This article was downloaded by: On: 24 November 2010 Access details: Access Details: Free Access Publisher Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 3741 Mortimer Street, London W1T 3JH, UK
AIDS Care
Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t713403300
Predicting quality of life for people living with HIV: international evidence from seven cultures
S. M. Skevingtonab; S. Norwegabc; M. Standagead; The WHOQOL HIV Group a WHO Centre for the Study of Quality of Life, University of Bath, Bath, UK b Department of Psychology, University of Bath, Bath, UK c Department of Clinical Psychology, University of Konstanz, Konstanz, Germany d School for Health, University of Bath, Bath, UK First published on: 14 May 2010
To cite this Article Skevington, S. M. , Norweg, S. , Standage, M. and The WHOQOL HIV Group(2010) 'Predicting quality
of life for people living with HIV: international evidence from seven cultures', AIDS Care, 22: 5, 614 — 622, First published on: 14 May 2010 (iFirst) To link to this Article: DOI: 10.1080/09540120903311466 URL: http://dx.doi.org/10.1080/09540120903311466
PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.
AIDS Care Vol. 22, No. 5, May 2010, 614 622
Predicting quality of life for people living with HIV: international evidence from seven cultures S.M. Skevingtona,b*, S. Norwega,b,c, M. Standagea,d and The WHOQOL HIV Group1 a WHO Centre for the Study of Quality of Life, University of Bath, Bath, BA2 7AY, UK; bDepartment of Psychology, University of Bath, Bath, BA2 7AY, UK; cDepartment of Clinical Psychology, University of Konstanz, Konstanz, Germany; dSchool for Health, University of Bath, Bath, BA2 7AY, UK
Downloaded At: 11:20 24 November 2010
(Received 13 February 2009; final version received 4 September 2009) The need for a validated quality of life (QOL) model focussing on people living with HIV/AIDS has led to an international re-evaluation and extension of the Chronic Illness Quality of Life model using complex latent modelling techniques. After reoperationalising six model variables and including independence and sex-life, the WHOQOL-HIV was administered to 1281 people with asymptomatic-HIV (42%), symptomatic-HIV (40%) or AIDS (18%; 34 years; 62% male) living in Australia, Brazil, India (north & south), Italy, Thailand and Ukraine. The overall model fit was acceptable. Social inclusion did not directly improve QOL, but increased positive feelings, social support and perceived improvements of access to health and social care; all three improved QOL. Social inclusion increased perceived physical health indirectly through positive feelings. Better physical health improved sex-life and gave greater independence; both improved QOL. Gender and disease stage models were acceptable, fitting best for men and asymptomatic-HIV. Similar aspects of QOL were depleted for women and some disease stages. Increased social support did not consistently improve independence or positive feelings. Positive feelings improved the sex-life of men and those with asymptomatic-HIV. This cross-cultural approach combining assessment with theory, could guide future international interventions and practice.
Keywords: quality of life; HIV; WHOQOL-HIV; CIQOL; cross-cultural; model
Despite growing research on the quality of life (QOL) of people living with HIV/AIDS (PLWHA), few models are available to guide disease management. Where models exist, they are largely concerned with HIV prevention. An exception is the Chronic Illness Quality of Life (CIQOL) model for PLWHA (Heckman, 2003) where life satisfaction (LS) was explained by AIDS-related discrimination. The model includes barriers to care, physical wellbeing, social support and engagement coping, but accounts for only one third of the variance in LS. Several observations are made about this work. First, the CIQOL model is a misnomer because only PLWHA were tested, so generalisation to other illnesses is problematic. Furthermore, although the model predicts LS, LS is just one component of QOL (Camfield & Skevington, 2008; Felce & Perry, 1996; Renwick, Brown, & Nagler, 1996). Therefore QOL needs to be tested within the model, or the title changed. Without evidence from other diagnoses, this should be specified as an HIV/AIDS QOL model. Second, Heckman’s US sample (n 275) was predominantly white (72%), so model cross-cultural applicability is questionable. Furthermore, it was not validated for different socio-demographic groups. Applicability to different sectors is essential to HIV *Corresponding author. Email: s.m.skevington@bath.ac.uk ISSN 0954-0121 print/ISSN 1360-0451 online # 2010 Taylor & Francis DOI: 10.1080/09540120903311466 http://www.informaworld.com
management e.g., women, in a field where men are focal. Third, as social constructs, CIQOL variables refer to subjective perceptions that cannot be directly observed and assessed, so using observed variables in structural equation modelling (SEM) (Structural equation modelling, 2005) is inappropriate. In such analyses, variables should be classified as latent; the five degrees of freedom reported for Heckman’s (2003) SEM indicate that latent constructs were not used. In this research, we adopt a full-latent SEM approach. Many QOL instruments for PLWHA were developed in the USA, Europe and South Asia, being translated from a source conceptualised by a different culture. Consequently new language versions often show lower semantic and conceptual equivalence, as evidenced by poorer psychometric properties (e.g., Spanish MOS-HIV, Xhosa SF-36; Skevington & O’Connell, 2003). Consequently, valid cross-cultural comparisons are problematic (Bowden & FoxRushby, 2003). International collaboration and new processes for developing the World Health Organisation Quality of Life Assessment (WHOQOL) instruments reduced equivalence problems by providing global concepts, items and protocols (Skevington, Sartorius, Amir, & the WHOQOL Group, 2004; WHOQOL Group, 1994, 1995, 1998a, 1998b), and
