the person-centred approach is really tailored to all individuals, though this it may simply be to low power, so would need confirmation when more data is available. There was however some indication that patients who are in union, who have current abuse of substances other than

alcohol, who have hepatitis B or C and who have not previously been on TB treatment, and those that are employed might have more challenges with adherence than other patients. Whilst it was not possible to formally compare the adherence in this cohort, we have collected information on adherence for other MDR/RR TB patients in Minsk. Median overall adherence for the 59 patients with a disorder due to the use of alcohol in the study

is 95% (IQR: 81- 98) compared to 90% (IQR: 84-99%) for 31 patients (Wilcoxon rank sum test, p-value = 0.9381) not in the programme, but who attend the same facilities, who have completed MDR/RR-TB treatment. The level of adherence for non-MSF MDR/RR-TB patients in the IPD who also had reported alcohol or substance use disorders but were not part of the MSF cohort due to their place of abode was exactly the same at 100%. This might indicate that this approach has a positive impact on adherence only in an out-patient setting – as adherence in IPD is always good, but has substantial limitations in terms of reducing the quality of life for patients kept in hospital.

TB TREATMENT OUTCOMES MDR/RR-TB treatment completion for patients starting treatment in 2018 in Belarus is 74%, and globally it was 57% (World TB report 1), indicating that the level of care for MDR/RR-TB in Belarus is generally higher than in other parts of the world. We do not have final treatment outcomes for patients in this study, but the Kaplan-meier analysis indicates that 76% are in care at six months and 63% are still on course for a positive outcome at 18 months which is therefore looking like it will be lower than the 70% average in Belarus. This possibly reflects the fact that MDR/ RR-TB patients in this cohort are the patients with most challenges to adhering to treatment, and some of these have simply not reached a place where they have the will and determination to stop or reduce their alcohol or substance abuse and to complete their TB treatment. It could also be an indication that, for some patient groups, the approach is not fully adapted to their needs in order to support them to complete treatment.

PREVALENCE OF ALCOHOL DISORDER AND OTHER VULNERABILITIES Using data from patients currently in IPD, we estimate that 40% of patients with MDR/RR-TB have a disorder due to the use of alcohol in Minsk, which is in line with previous estimates. A minority

of patients in our cohort did not openly admit to having high or problematic alcohol consumption, and were only diagnosed as such based-on suspicions from health workers, which in some cases meant having smelt or seen alcohol on the patient. During interviews, none of the women would talk openly about alcohol or drug abuse, and a minority of men would not open up about it either, reflecting the high level of stigma of being dependent on alcohol especially in the case of women. Health workers reported that despite the stigma associated with alcohol dependence, high levels of drinking are very socially acceptable, and that not drinking is itself sometimes stigmatised in Belarus. Furthermore, there is a lack of government policies to curb the quantity of drinking, and alcohol is very cheap and easily available. For some patients in the cohort, alcohol has at times been a major strategy to cope with stress. All participants are affected by other vulnerabilities such as unemployment, a past history of incarceration and many also had co-morbidities such as Hepatitis C, HIV, diabetes, heart disease or other mental health problems. Some of the patients are homeless and others have a lack of social and economic support. For those that did have close family members, some had interpersonal issues at home. For those in union, the quantitative analysis indicates they have worse adherence than those that are single. This may in some cases be linked to sharing and maintaining their alcohol disorder with their partner, but it may also be down to more complex issues of interpersonal distress.

IMPORTANCE OF CLOSE PERSONAL RELATIONSHIPS In the qualitative component, patients, described that their relationships with their family and in particular with their partners were very important motivations for them to either keep going with their TB treatment, or to reduce their alcohol consumption. For some patients, particularly who were forced into hospitalisation, close personal bonds were noticeably absent. This is

somewhat in contrast to the quantitative data where, those in union had on average worse adherence than those who were single. This could be explained by findings in other studies that describe how marital distress and problematic drinking can often exacerbate each other in a cyclical fashion (Rodriguez, Neighbors and Knee, 2014). Information collected through in-depth interviews and focus groups goes some way to explaining this relationship. Some patients described how shame of their alcohol problem or of TB in front of their loved ones was a large motivation for them to keep going with treatment, or stop drinking. It might be this feeling of shame that drives them further into a cycle of drinking or of not taking their pills. There is some evidence in the literature that alcohol and other substance disorders have often been linked to personal relationship distress, particularly during childhood, and associations have been described in both directions i.e. that personal relationship distress can increase the risk of alcohol dependence and alcohol dependence can increase the risk of personal relationship distress (Konkoly Thege et al., 2017; Fairbairn et al., 2018). Some patients described that they had been treated by practitioners in a way that perhaps echoed the poor treatment they had seen elsewhere, for example having their confidentiality betrayed or being treated as less than human. On the plus side, patients and practitioners described that, having a counsellor to help them with challenges in their close personal relationship was a key part of the person-centred care that they receive that was helpful to them. MOH staff also described that for some patients, the programme would be one of the few times in their life that they had someone to help them with their problems, and for some the first time they were treated as a human being. Unfortunately, there was some ambiguity expressed by MOH practitioners on the benefits of this humane person-centred approach, as it was feared in some cases it would lead to allowing the patient to do as they please, and therefore increasing the chances that they leave the programme.

GAPS IN TREATMENT APPROACHES Responses from patients in interviews indicated that in some cases the times between screening, diagnosis and treatment start was surprisingly long. Furthermore, several patients indicated that they probably caught TB from a drinking buddy – but that they had not been traced as a contact of that drinking buddy. Enhanced follow up of TB positive patients’ drinking buddies could therefore be recommended. Furthermore, quite a high proportion of patients have comorbidities, especially HIV, Hepatitis C or heart disease, and some patients complained that they were not treated as a whole human being by doctors but, as one put it, as “a piece of some incomprehensible flesh” [IDI6]. The programme could therefore perhaps be improved by extending the multi-disciplinary approach to be integrated with other medical specialities such as HIV or heart disease. Further assessment would be needed to assess what this would look like, and how it would need to be approached.

