Some first food for thought: selected
baseline findings from the Effective Aids
Treatment and Support in the Free State
  ...
Acknowledgement
• The financial support of The World Bank’s Research
  Committee, its Development Economics Research
  Gro...
Background (1)
• The Effective Aids treatment and support in the
  Free State (FEATS) study aims to:

  – Inform the effec...
Study design (1)
• Prospective cohort study

• Experimental design is a combination of:
  – Group time-series, quasi- or f...
Study design (2)
• ART nurses working at 12 phase-I ART clinics
  across five Free State districts recruited public
  sect...
Study design (3)

          Group A:                            Group B:

  216 ART households receiving          216 ART ...
Study design (4)
• Ethical clearance: Faculty of Health Sciences,
  University of the Free State

• Data collection (basel...
Figure 2: Study design (FEATS)

     Comparison households                             Comparison households



          ...
Three selected, potentially
     important questions…

– Nutritional status and food security

– Need and access @ support...
Background
… what challenges do food insecurity and malnutrition, worsened by the present
economic and food crises, pose i...
Figure 3: Adult nutritional status, by comparison group
    100%

                   17.2                           19.5  ...
Table 3: Correlates of malnourishment among adults
     Explanatory variable                      OR                  95% ...
Table 2: Nutritional status across ‘treatment career’

                        Baseline in      Pre-ART        Household  ...
Figure 4: Household food security, by comparison group
   60


                                                           ...
Figure 5: Adult and child food security, by group
  100%


   90%                                                         ...
Table 3: Correlates of household food security
                                                            Overall        ...
Concluding comments
• An integrated, comprehensive response to HIV
  and AIDS, requires effective, sustainable and
  scala...
Background (1)
• An integrated, multi-faceted response is required for an
  effective, sustainable ARV treatment programme...
Background (2)
• Rhine (2009: 369) points out that support groups in the
  context of the African HIV epidemic are more th...
Figure 3a: Access to support groups, treatment
           supporters and community health workers

                       ...
Figure 4: Access, need, cost and willingness-to-pay to
         belong to an HIV/AIDS support group
                      ...
Figure 5: Accessibility of support groups                                                                                 ...
Figure 9: Composition of support groups
                 100


                  90
                              95.8
   ...
Table 1: Correlates of need, access, payment and WTP
                                                                 NEED...
Limitations
• Generalisation: data does not allow generalisation beyond the South
  African and even perhaps beyond the Fr...
Concluding comments
• Large, unmet demand for HIV and AIDS support groups among public
  sector ART clients

• ART clients...
Why does mental health matter?
•   Globally, unipolar depressive disorders is projected to be the second
    primary cause...
Summary statistics
98.3%    African

76.0%    Female

65.9%     Single

Hospital Anxiety and Depression Scale (HADS):

31....
Table 1: Correlates of symptoms of anxiety and
          depression in public sector Free State ART clients
              ...
Prevalence of symptoms of anxiety and depression:

Estimated at 31.3% and 24.8%, respectively

Study in Brazil which also ...
Correlates of symptoms of both anxiety and depression:

• Lack of coping and severe side effects associated with an increa...
Limitations
•   Mental health and other scales: how valid and culturally sensitive and
    relevant are these measures to ...
Concluding comments
ART programmes should incorporate the following:


(1) a brief, standardised cultural-specific tool to...
Upcoming SlideShare
Loading in...5
×

06 Booysen Food For Thought

221

Published on

Published in: Health & Medicine
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
221
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
1
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