Downloaded At: 11:20 24 November 2010
AIDS Care these were applied in the WHOQOL-HIV (WHOQOL-HIV Group, 2003a, 2003b). The current research re-examines the CIQOL model using cross-cultural data from the WHOQOL-HIV obtained in seven cultures. In addition to reoperationalising existing variables, two QOL components relating to independence and sexlife were added to the model following literature reviews. Independence integrates mobility, daily activities, medication dependence and working capacity and is particularly salient when symptomatic-HIV progresses to AIDS. Mobility is essential to working capacity. As treatments improved, more PLWHA have re-entered the workforce. Despite discrimination, work is a means to survive, providing identity and status that improve QOL (Eller, 2001). Mobility is needed to fulfil daily activities e.g., visiting others, particularly for those living alone (Jacobson et al., 1997). Dependence on medication and treatment supports the maintenance of health and longevity (Kalichman & Catz, 2000). Sex-life could directly affect QOL, but is infrequently assessed for PLWHA (Skevington & O’Connell, 2003). Failure to assess QOL rather than sexual behaviour means that information that might have assisted adjustment to sexual difficulties remains scarce. Silence on this important basis of identity carries a potent message for those completing such questionnaires. Furthermore, sex-life is a sensitive indicator of wellbeing and health (Kalichman, 1998) so inactivity signals deterioration. The summarised advantages to using the WHOQOL-HIV permits the model proposed in this work to be validated using equivalent cross-cultural data. SEM analyses using a latent modelling approach was needed. Justified variables selected from a comprehensive international range could be added to enhance understanding and potentially improve model utility. Model generalisation over different populations needs validation for key socio-demographic and health status variables, and this is a new departure. If this data fit the model, then the convenient administration of a single, integrated multi-dimensional instrument would be pragmatic, promoting the dual use of model and measure by practitioners in the field. The present study aimed to test the fit of the CIOQOL model using cross-cultural data, following the reoperationalisation of variables and addition of two important WHOQOL-HIV dimensions: independence and sex-life. We also aimed to test the invariance of processes across gender groups and disease status models (asymptomaticHIV, symptomatic-HIV and AIDS).
615
Method Design Participating centres were located in Victoria, Australia; Bangalore and New Delhi, India; Naples, Italy; Porto Alegre, Brazil; Bangkok, Thailand and Dnipropetrovsk, Ukraine. National statistics were not always available to underpin representative sampling, so quotas guided collection: age (50% 30 years), gender (50% male) and HIV status (33% AIDS, HIV-symptomatic and HIV-asymptomatic). Most data were collected in primary care and hospital out-patient clinics. Asymptomatic-HIV participants knew they were infected and reported no symptoms. Symptomatic-HIV participants had minor symptoms/signs of the disease. People with AIDS had major signs e.g., weight loss, prolonged fever, Kaposi sarcoma, meningitis and TB (O’Connell, Saxena, & Skevington, 2004). Measure The WHOQOL-HIV is a subjective self-assessment of QOL, containing 120 items: 100 generic items (25 facets) from the WHOQOL-100 and a 20 HIVspecific items (five facets): HIV symptoms, social inclusion, death and dying, forgiveness, fear of the future (WHOQOL-HIV Group, 2003a, 2003b). The facets are scored in one of six domains: physical health, psychological, level of independence, social relationships, environment and spiritual, religious and personal beliefs. Psychometric properties showed good factorial validity (e.g., Comparative Fit Index (CFI) 0.97), and valid discrimination between sick and well groups. The present study presents results from a secondary analysis of international field test data (O’Connell et al., 2004). The original model was reoperationalised using measured dimensions from the WHOQOL-HIV. Some mapped closely onto CIQOL concepts: a social inclusion facet assessed AIDS discrimination; a health and social care facet addressed barriers to care, a physical health domain represented physical wellbeing and a social support facet measured social support. Others required minor adjustments to concept and/or measurement: a general facet on overall QOL and health replaced LS. As engagement coping referred to the process not outcomes of coping (unlike all other outcome variables in the CIQOL) it was replaced by a positive feelings facet. Although how people cope affects decisions about outcomes, this differs from having coped as a QOL outcome. Furthermore happiness is known as an outcome of good coping and emotional wellbeing (Fredrickson, 2002; Fredrickson & Joiner, 2002; Huppert & Whittington, 2003), so it
616
S.M. Skevington et al.
was hypothesised that positive feelings would directly affect general QOL, physical health and sexlife, and indirectly affect independence via physical health.