DIFFICULTY IN ASSURING PLANNED FOLLOW UP SESSIONS Considering the MSF intervention protocols, patients at high or moderate risk of adherence problems did not on average have the total number of scheduled counselling of contact sessions with MSF counsellors. The programme guidelines states that for high-risk patients this should be 1 counselling session per week and 2 further short contacts, whereas for moderate risk patients this should be 1 counselling and one further contact sessions per week. In reality patients tended to have 2 or three sessions in total per month. Patients who had more overall contacts, did in general have worse adherence than those with good adherence, in fact number of mental health sessions was the only variable significantly associated with non-adherence in the univariate analysis. This is likely explained by the fact that those with higher risk of adherence issues have more sessions scheduled. Lack of at-

tendance was usually linked to a patient not answering the phone, or not turning up for counselling, which is particularly common if and when patients are in binge drinking phases.

METHOD AND LOCATION OF DOTS This study did show that adherence to treatment was better whilst in IPD than for OPD treatment. Furthermore, adherence tended to dip to its lowest at around 4-9 months, the time when patients first leave the IPD for OPD care. However, many patients described the psychological difficulties of being locked up for a year or more. This is in line with global literature, where it is well documented that forced hospitalisation can lead to good TB outcomes, but it is also shown to lead to psychosocial problems such as loneliness caused by not being able to see a long-term partner, or issues linked to not being able to go about the usual social and economic activities (Oladimeji et al., 2016). This study in Nigeria also showed that women, and patients who were previously employed tended to have more psycho-social concerns during hospitalisation. The impact of a lengthy MDR/RR-TB treatment can be worse on those with limited social and financial support networks due to the loss of income and social life that happens when in hospital. Forced hospitalisation is in particular difficult for married women – potentially due to being taken away from their children. In-patient treatment for MDR/RR-TB that is characterised by isolation and where the services focuses predominantly on provision of drugs has been described to worsen feelings of depression and despair, furthermore patients who have co-morbidities are more likely to be depressed. Indeed, the article which describe the negative effect of forced hospitalisation is entitled “Death is a better option than being treated like this” (Huque et al., 2020). The findings of this study therefore suggest that forced hospitalisation rates in Belarus should be further reduced in favour of other options, especially since that the WHO recommends ambulatory care where possible (WHO, 2019).

Video DOT or home-based DOT, as opposed to facility-based outpatient DOT was preferable for many patients., as it gave them freedom to have their own lives, without being forced to come to the facility each day. But, the option of video DOT isn’t preferable for all patients – one patient preferred to come to the facility every day to meet her lover that she met whilst on treatment. The social support element is undoubtedly important, but could perhaps be provided in other ways such as with increased group counselling or health education. Regarding video or home-based DOT, one patient who admitted to treatment gaps, particularly when drinking, did notably say when asked if he would be more likely to take his pills if he took home 2 weeks of pills at once. The patient stated he would still not take his treatment during a binge drinking phase – but video DOT might, with the right phone or home based follow up, reduce the treatment gap after the end of a binge drinking phase.

DIAGNOSIS OF PSYCHIATRIC DISORDERS Personality disorder, according to the project psychiatrist, is under-diagnosed in some patients, and in particular amongst patients who abuse substances other than alcohol, who have an average adherence of only 68%. Those with a diagnosed personality disorder in the cohort had an average adherence of 97%, this is thought to be because for the most part these patients have been having treatment for their personality disorder, and/or they are in forced treatment so have good adherence. Around 20% of FHC patients are estimated to have a personality disorder according to the project psychiatrist. Whilst we cannot show this from the results, due to not having concrete information on the diagnosis, we hypothesis that it is untreated personality disorders that leads to poorer outcomes in particular with patients that abuse other drugs. The co-occurrence of personality disorders, especially borderline personality disorder and anti-social personality disorder, and alcohol disorders are well described

in the literature (Sher and Trull, 2002; Trull et al., 2018). Avoidant and paranoid personality disorders are also thought to co-occur frequently (Sher and Trull, 2002). Patients with both disorders have also been shown to have worse outcomes, than patients with only alcohol disorder. Crucially, impairment of decision making is a behaviour that tends to underlie both alcohol dependence and borderline personality disorder (DOM et al., 2006). Furthermore, it is generally advisable for patients with both a personality disorder and a substance abuse disorder to focus on the substance disorder as a priority (Links et al., 1995). But, those with borderline personality disorder in one study were found to be more likely to also incur treatment centre initiated drop out, due perhaps to being difficult to manage for the staff (Tull and Gratz, 2012). The psychiatrist in the project says that diagnosis is not currently possible as diagnosis requires very specific anamnestic data, including childhood behaviour, part of which should be collected from family members rather than the person themselves. Furthermore, another reason for under-diagnosis, is the feeling that official diagnosis will not always help a patient as much as it could do. For example, a patient who was an opioid addict described how ‘I stopped only when I myself came to it. Not because of being registered [as an addict].’ [IDI3] There are also considerable down sides described to being officially registered as having a psychiatric disorder, for example for alcohol or substance abuse you have to regularly go to the dispensary to give samples, and you are not allowed a driving licence, certain jobs are no longer possible (e.g. working at heights) and there can be some restrictions in parenting responsibilities including a ban on adoption or guardianship. Furthermore, employers may act in a way that is discriminatory against people with a psychiatric diagnosis. A diagnosis of personality disorder would have consequences for the person that depend on the severity of the disorder. If the person is frequently out of control or exhibits socially dangerous behaviour, etc., then the com-