06 Booysen Food For Thought

  1. 1. Some first food for thought: selected baseline findings from the Effective Aids Treatment and Support in the Free State (FEATS) study Frikkie Booysen, University of the Free State Alok Bhargava, Houston University Damien de Walque, The World Bank Mead Over, Center for Global Development Michele Pappin, University of the Free State 5th SAHARA conference on Socio-Cultural Responses to HIV 02 December 2009, Gallagher Estate, Johannesburg
  2. 2. Acknowledgement • The financial support of The World Bank’s Research Committee, its Development Economics Research Group (DECRG), and the Bank-Netherlands Partnership Programme (BNPP) • Fieldwork managers and fieldworkers of the Centre for Health Systems Research & Development (CHSR&D) • Study participants in the ART programme who willingly sacrificed their time and energy to participate in this research, and frankly shared their views and experiences. • The management and health care staff of the Free State Department of Health and of several local municipalities, who facilitated access to study participants
  3. 3. Background (1) • The Effective Aids treatment and support in the Free State (FEATS) study aims to: – Inform the effective and sustainable scale-up of ART as part of the National Strategic Plan (NSP) – Pilot and evaluate adherence and nutritional support for effective and sustainable ART – Investigate the positive/negative household externalities of ART
  4. 4. Study design (1) • Prospective cohort study • Experimental design is a combination of: – Group time-series, quasi- or field experiment (‘ARV treatment’) and – Zelen-type double randomised consent (‘peer adherence and nutritional support’) design
  5. 5. Study design (2) • ART nurses working at 12 phase-I ART clinics across five Free State districts recruited public sector ART patients into the study during a 12- month period [Oct ’07 to Oct ’08] • Inclusion criteria: – Adult (18+ years) – Initiated ART in past 4 weeks – Resides in community where clinic based • Comparison households: randomly sampled from relevant communities
  6. 6. Study design (3) Group A: Group B: 216 ART households receiving 216 ART households: A + treatment and support provided in bi-weekly visits by trained ARV the existing programme peer adherence supporter (PAS) Group C: Group D: 216 ART households: A + B + 208 comparison households nutritional supplement (canned randomly sampled from the food) delivered by PAS relevant communities
  7. 7. Study design (4) • Ethical clearance: Faculty of Health Sciences, University of the Free State • Data collection (baseline + two rounds of follow- up interviews): – Patient survey and clinical records in patient files – Household survey, including individual interviews with household members (10+ years) – Facility survey, including interviews with ARV coordinators and provider questionnaires
  8. 8. Figure 2: Study design (FEATS) Comparison households Comparison households Current public sector ART only = [A] Public sector clients enrolled in the [A] + Peer adherence support = [B] ART programme [B] + Nutritional support = [C] 1st follow-up 2nd follow-up Recruitment Baseline survey Randomisation survey (F1) survey (F2) April 2008 - January - June October 2007 - October 2008 October 2008 October 2009 2010
  9. 9. Three selected, potentially important questions… – Nutritional status and food security – Need and access @ support groups – Symptoms of anxiety and depression
  10. 10. Background … what challenges do food insecurity and malnutrition, worsened by the present economic and food crises, pose in the context of the HIV and AIDS epidemic? Prevention: Care and treatment: Impact mitigation: • unprotected transactional • PLWA have 10-30% • erodes household sex greater energy livelihoods, thus requirements constraining resilience and • distress migration → risk limiting coping/response of infection for migrants • more frequent and severe options and others OIs and more rapid progression to Aids • impacts on intra- • lower immune status household time allocation • Assist in curbing side • increased risk of vertical and other decisions effects and enhancing transmission of HIV adherence for those on • increased vulnerability of ARV treatment OVC • improved efficacy of ART • lower survival rates following ART initiation Source: Gillespie (2008)