Downloaded At: 11:20 24 November 2010
Data analysis Centre data were merged following assessment of normality; some variables showed acceptable positive skewness. Univariate analysis of variance (ANOVA; pB0.01) investigated model variable differences between gender and disease groups (Scheffe p B0.05). Pearson correlations (r) between predictor variables and general QOL examined consistency across groups (p B0.01), then proposed models analysed using AMOS (v 6.0). Non-normal distributions (Mardia’s coefficient) required a maximum likelihood estimation method with bootstrapping (Byrne, 2001). Indexes of fit, and criteria for a well-specified model were CFI ( 0.90 acceptable; 0.95 good); standardised root mean square residual (SRMR; B0.08); root mean square error of approximation (RMSEA; B0.06; Hu & Bentler, 1999). Squared multiple correlations (SMC) show the variance explained by independent (IV) variable(s) for the dependent (DV) variable (QOL). Indirect effects representing the mediated effect of a given variable in the IV DV relationships, were explored. Multi-sample invariance analysis tested for equality of constraints across gender and disease stage models. Differences in absolute and incremental fit indices investigated invariance (Little, 1997), and changes in CFI 5 0.01 between increasingly more constrained models demonstrate this (Cheung & Rensvold, 2002).
asymptomatic-HIV. PWA perceived the poorest access to health and social care, physical QOL and social support. Symptomatic participants reported poorer physical QOL and social support than asymptomatic. Asymptomatic participants reported the best QOL overall, positive feelings, independence and sex-life. Across disease status and gender groups, overall QOL was consistently, positively associated with social inclusion, access to care, physical health, independence, sex-life, positive feelings and social support. Despite differences, predictor model variables were all highly, positively correlated with overall QOL in different groups, allowing model testing for the total sample. Structural equation modelling (SEM) First, we tested the original model pathways (Heckman, 2003) using the reoperationalised WHOQOL-HIV variables. Although all specified paths were significant (z 1.96), the model only approached an adequate fit to the data [(x2 (242) 1918.72; CFI 0.88; SRMR 0.09; RMSEA 0.074 (90% CI 0.071 0.077)]. Fit indexes approached an acceptable fit, but there was room for model improvement. After including two additional variables; sex-life and independence, the retested model was an acceptable fit [(x2 (449) 2536.73; CFI 0.90; SRMR 0.06; RMSEA 0.061 (90% CI 0.058 0.063)] (Figure 1). Including these additional constructs within the model accounted for an additional 4% of the variance in QOL scores. Direct and indirect effects for the international sample
Results Sample The total sample contained 1281 PLWHA; 62% men, mean age 33.6 years (SD 9.3; 17 71). Australia contributed 253 (93% men); Brazil 244 (36% men); Italy 151 (70% men); Thailand 82 (60% men); Ukraine 300 (50% men), Bangalore, India 201 (62% men) and New Delhi, India 50 (72% men). About 42% were asymptomatic (32% Italy 60% New Delhi); 40% symptomatic (30% Australia & Italy 60% Brazil); 18% had AIDS (3% Brazil 38% Italy). Preliminary analyses Physical QOL was better for women than men (Table 1), but disease status groups differed on all variables: people with AIDS (PWD) and symptomatic-HIV reported lower social inclusion than
Tested pathways from the original CIQOL model confirmed that social inclusion did not directly predict overall QOL, supporting Heckman’s (2003) findings. Nor did social inclusion predict better physical health, as expected. Social inclusion did however positively predict social support, and more strongly with the present data. Furthermore, social support predicted better overall QOL and health. Greater social inclusion predicted better access to health and social care and as access increased, so did overall QOL. Also perceived access to care positively predicted social support, leading to better positive feelings. Positive feelings predicted overall QOL. Moderately strong pathways explaining the effect of social inclusion on social support, access to care and positive feelings, provide new findings, as engagement coping was not previously significant in this relationship. New findings relate to the addition of two novel variables in the model, showing that a better sex-life
AIDS Care
617