mission sets “stricter” follow up and restrictions: meaning a person cannot drive a car, carry out some types of work, use weapons, carry out military service. If the personality disorder is under control, then the restrictions will be set on an individual basis. Treatment of personality disorders, would also be fairly resource intensive – requiring the provision of reward and incentives for patients with personality disorders who have good adherence and compliance to other aspects of treatment. The provision of incentives for patients with personality disorders might also affect other patients’ motivation, so this might potentially need to be provided more broadly for the sake of equality. Furthermore, intensive psychotherapy or behavioural work for personality disorder would be needed, which requires specialised skills that are not currently available at the required level in the programme.

STAFFING LEVELS First, the high turnover of MSF staff was problematic for their MOH counterparts. Furthermore, staff working in the programme and in MOH indicated that in some areas, the programme is overstaffed. This is particularly noticeable in that there are too few patients per counsellor, but steps are being taken to take on more patients. Then in terms of distribution of staff, it seemed there was too high a rate of change for patients – they might start with one counsellor as an in-patient, then see another when they start in out-patient, and another if sent for forced hospitalisation, which seemed problematic for patients who struggled to trust clinicians. Given that patients have both MSF and MOH consultants, this could be another source of high changeover of staff from the patient perspective. The system of allocating patients to counsellors could therefore improve. MOH doctors on the other hand, seemed overstretched when treating patients, whilst MSF doctors did not have responsibility over patients so did not have to complete some of the tasks associated with this. In some ways this was positive as it gave MSF doctors time to spend with each patient

to provide a high level of personalised care, but at the same time caused issues as another doctor would be the one taking the treatment decision. It would be a lengthy process to change the system whereby MSF doctors could take legal responsibility of patients, but could be worth considering going forwards. Staff in the project estimate that a counsellor or a social worker could cover around 20 patients each. Though if many of the patients were high risk and complex, or if patients are divided between many different facilities then this number would need to be lower, in order to account for increased time spent with each patient or increased time travelling between facilities. One psychiatrist could manage around 20 patients who need psychiatric intervention, though again the psychiatrist’s work would depend on the level of intensity of work needed with each patient, so this figure could change if more intensive treatment were taken on. The programme can then be run with nurses and doctors in ratios that are standard to the country.

PSYCHOLOGICAL TESTS Whilst this study utilises information from systematic psychological tests such as the GAD7, PHQ9, ASSIST and AUDIT, the quality of data arising from these tests was not as strong as other data collected as part of this programme. Counsellor-educators described that baseline patients were not always open to filling their forms in a way that honestly and accurately represents their psychological state, and for many patients the form was not filled out at the scheduled 6 and 12 month follow up sessions, so follow up data is missing. The fact that the AUDIT and ASSIST scores’ median seem to worsen at 6 months may indicate that patients feel more comfortable with sharing accurate information as they gain trust in the program rather than actual deterioration. Some counsellor-educators, concluded that ‘If you just technically follow some guidelines, I believe the efficacy [of the intervention] will be pretty low. … If you add an emotional and personal component, and … if you dedicate your

time, then it works.’ This feeling might indicate that counsellors felt perhaps the complexities of the psychological state of these patients is not accurately captured in the standardised forms to measure alcohol use, depression and anxiety. Which may have led to its underuse, in combination with not wanting to antagonise the patients by making them fill out cumbersome forms, in favour of a more personalised approach. Whilst this attitude is commendable, it would be beneficial for the monitoring of the programme if the forms could be filled regularly, and if counsellors could encourage patients to fill it accurately.

ACCEPTABILITY OF THE PROGRAMME The qualitative component describes how the experience of vulnerability, despair and marginalisation or prejudice can negatively impact a patient’s

experience of MDR/RR-TB and gives insights into the use of alcohol or other substances. Conversely, patients and practitioners described a self-determination that, when matched with holistic and person-centred guidance from the practitioner team, seems to be what leads to patients in the programme achieving better acceptance and therefore outcomes. We have taken the definition of acceptability to be ‘a multi-faceted construct that reflects the extent to which people delivering or receiving a healthcare intervention consider it to be appropriate, based on anticipated or experienced cognitive and emotional responses to the intervention.’(Sekhon, Cartwright and Francis, 2018). Meaning, in this instance, that project acceptability for the majority of patients and practitioners is contingent on the person-centred approach realising the potential of the individual, which is then enhanced with provision of a

trusted support network to help navigate their treatment and challenges they face in their everyday lives. The support network would include those described as close loved ones by the patient and the more formal therapeutic interventions carried out by practitioners. Within the practitioner group good, effective guidance between health professionals was reported to optimise acceptance of the programme and lead to better outcomes for most patients. Overall therefore, the programme is acceptable for most patients in most circumstances. However, for some patients the programme could become more acceptable, notably for patients who use drugs other than alcohol, who have fractured or non-existent social support network, or who have a significant history of criminal behaviour which could potentially be indicators of undiagnosed and untreated personality disorder.