  11. 11. Figure 3: Adult nutritional status, by comparison group 100% 17.2 19.5 17.1 90% 27.6 80% 20.7 70% 22.9 25.7 20.1 60% 50% 40% 49.6 47.5 48.4 41.9 30% 20% 10% 12.5 10.1 10.5 8.7 0% ART patients (n=623) Other adults living with ART Adults in comparison South Africa (n=13,089) patients (n=667) households (n=399) Underweight (BMI < 18.5) Normal (18.5 = BMI < 25) Overweight (25 = BMI < 30) Obese (BMI = 30)
  12. 12. Table 3: Correlates of malnourishment among adults Explanatory variable OR 95% CI Female (yes/no) 0.469 *** 0.305 0.720 EQ-5D 0.634 ** 0.421 0.956 Smoke (yes/no) 1.621 * 0.954 2.755 Ownership of large livestock (number) 1.065 * 0.996 1.138 Ownership of small livestock (number) 0.897 ** 0.814 0.990 Household head's age 1.089 * 0.990 1.199 PLWA in household (yes/no) 1.728 * 0.996 2.998 Ever received food supplements (yes/no) 1.889 ** 1.022 3.491 Sample (n) 1,461 Wald chi2(p-value) 78.4 (<0.001) Pseudo R-square 0.080 Predicted success (%) 89.8
  13. 13. Table 2: Nutritional status across ‘treatment career’ Baseline in Pre-ART Household Post-interview South Africa clinical data initiation interview in clinical data (DHS 2004) Severely malnourished 5.9 7.6 2.9 3.4 Malnourished 10.2 8.5 8.0 5.9 Underweight 16.1 16.1 10.9 9.3 8.7 Normal 49.7 50.2 50.2 50.6 48.4 Overweight 19.1 18.7 22.1 21.2 25.7 Obese 14.8 14.8 16.6 18.7 17.1 Sample (n) 235 235 235 235 13,089.0 Total 100.0 100.0 100.0 100.0 100.0
  14. 14. Figure 4: Household food security, by comparison group 60 51.6 50 40 38.9 31.9 29.1 30 26.0 20 16.4 10 3.2 3.0 0 Food secure Food insecure without Food insecure with hunger Food insecure with severe hunger hunger Comparison households (n=185) Patient households (n=574)
  15. 15. Figure 5: Adult and child food security, by group 100% 90% 19.2 28.1 37.2 80% 49.3 70% 60% 48.5 22.9 50% 52.4 40% 29.9 30% 20% 39.9 32.3 10% 20.8 19.5 0% Comparison households Patient households Comparison households Patient households Adults Children High Low Very low
  16. 16. Table 3: Correlates of household food security Overall Adult Children Explanatory variable OR 95% CI OR 95% CI OR 95% CI Owns land (yes/no) 0.557 0.220 1.410 1.097 0.514 2.340 0.284 * 0.079 Household head's education (comparison = none) Primary 1.166 0.547 2.485 1.327 0.644 2.732 1.590 0.572 Some secondary 1.896 0.860 4.178 2.163 ** 1.016 4.606 3.230 ** 1.107 Matric/grade 12 2.391 * 0.931 6.136 2.514 ** 1.031 6.131 4.546 ** 1.372 Tertiary 5.592 ** 1.388 22.536 4.613 ** 1.181 18.020 11.659 *** 2.387 Household head = African 0.233 * 0.051 1.065 0.276 * 0.067 1.143 0.401 0.093 Patient household (yes/no) 0.647 * 0.398 1.050 0.576 ** 0.364 0.913 0.690 0.369 Sample (n) 674 463 660 Wald chi2(p-value) 74.05 (<0.001) 42.26 (0.017) 67.39 (<0.001) Pseudo R-square 0.087 0.081 0.107 Predicted success (%) 77.5 74.6 78.4
  17. 17. Concluding comments • An integrated, comprehensive response to HIV and AIDS, requires effective, sustainable and scalable nutritional, food security and dietary/lifestyle interventions to enhance both the prevention and treatment efforts in the fight against HIV and AIDS … what research, or more specifically, programme evaluations, are required to help inform the above response?
  18. 18. Background (1) • An integrated, multi-faceted response is required for an effective, sustainable ARV treatment programme • Sustained, long-term adherence crucial • HIV/AIDS support groups a potentially important source of adherence and psychosocial support to ART clients • Yet, little is known regarding the demand, access and willingness-to-pay for belonging to an HIV and AIDS support group