Table 1. Model variable differences and correlations with overall QOL relating to HIV status and gender groups. HIV asymptomatic (n 533)
HIV symptomatic (n 496)
Social inclusion M (SD) r
3.38 (0.76) 0.49*
3.15 as (0.75) 0.49*
3.07 as (0.85) 0.50*
Health & social care M (SD) r
3.26 (0.71) 0.34*
3.17 (0.76) 0.35*
Physical health M (SD) r
3.63 (0.66) 0.57*
Social support M (SD) r
Male (n 793)
Female (n 481)
16.94**
3.28 (0.83) 0.56*
3.17 (0.68) 0.39*
6.35
3.00 as, s (0.78) 0.45*
9.62**
3.22 (0.77) 0.47*
3.12 (0.72) 0.18*
4.8
3.13 as (0.64) 0.69*
2.87 as, s (0.66) 0.68*
105.46**
3.22 D (0.75) 0.70*
3.35 (0.78) 0.45*
3.18 as (0.73) 0.46*
2.96 as, s (0.76) 0.54*
19.68**
3.22 (0.82) 0.52*
3.20 (0.73) 0.38*
0.15
Positive feeling M (SD) r
3.04 (0.76) 0.66*
2.66 as (0.81) 0.60*
2.66 as (0.70) 0.61*
38.93**
2.86 (0.79) 0.68*
2.77 (0.72) 0.59*
4.24
Independence M (SD) r
3.66 (0.82) 0.59*
3.03 as (0.76) 0.57*
2.89 as (0.76) 0.69*
111.86**
3.27 (0.86) 0.68*
3.29 (0.85) 0.58*
0.26
Sex life M (SD) r
2.93 (0.85) 0.48*
2.69 as (0.81) 0.45*
2.68 as (0.82) 0.40*
13.37**
2.75 (0.89) 0.50*
2.87 (0.75) 0.41*
6.31
General QOL M (SD)
3.24 (0.78)
2.81 as (0.76)
2.72 as (0.84)
54.07**
3.03 (0.87)
2.92 (0.71)
5.43
Downloaded At: 11:20 24 November 2010
Variable
AIDS (n 227)
F
F
3.44 D (0.64) 19.22** 0.63*
*pB0.01. **pB0.001. Note: r, Pearson’s correlation with QOL; SD, standard deviation; D, pairwise comparison differ at pB0.01; as, differs from asymptomatic at B0.05; s, differs from symptomatic at B0.05. Only one variable is indicated in each case in order of appearance.
and greater independence positively predicted overall QOL. However, as expected, social inclusion did not directly influence either. Better physical health predicted independence, and a moderately better sex-life. Furthermore, positive feelings positively predicted sex-life. Adding these new components provides a more complete and detailed picture of how physical health relates to QOL. Two moderately strong pathways confirmed by the present study were not confirmed by Heckman (2003); social inclusion predicted positive feelings, and positive feelings predicted physical health. In contrast, the present data did not confirm Heckman’s findings that social inclusion and access to care independently predicted physical health, or that physical health predicted QOL. Two pathways predicted but not confirmed by either study were that physical health predicted social support, and access to care predicted positive feelings.
An indirect relationship between social inclusion and overall QOL showed that this was primarily through three distinctive pathways, namely through social support via perceived access to health and social care, and through positive feelings (Table 2 and Figure 1). The expected direct effect between social inclusion and physical health was instead, primarily accounted for by indirect effects through positive feelings. Social inclusion indirectly affected each additional component: it predicted independence levels through social support, and sex-life both through improved positive feelings and better subjective physical health (Figure 1). Positive feelings influenced both sex-life and independence through the same common mediating effects of physical health. Although it was expected that better subjective physical health would predict QOL, this was only indirect and primarily occurred through two pathways: physical health strongly led to more independence, and to a better sex-life.
618
S.M. Skevington et al. 0.06 (0.02) 0.78 0.63
0.21
Positive feelings (SMC = 0.39)
Social support (SMC = 0.60)
Level of independence (SMC = 0.96)
0.16 (0.07)
0.39 (0.04) 0.14 (0.03)
0.36 (0.03)
0.62 (0.03)
0.44
0.50 (0.06)
62 (0.04)
Quality of life (SMC = 0.81)
0.96 (0.01)
0.19 (0.04) Social inclusion
0.26 (0.04) 0.46 (0.03)
0.79 0.88
Health and social care (SMC = 0.22) Downloaded At: 11:20 24 November 2010
0.84 0.12 (0.03)
0.17 (0.03)
Sex life (SMC = 0.29)
Physical health (SMC = 0.38)
0.40 (0.04)
Figure 1. Standardised solution for the revised model using the WHOQOL-HIV. Note: For visual simplicity, factor indicators and their respective errors are not reported, but are available from the corresponding author. All standardised estimates are signiďŹ cant (p B0.05). The bootstrap estimate of the standard error for each parameter is shown in parenthesis, whereas the proportion of the variance explained for each dependent variable is denoted in each endogenous variable by its squared multiple correlation (SMC) value.