SUGGESTIONS FOR IMPROVEMENTS TO THE PROGRAMME POLICY

•

Uptake and scale up of the programme across Belarus and in countries with similar patients who exhibit disorders due to the use of alcohol and who are likely to be at risk of not completing MDR/RR-TB treatment Carry out an assessment to look into whether rolling out the programme to patients with drug sensitive (DS) TB, and suspected alcohol or substance use disorder would be worthwhile. Also consider which elements of the person-centred approach can be incorporated into DS-TB care. Assess whether the multi-disciplinary approach could be integrated with other diseases such as HIV, Hepatitis C and Heart disease given the high proportion of patients with co-morbidities and the struggles patients faced in doctors be-

• •

ing disinterested in medical concerns other than TB Creation of guidelines for the management of psychiatric disorders for TB patients Ensure patients are better involved in the clinical decision-making process. This could include improved discussions with patients about clinical issues in appointments, including helping them better understand the different professionals that are part of the multi-disciplinary approach. Furthermore, it could mean inclusion of patients (either expert patients or the patient in question) into the Consilium decision making process to ensure treatment decisions fully consider the patient’s point of view.

PRACTICE

•

Increase in group and family interventions within the programme, as

•

for the moment, only a small proportion of patients currently benefit from this. Family interventions can help reduce stress and stigma for the patients. Patients have previously expressed a lack of interest in group sessions, so staff could try to support the instigation of a patient lead or peer-to-peer support group. Patients reported learning from the experience of others around them, and practitioners suggested that for patients who refused to listen to health advice benefit from interventions with a group of practitioners at the same time. Adaptation of the programme for patients with suspected or confirmed personality disorders. Patients could be screened for personality disorder at entry into the programme, and then patients could receive an enhanced care

•

package. This could include an incentive or a token system, behavioural work or in-depth psychotherapy. Incentives or token work could include personalised bonuses for patients, even though this may require flexibility of staff and clear frames of agreement between MSF, MOH and the patient. More research may be required to adapt the programme for patients with personality disorders. Continue to avoid formal diagnoses of psychiatric disorders for patients, given the negative consequences of a diagnosis, and limited benefits – unless the lack of diagnosis means the patient cannot access treatment for their disorder. Increased utilisation of video DOT for patients with poor adherence, as well as those with good adherence. The strategy could be to give video DOT at the first session, so that a patient can immediately start looking for work, then DOT could only be withdrawn if it is not working. This could be combined with more home visits for patients who need it. Forced hospitalisation rate for patients should be further reduced, given the impact that forced hospitalisation of up to two years can have on a patient’s life, financial situation and wellbeing. Enhanced health promotion and contact tracing among drinking partners of people with TB, not simply family contacts as drink-

ing partner should be considered an extended part of the key social network. This should be carried out in a way that minimises stigma for the index case however.

INTERNAL PROJECT RECOMMENDATIONS (NOT IN SUMMARY VERSION):

•

The internal monitoring process of the project could be improved. This could be achieved by reinforcing the importance of regular systematic psychological tests to monitor the progress of the programme overall, even if the benefit of such tests to the individual patient or counsellor may be minimal. This can include automated reminders, counsellors setting dates with patients for follow up meetings and ensuring someone is responsible for data completeness. It could also include more regular feedback and monitoring amongst the team on ways to improve the project. Capture quantitative data on the presence of a support person during TB treatment going forwards to monitor whether patients have a support person, or are able to find one with the help of counsellors and social workers Reduce turnover or change of staffing from the patient perspective, so that patients are able to have more consistency in counsellors, doctors and nurses in order to

•

build a strong trusting relationship with them. A policy could be put in place that a patient should only have 2 counsellors over the course of the programme. For the multi-disciplinary approach to be more effective, the MSF doctors administering it should be given more responsibility for TB treatment decisions and monitoring. Alternative MOH doctors could be trained and empowered to take on the programmes approach. This would require more resourcing, to ensure the patient clinician ratio is small enough. Increased sensitisation for MOH MDR/RR-TB staff in Belarus to understand the high levels of adherence achieved by patients in the programme, combined with the ‘soft’ outcomes of helping patients to navigate the challenges in their lives. This could be best done in an unobtrusive form, such as through personal dialogue, informal dialogues, or joint home visits. Consider changing the role of counsellor-educators in the project to psychologists, so that more types of psychological intervention could be carried out during counselling sessions for patients who are currently under-treated. This might involve upskilling some of the current staff to be able to apply certain techniques, such as CBTpd (work with core beliefs and behavioural work)

CONCLUSION In this interim analysis, patients in this cohort, who face complex health and social issues, have very promising levels of adherence to MDR/RR-TB treatment. MDR/RR-TB outcomes however for this cohort look slightly lower than the Belarus average to date, which is possibly an indication that the programme simply takes patients with the highest risks of not being able to com-

plete treatment. Further research, with a controlled design, would be needed to ascertain whether the programme has a clear improvement on treatment outcomes. This study does however demonstrate that a person-centred approach using psycho-social support and harm reduction to support patients with MDR/RR-TB and alcohol use disorder is feasible and acceptable in this setting

and can have a positive impact for the patient. The holistic approach, which can help the patient solve a variety of personal and socio-economic problems, first may lead to an improvement in a patient’s quality of life, and secondly this could mean the patient is more able to be adherent to TB treatment. Patients appreciated having support from someone they were able to trust, who

treated them like a human being, and who could give them professional help with interpersonal problems, with administrative tasks, or with finding a job. This evaluation identified some areas of improvement for the person-centred approach, including adapting the programme to patients with suspected personality disorders, to minimise forced hospitalisation for these particularly

vulnerable patients, who are particularly challenging for health practitioners. More research would be needed to better understand how to adapt the programme for patients with personality disorders. Increased uptake of video DOTS could also be beneficial for many patients. We recommend that findings from this study are used to inform policy changes that encourage this mul-

ti-centred person-centred approach to be scaled up across Belarus, and to be incorporated as a national programme, with close M and E and with a scaling down of forced hospitalisation. Countries with similar challenges with TB treatment and patients who face complex social situations combined with disorders due to use of alcohol could also benefit from the approach.