  19. 19. Background (2) • Rhine (2009: 369) points out that support groups in the context of the African HIV epidemic are more than “spaces for discussion of social and health well-being” or “institutions functioning solely to cultivate self-responsible and economically empowered patients” and fulfill other functions in communities • Simoni et al (2007) conducted a RCT to evaluate the use of peer support groups as a tool for enhancing adherence to ART, but found no significant treatment effects… yet, the authors attribute their null finding to the short duration and low intensity of the intervention as well as the heterogenous nature of their study population … a preliminary search of the literature failed to reveal any other evaluations of group-based adherence support interventions for antiretroviral treatment
  20. 20. Figure 3a: Access to support groups, treatment supporters and community health workers Community health Support group Treatment supporter worker Previously only 2.0 2.3 2.0 Currently 7.3 59.4 4.0 Never 90.7 38.3 94.0 Total 100.0 100.0 100.0
  21. 21. Figure 4: Access, need, cost and willingness-to-pay to belong to an HIV/AIDS support group Willing to pay Willingness-to-pay Access (yes/no) Need (yes/no) Pay (yes/no) Cost (ZAR) (yes/no) (ZAR) Previously 84.6% 54.5% Mean: R30.00 [13/655 = 2.0%] [11/13] [6/11] Median: R15.00 IQR: R10.00 - R30.00 Currently 58.3%* Mean: R18.90 10.4%** Mean: R50.00 [48/655 = 7.3%] [28/48] Median: R9.00 [5/48] Median: R30.00 IQR: R8.00 - R20.00 IQR: R20.00 - R40.00 Never 68.0% 61.4% Mean: R24.78 [594/655 = 90.7%] [404/594] [248/404] Median: R20.00 IQR: R10.00 - R28.00 68.4% 55.9%*** Mean: R25.39 Total [415/607] [259/463] Median: R20.00 IQR: R10.00 - R30.00 Notes : * Proportion of current support group members reporting either actual costs for last visit or normally incurring cost for transport, food or membership to participate in meetings; ** Reflects proportion of current support group members who are willing to pay an additional amount to belong to a support group; *** Reflects proportion willing to pay among (a) current support groups members and (b) non-members expressing a demand for membership.
  22. 22. Figure 5: Accessibility of support groups Figure 6: Organisation of support groups 100 100 90 90 80 80 72.9 70 70 64.6 Percentage (%) 60 Percentage (%) 60 50 46.8 50 40 40 30 23.4 30 20 18.8 20 10 12.5 2.1 10 4.2 0 Travel more than short walk Pay for transport Pay for food Pay membership fee 0 from home DoH or clinic staff NGO Self Church Figure 7: Meeting place of support groups Figure 8: Frequency of support group meetings 100 100 90 90 80 80 70 66.6 70 Percentage (%) Percentage (%) 60 60 50 50 40 40 31.3 30 30 27.1 20.8 20.8 20 20 10.7 8.3 10 6.2 10 4.1 4.1 0 0 Clinic Other health Church NGO School Other More than once per week Once per week Two or three times per Once per month facilities month
  23. 23. Figure 9: Composition of support groups 100 90 95.8 Composition: 80 70.2 Comprise persons on ART, HIV- 70 positive persons not yet on ART, 60 Percentage (%) 56.2 50 52 HIV-negative persons, and ones 40 whose status is unknown to the 30 member who responded to the 20 survey… this has NB implications? 10 0 HIV+ and on ART HIV+ but not on ART Known to be HIV-negative HIV status unknown Figure 10: Participation of support groups in prevention activities 100 Activities: 90 80 76.6 Relatively high participation in 70 range of prevention activities… Percentage (%) 60 this has NB implications? 50 46.6 40 30.2 30 20 10 0 Distribute condoms Visit schools and other institution to Visit bars, hotels and other hot spots talk about HIV and AIDS to talk about HIV and AIDS
  24. 24. Table 1: Correlates of need, access, payment and WTP NEED ACCESS PAY WTP Age 0.061 0.032 -0.084 -0.068 Age2 -0.001 * 0.000 0.001 0.001 Education (comparison = none) Some primary -0.135 -0.701 0.463 0.397 Some secondary -0.172 -0.584 0.650 0.553 ** Grade 12 or higher -0.341 -0.953 * 0.707 0.615 ** Household size -0.012 0.040 0.123 *** -0.028 Dependency ratio 0.346 -0.146 -0.546 0.614 * Marital status (comparison = single) Married and cohabiting -0.447 *** 0.435 * -0.002 -0.044 Married but not cohabiting -0.573 *** 0.760 ** 0.117 0.073 Divorced, separated or widowed 0.031 0.139 0.074 0.085 Stigma 0.062 0.033 -0.119 -0.196 *** Disclosure (comparison = disclosed to none) Disclosed to some only -0.256 -0.255 0.209 Disclosed to all 0.827 ** -0.174 -0.198 Avoidant coping 0.027 -0.143 *** 0.070 0.038 Supportive coping 0.335 *** -0.038 0.112 -0.029 Positive coping -0.470 * 0.135 0.303 0.208 Acceptance coping 0.953 * -0.365 Health-related quality of life (EQ-VAS) -0.008 ** -0.003 0.002 0.004 Reported symptions of depression (yes/no) -0.347 0.154 0.615 ** 0.592 ** Reported symptions of anxiety (yes/no) 0.127 -0.427 0.297 -0.083 Access to treatment buddy (yes/no) -0.012 0.576 *** 0.016 0.116 Visited by community health worker (yes/no) -0.270 0.455 -0.767 * 0.204 Previously belonged to a support group (yes/no) 1.024 ** -0.473 0.308 Treatment duration (months) 0.041 0.002 -0.142 ** 0.145 ** Previously on ART (yes/no) 0.021 0.486 1.015 ** 0.172 Sample (n) 510 394 372 240 Wald chi2 or F statistic (p-value) 75.31 (<0.001) 50.04 (0.003) 45.15 (0.028) 2.13 (0.001) Pseudo R 2 or R 2 0.106 0.130 0.092 0.174 Successful prediction (%) 70.6 89.3 63.7 Note: Results are for probit [NEED, ACCESS, PAY] and linear [WTP] regression models respectively. WTP model include non-zero WTP values only. Models were also adjusted for gender, self-reported adherence, income, employment status, and time known HIV+ status.
  25. 25. Limitations • Generalisation: data does not allow generalisation beyond the South African and even perhaps beyond the Free State setting • Selection processes: choices of demand → access and of pay → WTP may be characterised by selection, thus requiring estimation of Heckman selection models • Network-like nature of support: use of SEM may be feasible given multiplicity of (other) support types, which may be substitutes or complements for support groups • Simultaneity and/or endogeneity: cross-sectional data cannot reveal dynamics of demand for and participation in HIV and AIDS support groups among public sector ART clients • Study dynamics: introduction during experimental phase of the study of individual peer adherence support are likely to crowd out demand and WTP for support groups
  26. 26. Concluding comments • Large, unmet demand for HIV and AIDS support groups among public sector ART clients • ART clients attribute relatively large, non-zero benefits to membership of a support group • Nature of membership and activities of existing support groups suggest that such groups may play an important role, not only in treatment support, but also in HIV prevention activities • Importance of coping, disclosure, stigma and depression as correlates of demand, access and WTP hint at the important role of support groups as means of psycho-social support to ART clients • Importance of past support group membership and access to treatment supporter implies that complementarity and substitutability of different means of (adherence) support should be investigated
  27. 27. Why does mental health matter? • Globally, unipolar depressive disorders is projected to be the second primary cause of burden of disease by 2030 next to HIV and AIDS (Mathers and Loncar, 2006) • In South Africa, it is estimated that in the year 2000, HIV and AIDS represented the most important cause of loss of disability adjusted life year, and unipolar depressive disorders the tenth leading cause (Bradshaw et al, 2003) • National Income Dynamic Study [CES-D 10]: prevalence of depression among women and men 36% and 27% (Ardington and Case, 2009) • HIV and AIDS and mental illness are interconnected. Depression is associated with risky sexual behaviour (WHO, 2001), which may result in contracting HIV. HIV and AIDS may lead to symptoms of anxiety and depression when people learn their HIV+ status … as a result, mental health matters for long-term treatment adherence and sustainable, effective treatment programmes
  28. 28. Summary statistics 98.3% African 76.0% Female 65.9% Single Hospital Anxiety and Depression Scale (HADS): 31.3% Symptoms of anxiety 24.8% Symptoms of depression