Gender Table 2. Standardised parameter estimates of indirect effects for total sample. Parameter Indirect effects Social inclusion Social inclusion Social inclusion Social inclusion Social inclusion Social inclusion Health & social care Health & social care Health & social care Health & social care Health & social care Social support Social support Social support Social support Positive feelings Positive feelings Positive feelings Physical health
Effect Social support Positive feelings Physical health Sex life Level of independence QoL Positive feelings Physical health Sex life Level of independence QoL Physical health Sex life Level of independence QoL Sex life Level of independence QoL QoL
0.12 0.12 0.38 0.27 0.41 0.60 0.04 0.03 0.02 0.04 0.07 0.10 0.07 0.10 0.13 0.25 0.59 0.26 0.39
Note: All standardized indirect effects are significant (pB0.05).
Equivalence between gender groups was tested via independent QOL models for men and women. The baseline model showed an acceptable fit [(x2 (899) 3235.82; CFI 0.89; SRMR 0.05; RMSEA 0.045 (90% CI 0.044 0.047)] (Table 3). Baseline models Table 3. Standardised parameter estimates for men and women. Path
Men
Women
Social inclusion to health & social care Social inclusion to social support Health & social care to social support Social inclusion to positive feelings Social support to positive feelings Positive feelings to physical health Physical health to level of independence Social support to level of independence Physical health to sex life Positive feelings to sex life Positive feelings to quality of life Level of independence to quality of life Social support to quality of life Health & social care to quality of life Sex life to quality of life
0.58 0.64 0.23 0.56 0.17 0.67 0.95 0.09 0.38 0.25 0.44 0.33 0.09 0.16 0.11
0.21 0.57 0.36 0.37 0.12* 0.52 0.99 0.02* 0.41 0.11* 0.29 0.45 0.28 0.16 0.15
*Path not significant (z B1.96).
AIDS Care
619
Table 4. Results of the SEM multi-sample invariance analysis for the modiďŹ ed CIQL across gender groups. Model tested Step Step Step Step
1 2 3 4
x2
df
Dx2
CFI
SRMR
3235.82 3370.36 3397.75 3424.25
899 923 935 936
134.54* 27.39* 26.50*
0.89 0.89 0.89 0.89
0.05 0.06 0.07 0.07
RMSEA (90% CI) 0.045 0.046 0.046 0.046
(0.044 0.047) (0.044 0.047) (0.044 0.047) (0.044 0.048)
Downloaded At: 11:20 24 November 2010
*pB0.05. Note: Step 1, baseline; Step 2, measurement weights constrained; Step 3, structural weights constrained; Step 4, structural covariances constrained.
Independent results from each disease stage model supported all significant pathways between social inclusion and social support, access to care and positive feelings, plus all five pathways predicting overall QOL in Figure 1. There were a few expected differences between stages. Except during AIDS, social support did not predict positive feelings, but social support and sex-life predicted overall QOL, nor did social support predict independence in symptomatic participants, like other stages. Positive feelings only predicted sex-life in asymptomatic participants.
for partial invariance of gender models did not substantially depart from the total sample model (Table 4). All significant paths in the total model were significant for men; for women, three of these were not significant (Table 3). In women, positive feelings and independence did not predict increased social support, although social support predicted overall QOL. Positive feelings did not predict sex-life, but did predict QOL. Health care access predicted more social support for women than men (Table 3). Social inclusion more strongly predicted access to care and positive feelings for men than women.
Discussion HIV status
With 33 million people living with HIV globally, improving QOL is a priority for international organisations (UNAIDS & WHO, 2007; WHO & UNAIDS, 2000). However few models exist to predict the QOL of PLWHA, so the CIQOL (Heckman, 2003) is a valuable addition to knowledge. The present study evaluates internationally important factors in living with HIV that affect QOL, by reoperationalising and extending this model and improving the modelling. Even with more complex latent modelling, an
Differences between baseline models for three disease status groups were compared with results approaching an adequate fit to the data [(x2 (1347) 3704.99; CFI 0.88; SRMR 0.07; RMSEA 0.037 (90% CI 0.036 0.039)] (Table 5). Across samples, several pathways were not significant, so were unconstrained in the subsequent multi-sample invariance analysis with partial invariance supported across groups (Table 6).
Table 5. Standardised parameter estimates for the disease status samples. Path Social inclusion to health & social care Social inclusion to social support Health & social care to social support Social inclusion to positive feelings Social support to positive feelings Positive feelings to physical health Physical health to level of independence Social support to level of independence Physical health to sex life Positive feelings to sex life Positive feelings to quality of life Level of independence to quality of life Social support to quality of life Health & social care to quality of life Sex life to quality of life *Non-significant path (zB1.96).