REFERENCES Alena Skrahina and Hurevich H, Zalutskaya A, Sahalchyk E, Astrauko A, Hoffner S, et al. (2013) ‘Multidrug-resistant tuberculosis in Belarus: the size of the problem and associated risk factors’, Bull World Health Organ., 1;91(1)(36–45). Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3537245/ (Accessed: 31 December 2020). Bello, N. T. (2019) ‘Update on drug safety evaluation of naltrexone/bupropion for the treatment of obesity’, Expert Opinion on Drug Safety, 18(7), pp. 549–552. doi: 10.1080/14740338.2019.1618268. Bradley, E. H., Curry, L. A. and Devers, K. J. (2007) ‘Qualitative data analysis for health services research: developing taxonomy, themes, and theory’, Health Services Research, 42(4), pp. 1758–1772. doi: 10.1111/j.1475-6773.2006.00684.x. Craig, G. M. et al. (2017) ‘Tuberculosis stigma as a social determinant of health: a systematic mapping review of research in low incidence countries’, International journal of infectious diseases: IJID: official publication of the International Society for Infectious Diseases, 56, pp. 90–100. doi: 10.1016/j.ijid.2016.10.011. DOM, G. et al. (2006) ‘BEHAVIOURAL ASPECTS OF IMPULSIVITY IN ALCOHOLICS WITH AND WITHOUT A CLUSTER-B PERSONALITY DISORDER’, Alcohol and Alcoholism, 41(4), pp. 412–412. Dua, T. et al. (2011) ‘Evidence-based guidelines for mental, neurological, and substance use disorders in low- and middle-income countries: summary of WHO recommendations’, PLoS medicine, 8(11), p. e1001122. doi: 10.1371/journal. pmed.1001122. Fairbairn, C. E. et al. (2018) ‘A meta-analysis of longitudinal associations between substance use and interpersonal attachment security’, Psychological Bulletin, 144(5), pp. 532–555. doi: 10.1037/bul0000141. Huque, R. et al. (2020) ‘“Death is a better option than being treated like this”: a prevalence survey and qualitative study of depression among multi-drug resistant tuberculosis in-patients’, BMC Public Health, 20(1), p. 848. doi: 10.1186/s12889020-08986-x. Imtiaz, S. et al. (2017) ‘Alcohol consumption as a risk factor for tuberculosis: Meta-analyses and burden of disease’, European Respiratory Journal, 50. doi: 10.1183/13993003.00216-2017. Konkoly Thege, B. et al. (2017) ‘Relationship between interpersonal trauma exposure and addictive behaviors: a systematic review’, BMC psychiatry, 17(1), pp. e164–e164. Lingford-Hughes, A. R. et al. (2012) ‘BAP updated guidelines: evidence-based guidelines for the pharmacological management of substance abuse, harmful use, addiction and comorbidity: recommendations from BAP’, Journal of Psychopharmacology (Oxford, England), 26(7), pp. 899–952. doi: 10.1177/0269881112444324. Links, P. S. et al. (1995) ‘Borderline personality disorder and substance abuse: Consequences of comorbidity’, The Canadian Journal of Psychiatry, 40(1), pp. 9–14. Lönnroth, K. et al. (2008) ‘Alcohol use as a risk factor for tuberculosis - a systematic review’, BMC public health, 8, p. 289. doi: 10.1186/1471-2458-8-289. Lönnroth, K. et al. (2009) ‘Drivers of tuberculosis epidemics: the role of risk factors and social determinants’, Social Science & Medicine (1982), 68(12), pp. 2240–2246. doi: 10.1016/j.socscimed.2009.03.041.

Mathew, T. et al. (2006) ‘Causes of death during tuberculosis treatment in Tomsk Oblast, Russia’, The international journal of tuberculosis and lung disease : the official journal of the International Union against Tuberculosis and Lung Disease, 10, pp. 857–63. Mathew, T. A. et al. (2009) ‘Integration of alcohol use disorders identification and management in the tuberculosis programme in Tomsk Oblast, Russia’, European Journal of Public Health, 19(1), pp. 16–18. doi: 10.1093/eurpub/ckn093. Medecins Sans Frontieres, Operational Centre Amsterdam. (2020) ‘MSF OCA Strategic Plan, 2020-2023.’ NICE (2011) ‘Alcohol-use disorders: diagnosis, assessment and management of harmful drinking (high-risk drinking) and alcohol dependence’. NICE. Available at: https://www.nice.org.uk/guidance/cg115 (Accessed: 31 December 2020). Oladimeji, O. et al. (2016) ‘Psychosocial wellbeing of patients with multidrug resistant tuberculosis voluntarily confined to long-term hospitalisation in Nigeria’, BMJ Global Health, 1(3), p. e000006. doi: 10.1136/bmjgh-2015-000006. Rehm, J. et al. (2009) ‘The association between alcohol use, alcohol use disorders and tuberculosis (TB). A systematic review’, BMC public health, 9, p. 450. doi: 10.1186/1471-2458-9-450. Rodriguez, L. M., Neighbors, C. and Knee, C. R. (2014) ‘Problematic alcohol use and marital distress: An interdependence theory perspective’, Addiction Research & Theory, 22(4), pp. 294–312. doi: 10.3109/16066359.2013.841890. Rogers, C. R. (1946) ‘Significant aspects of client-centered therapy.’, American Psychologist, 1(10), pp. 415–422. doi: 10.1037/h0060866. Sekhon, M., Cartwright, M. and Francis, J. J. (2018) ‘Acceptability of health care interventions: A theoretical framework and proposed research agenda.’, British journal of health psychology, 23(3), p. 519. Sher, K. J. and Trull, T. J. (2002) ‘Substance use disorder and personality disorder’, Current Psychiatry Reports, 4(1), pp. 25–29. doi: 10.1007/s11920-002-0008-7. Stewart, David W, Shamdasani, Prem N., and Rook, Dennis: (2015) Focus Groups: Theory and Practice (Applied Social Research Methods). Available at: https://www.abebooks.com/9780761925828/Focus-Groups-Theory-Practice-Applied-0761925821/plp (Accessed: 31 December 2020). Trull, T. J. et al. (2018) ‘Borderline personality disorder and substance use disorders: an updated review’, Borderline Personality Disorder and Emotion Dysregulation, 5(1), p. 15. doi: 10.1186/s40479-018-0093-9. Tull, M. T. and Gratz, K. L. (2012) ‘The impact of borderline personality disorder on residential substance abuse treatment dropout among men’, Drug and Alcohol Dependence, 121(1), pp. 97–102. doi: 10.1016/j.drugalcdep.2011.08.014. Vi, R. et al. (2018) ‘The American Psychiatric Association Practice Guideline for the Pharmacological Treatment of Patients With Alcohol Use Disorder.’, The American Journal of Psychiatry, 175(1), pp. 86–90. doi: 10.1176/appi.ajp.2017.1750101. WHO (2019) WHO | WHO consolidated guidelines on drug-resistant tuberculosis treatment. World Health Organization. Available at: http://www.who.int/tb/publications/2019/consolidated-guidelines-drug-resistant-TB-treatment/en/ (Accessed: 2 February 2021). World Health Organisation (2020) ‘Global Tuberculosis Report 2020’. Available at: https://apps.who.int/iris/bitstream/ handle/10665/336069/9789240013131-eng.pdf (Accessed: 31 December 2020). World Health Organization (2018) Global status report on alcohol and health.