  29. 29. Table 1: Correlates of symptoms of anxiety and depression in public sector Free State ART clients Anxiety Depression OR 95% Conf. Interval OR 95% Conf. Interval Coping scale 0.029 *** 0.008 0.115 0.060 *** 0.015 0.236 Stigma scale 1.553 *** 1.152 2.093 1.285 0.914 1.806 Condom use and partner knows HIVstatus (comparsion group = no sex) Always use condom and partner knows HIV status 1.332 0.714 2.486 1.154 0.613 2.173 Inconsistent condom use and partner knows HIV status 0.812 0.280 2.351 0.359 * 0.113 1.141 Always condom use and partner don't know status 0.446 0.048 4.100 0.172 0.010 2.977 Inconsistent condom use and partner don't know HIV status 0.466 0.086 2.531 0.284 0.053 1.516 Socio-economic status (comparison group = grant holder) Other grant holder in household 1.422 0.482 4.198 0.491 0.154 1.563 Employment 0.748 0.341 1.644 1.289 0.585 2.838 Support within the household 0.486 * 0.231 1.027 0.812 0.388 1.700 Support outside the household 0.862 0.283 2.626 1.848 0.631 5.407 Other support 1.042 0.363 2.989 1.863 0.681 5.091 Time since first HIV+ test 1.008 0.996 1.020 1.012 ** 1.001 1.022 Self reported side effect (comparison group = none) Somewhat disruptive side effects 1.740 0.872 3.474 1.193 0.578 2.464 Very disruptive side effects 3.905 *** 1.893 8.057 2.136 ** 1.016 4.493 Hospitalisation in past 6 months (yes/no) 2.140 * 0.906 5.054 EQ-5D 0.564 * 0.309 1.027 Sample size 394 408 Wald-statistic 81.5 64.3 P-value <0.001 <0.02 R2 0.2 0.2 % successfully predicted 78.6 79.2
  30. 30. Prevalence of symptoms of anxiety and depression: Estimated at 31.3% and 24.8%, respectively Study in Brazil which also used the HADS to assess symptoms of anxiety and depression among patients who initiated ART reported similar prevalence rates namely, namely 35.8% and 21.8% (Nogueira Campos et al, 2006) However, other studies from South Africa report much lower rates of anxiety (Olley et al, 2003) and much higher rates of depression (Olley et al, 2003; Moosa et al, 2005; Simbayi et al, 2007) … why the latter differences?
  31. 31. Correlates of symptoms of both anxiety and depression: • Lack of coping and severe side effects associated with an increased likelihood of symptoms of both anxiety and depression • Symptoms of anxiety: Symptoms of depression: reliance on a social welfare grant earlier HIV diagnosis HIV/AIDS-related stigma inconsistence condom use poorer health status recent hospitalisation … when we use a biprobit model to jointly estimate the coefficients on the correlates and symptoms of anxiety and depression, given the high correlation between the two, the results remain similar…
  32. 32. Limitations • Mental health and other scales: how valid and culturally sensitive and relevant are these measures to our particular study population? • Selection bias: Study participants represented those HIV-positive individual who gained access to the public sector antiretroviral treatment programme and successfully completed drug readiness training. Patients suffering from anxiety and depression may have been be less likely to seek care, to adhere and complete drug readiness, and to initiate treatment, hence possibly being under represented in the study. • Cross-sectional data: follow-up interviews currently in field will help explore role of various causal mechanisms and/or processes, including symptoms of anxiety, depression and other mental illness as both cause and effect of other important treatment dynamics and experiences
  33. 33. Concluding comments ART programmes should incorporate the following: (1) a brief, standardised cultural-specific tool to diagnose anxiety and depression; (2) screening for anxiety and depression; (3) appropriate treatment for those diagnosed with anxiety and depression; (4) additional, equitable distributed financial and human resources for mental health; (5) use of state-of-the-art and best-practice in order to ensure treatment success; (6) equipping ART patients with positive coping skills during drug readiness training; (7) tailor-made and tested anti-stigma programmes.
  1. A particular slide catching your eye?

    Clipping is a handy way to collect important slides you want to go back to later.

×