HIV no symptoms
HIV symptoms
AIDS
0.49 0.63 0.22 0.46 0.17* 0.53 0.95 0.06 0.27 0.30 0.40 0.32 0.18 0.12 0.16
0.42 0.55 0.30 0.58 0.00* 0.62 0.99 0.00* 0.55 0.06* 0.39 0.38 0.18 0.20 0.09
0.49 0.72 0.19 0.37 0.38 0.61 0.87 0.17 0.41 0.13* 0.29 0.47 0.07* 0.20 0.09*
620
S.M. Skevington et al.
Table 6. Results of the SEM multi-sample invariance analysis for the CIQOL across disease status samples. Model tested Step Step Step Step
1 2 3 4
x2
df
Dx2
CFI
SRMR
3704.99 3805.25 3841.69 3849.95
1347 1395 1415 1417
100.27* 36.44* 8.26*
0.88 0.88 0.88 0.88
0.07 0.08 0.08 0.08
RMSEA (90% CI) 0.037 0.037 0.037 0.037
(0.036 0.039) (0.036 0.039) (0.036 0.038) (0.036 0.038)
Downloaded At: 11:20 24 November 2010
*pB0.05. Note: Step 1, baseline; Step 2, measurement weights constrained; Step 3, structural weights constrained; Step 4, structural covariances constrained.
acceptable fit was found providing a theoretical underpinning for developing QOL interventions and a means to test them. Testing a ‘‘universal’’ model was possible due to data available from seven diverse cultures from a quality cross-cultural measure with good equivalence between languages. Furthermore, the WHOQOL-HIV manual can be used to derive further language versions to investigate the impact of HIV on QOL in infected communities where this is unknown. Introducing two important QOL dimensions of sex-life and independence, and replacing LS with QOL per se, contributed to the model fit being acceptable. Although social inclusion did not predict QOL, social inclusion predicted more positive feelings, which in turn, predicted better overall QOL and health. The present model supports the view that happiness buffers the negative effects of social exclusion on QOL, although Heckman (2003) did not find that AIDS discrimination affected coping engagement. Positive coaching techniques e.g., expressing gratitude (Seligman, 2008) offer new strategies to defray the negative impact of exclusion. Twin mechanisms represent buffers against the negative impact of social exclusion on QOL. Social inclusion reduced barriers to accessing quality care and also improved social support; both mechanisms subsequently improved QOL. Furthermore, there were strong mutually reinforcing positive interrelationships between all these variables. Interventions to counter stigma may therefore need to actively and jointly harness resources from informal support and formal health care to enhance life quality. There were several unexpected findings, but these departures may be as much connected with the international perspective adopted here, as with sampling issues, as most published studies have adopted a mono-cultural and usually Western approach. This cross-cultural data showed that physical QOL played a more minor role in explaining the relationship between social inclusion and QOL, and physical health perceptions were only indirectly affected by social inclusion. Instead, inclusion enhanced positive feelings, which
improved perceived physical health (see Carver et al., 1993; Fredrickson, 2002; Taylor, Kemeny, Reed, & Aspinwall, 1991). Unexpectedly, physical QOL did not directly improve overall QOL, but facilitated greater independence and a better sex-life. Taking account of both additional dimensions improved QOL prediction. In line with Thomas et al. (2005), our findings indicate the importance of addressing a diversity of QOL domains in stigma interventions, not physical health exclusively. Models of gender and disease stage showed only minor departures from the total sample, and for men, QOL was almost identical. However, gender makes a difference (Green, 1996) as three predominantly psychosocial pathways were not significant for women. Although often reported elsewhere (e.g., Patrick, Cottrell, & Barnes, 2001), social support did not improve women’s QOL; nor did women derive independence and positive feelings from social support, or happiness from their sex-lives. Without longitudinal data it is not known whether several key QOL resources are shut down or withdrawn, depleting the richness of women’s lives. As large female samples were recruited in Brazil and Ukraine, this suggests that previous findings may be more culturebound than formerly appreciated. A similar pattern was found when disease stages were compared, where asymptomatic-HIV best fit the model. While cross-sectional data prevents reliable causal conclusions, psychosocial issues appeared to be adjusted with disease progression. In pre-AIDS stages, social support does not predict positive feelings, but does improve QOL. However those with AIDS derived positive mood from social support, even though support no longer improved QOL and health. Here the models underscore previous findings about the importance of social support throughout the illness. However provision is complex as PLWHA often have disproportionately fewer significant others than usual, as the providers of front-line support may be dying from HIV themselves (Vedhara & Folkman, 2000). However social support seems to disrupt independence once symptoms
Downloaded At: 11:20 24 November 2010
AIDS Care appear. Social interventions designed to connect PLWHA to empathic sources of informational, instrumental and emotional support through ‘‘buddying’’ programmes and internet communities (Burrage & Demi, 2003) could incorporate this. Furthermore, greater happiness only lead to a better sex-life when asymptomatic, suggesting a disrupted mechanism as symptoms develop. The current study model fit the data well supporting most direct effects that were medium to large. Including two additional facets improved the model fit. However there are limitations. Two pathways were predicted but unconfirmed by either study suggesting redundancy in the model. Insufficient data from Thailand and Delhi precluded a reliable cross-cultural test of this complex model, and a culture gender interaction. The effect of substituting engagement coping is unknown and deserves further investigation. Quota rather than representative sampling, and small proportions of AIDS and women participants were problems. Model generalisation may be limited by increased availability of ART since 2004, affecting health care access e.g., Brazil (UNAIDS & WHO, 2003). New centres could cross-validate this international model; sub-Saharan Africa is a priority, and the WHOQOL-HIV is in development in Zimbabwe and Zambia.