ANNEXES ANNEX 1: RISK FACTOR ASSESSMENT For general risk assessment, counsellors consider the following factors: Socio economic

Risk assessment

Level of social and family support Income/ financial stability/ employment situation

Low — moderate — high

History of incarceration Education level – understanding of the disease and treatment Housing/ living situation Psychiatric/ Psychologic patient related Self-efficacy

Low — moderate — high

Motivation Level Mental health/ psychiatric comorbidities Forgetfulness, Amnesia, Dementia

Yes — No

Disruptive sleep Suicidal ideation/ history Health-Care Distance to DOT corner Relationship to healthcare providers

Good — Bad

Experience with health care providers General availability of medication Treatment related Comorbidities Side-effects

Yes — No

Fatigue Length and complexity of treatment

AUD risk assessment is based on the following test results

Risk

Low

Moderate

High

AUDIT Score

≤7

8-15

≥ 16

Substance

Low

Moderate

High

Alcohol

0-10

11-26

≥ 27

Tobacco

0-3

4-26

≥ 27

Drugs

0-3

4-26

≥ 27

ANNEX 2 RECRUITMENT OF PARITICIPANTS FOR INTERVIEWS AND FOCUS GROUPS Details on practitioner recruitment and focus group discussion data collection process can be seen below ( Figure 1 and Figure 2)

Invitation email to gatekeepers (heads of department at MSF and MoH) to obtain list of potential participants and schedule FGD time by study investigator (project epidemiologist

Invitation to FGD sent by individual emails by study investigator (MSF psychiatrist)

Provide information sheet and consent form to potential participants via email (MSF psychiatrist)

Conduct FGDs via skype

Obtain consent forms via email

MSF psychiatrist to answer questions and invite to scheduled FGD via telephone

Enter and transcribe translate (anonymise) data

Analyse data

Share initial analysis for feedback with participants for verification

Archive analysed data and destroy according to protocol

Sharing results and write up

FIGURE 1. RECRUITMENT FLOW CHART FOR FOCUS GROUP DISCUSSION WITH HEALTH CARE PROVIDERS INVOLVED IN MULTI-DISCIPLINARY PERSON CENTRED-CARE AND HARM REDUCTION INTERVENTION Details of the purposive sampling can be seen below: Focus group A: MSF staff ALL MSF STAFF WHO CARE FOR PATIENTS WITH MDRTB AND SUBSTANCE ABUSE DISORDER WERE INVITED TO PARTICIPATE

2 TB DOCTORS

2 NURSES

5 COUNSELLORS

We sought voluntary consent from the above, and if possible, we will select 2 doctors, 2 nurses and 4 counsellors – this will be based on their role, the activities they carry out and their years of experience with MSF.

Focus group B: MOH staff The study investigators planned to invite the following participants to a separate focus group for MOH staff:

MOH STAFF WHO CARE FOR MDRTB PATIENTS

IPD

2 TB Doctors

2 Nurses

OPD (TB2 AND TB3)

1 Psychologist

1 Psychiatrist

2 Doctors

2 Nurses

1 counsellor

For IPD and OPD we selected a total of 4 each from the above. Selection was based on the person’s role, the activities they carry out and their years of experience with MDRTB patients.