Note 1. The WHOQOL-HIV group comprises a coordinating group of collaborating investigators in the field sites, and a panel of consultants. Dr. S. Saxena directed the project which was initiated by Dr. R. Billington and Dr. J. Orley. Technical assistance was given to the project by Ms M. Lotfy and Dr. K. O’Connell. The field work reported here was carried out in the following field centres: Mr M. Bartos, Centre for the Study of Sexually Transmissible Diseases, La Trobe University, Victoria, Australia; Dr. P. Chandra, National Institute of Mental Health and Neuroscience (NIMHANS), Bangalore, India; Dr. R. Bhargava, Department of Psychiatry, All India Institute of Medical Sciences, New Delhi, India; Prof. F. Starace, Consultation and Behavioural Epidemiology Service, Naples, Italy; Dr. M. Fleck, Department of Psychiatry and Legal Medicine, University of the State of Rio Grande do Sul, Porto Alegre, Brazil; Dr. K. Meesapya, Branch of Preventive Mental Health, Department of Mental Health, Ministry of Public Health, Bangkok, Thailand and Dr. S. Pkhidenko, Dniepropetrovsk State Medical Academy, Dniepropatrovsk, Ukraine.
621
References Bowden, A., & Fox-Rushby, J. (2003). A systematic and critical review of the process of translation and adaptation of generic health-related quality of life measures in Africa, Asia, Eastern Europe, the Middle East, and South America. Social Science and Medicine, 57(7), 1289 1306. Burrage, J., & Demi, A. (2003). Buddy programs for people infected with HIV. Journal of the Association of Nurses for AIDS Care, 14(1), 52 62. Byrne, B.M. (2001). Structural equation modeling with AMOS basic concepts, applications, and programming. NJ and London: Lawrence Erlbaum. Camfield, L., & Skevington, S.M. (2008). On subjective well-being and quality of life. Journal of Health Psychology, 13(6), 764 775. Carver, C.S., Pozo, C., Harris, S.D., Noriega, V., Scheier, M.F., Robinson, D.S., et al. (1993). How coping mediates the effect of optimism on distress: A study of women with early stage breast cancer. Journal of Personality and Social Psychology, 65, 375 390. Cheung, G.W., & Rensvold, R.B. (2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling, 9, 233 255. Eller, L.S. (2001). Quality of life in persons living with HIV. Clinical Nursing Research, 10, 401 423. Felce, D., & Perry, D. (1996). Exploring current conceptions of quality of life - a model for people with and without disabilities. In R. Renwick, I. Brown, & M. Nagler (Eds.), Quality of life in health promotion and rehabilitation: Conceptual approaches, issues, and applications (pp. 51 62). Thousand Oaks, CA: Sage. Felce, D., & Perry, J. (1996). Exploring current conceptions of quality of life. In R. Renwick, I. Brown, & M. Nagler (Eds.), Quality of life in health promotion and rehabilitation (pp. 51 62). Thousand Oaks, CA: Sage. Fredrickson, B.L. (2002). Positive emotions. In C.R. Snyder & S.J. Lopez (Eds.), Handbook of positive psychology (pp. 120 134). Oxford: Oxford University Press. Fredrickson, B.L., & Joiner, T. (2002). Positive emotions trigger upwards spirals toward emotional well-being. Psychological Science, 13, 172 175. Green, G. (1996). Stigma and social relations of people living with HIV: Does gender make a difference? In C. Hankins & L. Sherr (Eds.), AIDS as a gender issue: Psychosocial perspectives (pp. 46 63). Philadelphia, PA: Taylor & Francis. Heckman, T.G. (2003). The chronic illness quality of life (CIQOL) model: Explaining life satisfaction in people living with HIV disease. Health Psychology, 22, 140 147. Hu, L., & Bentler, P.M. (1999). Cut-off criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1 55. Huppert, F.A., & Whittington, J.E. (2003). Evidence for the independence of positive and negative well-being:
Downloaded At: 11:20 24 November 2010
622
S.M. Skevington et al.
Implications for quality of life assessment. British Journal of Health Psychology, 8, 107 122. Jacobson, D.L., Berzon, R.A., Revicki, D.A., van der Horst, C., Fichtenbaum, C.J., Saag, M.S., et al. (1997). The effect of mode of administration on medical outcomes study health ratings and EuroQol scores in AIDS. Quality of Life Research, 6, 3 10. Kalichman, S.C. (1998). Understanding AIDS: Advances in research and treatment. Washington, DC: American Psychological Association. Kalichman, S.C., & Catz, S.L. (2000). Stressors in HIV infection. In H. Nott & K. Vedhara (Eds.), Psychosocial and biomedical interaction in HIV infection (pp. 31 60). Amsterdam, The Netherlands: Harwood Academic. Little, T.D. (1997). Mean and covariance structure (MACS) analysis of cross-cultural data: Practical and theoretical issues. Multivariate Behavioral Research, 32, 53 76. O’Connell, K.A., Saxena, S., & Skevington, S.M. (2004). WHOQOL-HIV for quality of life assessment among people living with HIV and AIDS: Results from the field test. AIDS Care, 16, 882 889. Patrick, J.H., Cottrell, L.E., & Barnes, K.A. (2001). Gender, emotional support, and well being among the rural elderly. Sex Roles, 45(1 2), 15 29. Seligman, M.E.P. (2008). Positive health. Applied Psychology, 57, 3 18. Skevington, S.M., & O’Connell, K.A. (2003). Measuring quality of life in HIV & AIDS: A review of the recent literature. Psychology and Health, 18(3), 331 350. Skevington, S.M., Sartorius, N., Amir, M., & the WHOQOL Group. (2004). Developing methods for assessing quality of life in different cultural settings: The history of the WHOQOL instruments. Social Psychiatry & Psychiatric Epidemiology, 39(1), 1 8. Structural Equation Modeling (2005). Retrieved www2. chass.ncsu.edu/garson/pa765/structure.htm Taylor, S.E., Kemeny, M.E., Reed, G.M., & Aspinwall, L.G. (1991). Assault on the self: Positive illusions and adjustment to threatening events. In J. Strauss & G.R. Goethals (Eds.), The self: Interdisciplinary approaches (pp. 239 254). New York: Springer.
Thomas, B.E., Rehman, F., Surayanarayanan, D., Josephine, K., Dilip, M., Dorairaj, V.S., et al. (2005). How stigmatizing is stigma in the life of people living with HIV: Positive individuals from Chennai South India. AIDS Care, 17(7), 795 801. UNAIDS & WHO (2003). Joint United Nations Programme on HIV/AIDS. AIDS Epidemic Update 2003. Retrieved from www.unaids.org/wad2003/report.html UNAIDS & WHO (2007). Joint United Nations Programme on HIV/AIDS. AIDS Epidemic Update 2007. Retrieved www.unaids.org/wad2007/report.html Vedhara, K., & Folkman, S. (2000). Psychosocial determinants of emotional well-being in HIV infection. In K.H. Nott, & K. Vedhara (Eds.), Psychosocial and biomedical interactions in HIV infection (pp. 137 163). Amsterdam, the Netherlands: Harwood Academic. WHO & UNAIDS (2000). Key elements in HIV/AIDS care and support. World Health Organisation working document. Geneva: Author. WHOQOL Group (1994). Development of the WHOQOL: Rationale and current status. International Journal of Mental Health, 23(3), 24 56. WHOQOL Group (1995). The World Health Organisation Quality of Life assessment (WHOQOL): Position paper from the World Health Organisation. Social Science & Medicine, 41(10), 1403 1409. WHOQOL Group (1998a). Development of the WHOQOL-BREF Quality of Life Assessment. Psychological Medicine, 28, 551 558. WHOQOL Group (1998b). The World Health Organization Quality of Life Assessment (WHOQOL): Development and general psychometric properties. Social Science & Medicine, 46(12), 1569 1585. WHOQOL-HIV Group (2003a). Initial steps to developing the World Health Organization’s Quality of Life Instrument (WHOQOL) module for international assessment in HIV/AIDS. AIDS Care, 15, 347 357. WHOQOL-HIV Group (2003b). Preliminary development of the World Health Organisation’s Quality of Life HIV instrument (WHOQOL-HIV): Analysis of the pilot version. Social Science & Medicine, 57, 1259 1275.