ANNEX 3: IN-DEPTH INTERVIEW GUIDE FOR PATIENTS 1. Background The in-depth interview for people living with tuberculosis and participating in this feasibility study looking at acceptance of a person-centred care (PCC) programme to improve DR-TB treatment outcomes is carried out in a setting that minimises the influence of the researcher and encourages and stimulates an interviewee. The interview aims to encourage participants to speak about aspects of TB treatment within the social context of their adherence to treatment where specific support to substance abuse is integral to treatment support. The topic guide acts as an aide-memoire for the interviewee, outlining areas or topics of interest framed from results of the survey responses. Interviewees are free to expand on their discussion of any area, or to skip over one completely. In this way the aim of in-depth interviews is to understand more fully interviewee’s perceptions, beliefs, experiences and processes rather than to arrive at a ‘single truth’ across the cohort. 2. Introduction (5 mins max) • Thank the participant for agreeing to take part in this research • Introductions: explain who you are and share some of your experience • Create a relaxed atmosphere; offer the participant something to drink when this is possible • Tell each participant: “I (We) would like to talk to you about your experiences during TB treatment and to what extent the care and approach to difficulties you face is improving this experience. We would like to hear your stories, what you think is working well with regards to how you feel after treatment and the challenges you face. This interview will contribute to a better understanding of how people living with tuberculosis experience treatment in their social context. The interview will take approximately 45 - 60 minutes and can be stopped by you at any time without any consequences. If you would like to continue there is a form we have to complete to check that you have all the information you need before we start. Would you like to continue? • Make sure the participant has been informed and has consented verbally to participate in the research study • NOTE: Turn on the recorder and test it is recording (avoid placing cell phones close to recorder!) Visual aids alongside the topic areas may be used to stimulate perspectives of their health with regard to interviewees’ experience of TB treatment. 3. Domains:* • living with TB and substance use; impact on social life; • failures with treatment; social network support; • adherence events or acceptance toward treatment approaches. * ТDomains are chosen based on quantitative survey question framework but may be adapted post initial survey analysis.

Introduction

-Study aim -Why invited to participate -Consent

Domain

Prompts

Grand opening question: Tell me a little bit about yourself and how you came to be in this TB Programme.

[picking up on anything to be expanded in interviewee’s answer linked to main domains]

General health

Views on current general health how do they see their health and social life: what is their story of living with TB- what are their worries and how do they manage these worries about health- previous experience and expectation of TB treatment, what influences the experience – positive and negative?

Treatment journey

Treatment life and timetable – what dominates their sense of success to enduring TB each day, week- how often. How treatment and [specific support toward substance abuse] fits into everyday life (challenges, things that help accept and keep to treatment plan)?

Taking care of your health

Health-risk behaviours affecting TB treatment [prompts: (e.g. smoking; regular dependency on alcohol or other)]. Understanding of preventative health measures How is taking care of themselves managed – routine?

Social life and network of support

What does support feel like? Family support and home life-relationship to support and acceptance of treatment. Knowledge of carer?

Adherence events

Main concerns day to day about treatment regime – what might impact adherence that is positive or negative- [food tolerability of treatment; impact of side effect on treatment success; aspects that help or hinder?]

Acceptance of programme

Perceived benefits of treatment including all aspects of the person-centred care approach. Other health-promoting measures followed during TB treatment (prompts: e.g. foods, fads, traditions, food supplements).

4. Any additional information participant would like to share and add that was not mentioned so far Thank you again for participating in this interview, any questions?

ANNEX 4: FOCUS GROUP DISCUSSION GUIDE FOR PRACTITIONERS Introduction (5 mins max)

• • • •

Thank the participants for agreeing to take part in this research Introduce yourself Create a relaxed atmosphere Tell the group “I (We) would like to talk to you about the topics related to the “the person centred care programme for patients with drug resistant tuberculosis” run by Médecins Sans Frontières and the Ministry of health. We are interested in both your particular experiences working in the programme and outlook related to its implementation. This discussion will contribute to a better understanding of how the activities and processes of the comprehensive psychosocial and harm reduction intervention programme contribute to better outcomes for patients.

The interview will take approximately 45 - 60 minutes with anyone in the group deciding to leave at any time if they no longer want to contribute at any time without any consequences.

• •

Make sure the group participants have been informed about the study and have consented verbally to participate in the research NOTE: check consent for recording and test it is recording. Go through etiquette when using online platform (e.g. verbal responses, privacy, respect and non-judgment).

Introduction

- Study aim - Why invited to participate - Consent and respect within the group

History – please share any ideas you have about how this program came about and why different to other types of MDRTB programme (warmer – grand opening question)

Dif with global fund programme What is an outreach brigade?

Description of the programme

We would like you to talk about the programme Describe the different aspects and experience of these?

How is patient’s health and wellbeing looked after? What specific issues do you think are particular to the patients is it different for women and men

How may implementation feasibility impact on policy change toward this model of care

What does the relationship look like How are the two organisations working together Describe training support needed to run the programme Describe quality controls in place linked to procedures skills development

Here we are looking at practicalities – theory versus practice

Of the components described what happens in practice ? Staff ratios; multidisciplinary element; how is care model realised competencies needed ethics as part of practice safety and confidences

Challenges

What is and is not working

How do you manage challenges? What may help overcome theseTake examples of what support looks like (consistent follow up by one counsellor , family support networks) What is it like for patients- Can we do better in terms of consistent follow up of patients by someone they trust?

Treatment adherence

From your perspectives describe issues of adherence and how this approach might influence this (as in what affects whether a patient is able to adhere to treatment?

Look at responses from different roles (doctor, counsellor etc) Counsellor perspectives on shame

Success stories

What does success look like

Ask for descriptions of success (and failure) related to cases that attribute this programme model to that success Looking here for learning from challenges too

Support for better outcomes

Description from each practitioner or program manager experience of how to support better outcomes

Is there more we should / could be doing for patients who have problems other than alcohol dependence (especially personality disorders)?

Aspects of MSF and MoH

Feasibility

General observations moderator

ANNEX 5: ADVERSE EVENTS LIST OF ADVERSE EVENTS

колич.

Other

Hypomagnesemia

Aspartate Aminotransferase (AST or SGOT) Increased

Increased liver enzymes

Anemia

Prolonged QT Interval

Chronic Kidney Disease

Gamma-Glutamyltransferase (GGT) Increased

Hyperbilirubinemia (when other liver function are in the normal range)

Hypocalcemia (corrected for albumin)

Paresthesia (Burning, Tingling, etc.)

Platelets Decreased

Alanine Aminotransferase (ALT or SGPT) Increased

Depressed Level of Consciousness

Hearing impairment

Hyperkalemia

Optic Nerve Disorder

LIST OF SEVERE ADVERSE EVENTS

колич.

Other

Platelets Decreased

Prolonged QT Interval

Nurse passes admission information (patient file data, anamnesis) to Multi-Disciplinary MSF Team

Basic information for patient about patient centered care (Doctor, Nurse, Counselor, Psychiatrist, Social Worker)

Initial Assessment (day 1-5 after admission) Collection of basic psychosocial information (structured interview)

Support patient to adapt to environment, build trust, de-stress, answer basic questions

Basic screening for depression and anxiety (PHQ-9, GAD-7)

General decision if patient is eligible for psychosocial support

Risk Assessment (day 1-8 after initial assessment) Adherence risk assessment (structured interview)

AUD/ SUD Risk Assessment (Prescreen & Audit Questionnaire)

Determination of risk – and support level (Low – Moderate – High)

MoH/MSF MDM decision about PS support plan

Impl. MSF Fut. MoH

Intoxicated patients monitored by MoH staff, if necessary/ detoxprotocol

Impl. MSF Fut. MoH

Patient Admission to IPD

MoH Doctor, Nurse and MSF

ANNEX 6: PLANNED FREQUENCIES OF COUNSELLING BY BASELINE RISK LEVEL AND TREATMENT PLAN

Psychiatrist involvement (anti craving medication, psychiatric medication)

Follow up DR-TB Education

(Patients informed consent)

Referral to Intervention (max 4 weeks after admission) Physician/ Psychiatrist decides about medication (anti craving, psychiatr. drugs)

Establish Multi Disciplinary Intervention Team (TB-Doc., Psychiatrist, Counsellor, Social worker, Nurse)

Initial patient-centered intervention plan

Weekly update in MDM (MoH/ MSF)

Impl. MSF Fut. MoH

AUD/ SUD Education

MSF + MoH

Brief Intervention (during 3 weeks after admission)

DR-TB Education Health Education Workbooks

DR-TB Support Groups AUD Support Groups

Individual Counselling

AUD Education Harm Reduction Diaries

Anti-craving medication

Social Work support (Community activities)

Discharge Planning (4 weeks before e.o.t.)

Absconded Discharge (immediate notification) Disciplinary Discharge (notification + interventions about 1, 2, 3 warnings)

Ambulatory Care Intervention Team involving OPD health care practitioners

Continuous support through counselling and social work

Anti-craving medication if indicated on patient’s demand

Arrangement of Community support, Selfhelp groups…

Evaluation 6 monthly PS evaluation PHQ-9, GAD-7, ASSIST or AUDIT

Medical Evaluation

MSF + MOH

Handover to OPD Team (Doc. Counsellor, Nurse, Social worker) consilium, documents, medication

MSF + MOH

Информация о выписке

MSF + MOH

Discharge plan created by patient with intervention team

Intensified Social work & counselling support (Activity plan, Contact plan, Support agenda)

MSF

Interventions (will be implemented gradually)

Treatment Phase First 3 months of inpatient treatment (IPD, FHC) Rest of inpatient stay (IPD, FHC) First 30 days of ambulatory phase (OPD)

Intervention Type

Low Risk

Moderate Risk

High Risk

Counseling

1-2x per month

1x per week

Patient Contact

1x per month

1x per week

2x per week

Collateral MoH

1x per month

1x per week

2x per week

Counseling

1-2x per month

1x per 2 weeks

1x per week

Patient Contact

1x per month

1-2x per 2 weeks

2x per week

Collateral MoH

1x per month

1x per 2 weeks

2x per week

Counseling

1-2x per month

1x per week

Patient Contact

1x per month

1x per week

2x per week

Collateral MoH

1x per month

1x per week

2x per week

Ambulatory phase to completion of treatment (OPD)

Counseling

1-2x per month

1x per 2 weeks

1x per week

Patient Contact

1x per month

1x per 2 weeks

2x per week

Collateral MoH

1x per month

1x per 2 weeks

2x per week

Discharge phase; last 4 weeks of treatment (all facilities)

Social work counselling

2x per week

1x per week

On request

ANNEX 7: BOX PLOTS OF ADHERENCE LEVELS BY FACTORS ASSOCIATED WITH ADHERENCE

ANNEX 8: GRAPHS OF ADHERENCE OVER TIME BY FACTORS ASSOCIATED WITH ADHERENCE

EN

Articles inside

IDENTIFIED GAPS AND SUGGESTED PROGRAMME IMPROVEMENTS ��

SELF-DETERMINATION, GUIDANCE AND ACHIEVING A STATE OF HEALTH ��

TB TREATMENT ADHERENCE ��

BASELINE CHARACTERISTICS OF STUDY PATIENTS ��

TB TREATMENT OUTCOMES��

MENTAL HEALTH FOLLOW UP DATA ��

EN

Articles inside

IDENTIFIED GAPS AND SUGGESTED PROGRAMME IMPROVEMENTS ��������������������������������������������������������������������������

SELF-DETERMINATION, GUIDANCE AND ACHIEVING A STATE OF HEALTH ����������������������������������������������������������������

IDENTIFIED GAPS AND SUGGESTED PROGRAMME IMPROVEMENTS ��

SELF-DETERMINATION, GUIDANCE AND ACHIEVING A STATE OF HEALTH ��