Quality of Life Change in Zimbabwean Patients at

Quality of Life Change in Zimbabwean Patients at Western versus Traditional African Medical Care Sites Tonya N. Taylor, Curtis Dolezal, Susan Tross, and William C. Holmes HIV Center for Clinical and Behavioral Studies June 25, 2009

Background l 22 million people living with HIV/AIDS in sub-Saharan Africa have little to no access to life saving antiretroviral treatment (ART). – only 28% of those needing ART in the region are currently covered. l Structural and socioeconomic constraints prevent many from accessing HIV prevention, testing, and treatment. l Consequently many delay in their engagement into treatment

Background cont. l To remedy these challenges, many have called for community-based approaches, including the use of traditional healers l Traditional Healing is an integral part of African culture and society and is the primary system of health care for more than 80% of rural Africa. l Many Africans with HIV turn to traditional healing throughout their disease course. l There are no data on the potential role of Traditional Healing on the delivery of HIV care (or ART rollout) in Africa or other resource poor settings

Objectives l To explore differences in HIV-related quality of life (Qo. L) with varying medical care sites. l We examined self-perceived Qo. L changes over 1 month in a presumptively HIV+ patients from traditional African care (TAC) and western medical care (WC) sites using the HAT-Qo. L and the MOS-HIV l We hypothesized that in individuals seeking TAC would significantly improve in Qo. L compared to those seeking WC l We conjectured that the mental health benefits of culturally contextualized psychological support of traditional healers would result in a broad perception of improved physical functioning and well-being l If confirmed, we anticipated that these findings would be among the first to provide data on the potential cultural dynamics that are missing from current policies on ART rollout for Africa.

Traditional Healing and HIV Care in rural Zimbabwe l Not a substitute for western therapeutics, but an alternative explanatory model for the diagnosis and management of illness l Diagnose its socio-etiology, answering the more culturally relevant question: why did this happen to me? l Biomedicine can address HIV's physical symptoms, but its social and cultural cause and subsequent meanings can only be revealed through the metaphysical insights of traditional healing

Un’anga l Health and healing are maintained by a balance of coolness from social harmony that counteracts the heat of social conflict caused by witchcraft, greed and envy l Six types of healers (n’anga) – – – spirit medium (svikiro) diviner (mushoperi) herbalist (godobori) faith healers (muporofita) midwife (nyamukuta) injectionist l N’angas help individuals examine the interpersonal context of their illness experience l e. g. resolve disputes or help individuals cope with grief and loss

Quality of Life & HIV/AIDS l Qo. L, hereafter is defined as the degree of well-being felt by an individual or group of people that includes social, economic and political factors. l Qo. L (especially in Public Health) is often measured as physical, psychological and social well-being. l Since the advent of the AIDS pandemic, attention to Qo. L with HIV infection has remained a strong area of study l The ability to assess HIV-specific QOL in a culturally sensitive manner is imperative. However, there are few QOL instruments validated for use in sub-Saharan Africa.

Quality of Life Measures Medical Outcomes Study HIV Health HIV/AIDS-Targeted Quality of Life Survey (MOS-HIV) Wu et al 1991 Instrument (HAT-Qo. L). Holmes & Shea, 1997 l 11 dimensions: – general health perceptions l 9 dimensions: – physical function – overall function – role function – disclosure worries – social function – health worries – cognitive function – financial worries – Pain – HIV mastery – mental health – life satisfaction – energy/fatigue – medication worries – health distress – provider trust – quality of life – sexual function – health transition

Adaptation of the Instruments l The International Quality of Life Assessment Project Translation Procedures were used to adapt the QOL instruments. l These processes were carried out by six Linguistic graduate students from the University of Zimbabwe, two bilingual health care workers, and the original instrument developers: Drs. William Holmes and Christopher Mast (for Dr. Albert Wu). l The reliability and validity of the Shona adapted QOL instruments are reported elsewhere (Taylor, Dolezal, Tross and Holmes, in AIDS Care).

Methods

Chipinge - A Place of Troubles l Border community l Commercial farming – Migrant labor – Long distance trucking l High illiteracy l Extreme rural poverty l Gender inequality l Military camp l Commercial sex workers

Participants l 400 Shona patients (200 TAC and 200 WC) were consecutively sampled as they sought treatment at 12 sites for one of the following infections associated with HIV: – TB, recurring STIs, unusual skin problems (especially herpes zoster) and chronic diarrhea l 24 healers recruited to their self-reported expertise in the treatment of HIV or AIDS l Inclusion criteria include: being an adult with one of the aforementioned an OI associated with HIV, and presenting for care and treatment. l All respondents read and signed a written consent form.

Procedures l WC Participants were recruited by their primary care provider (doctor or nurse) and invited to participate in the study. l TAC Participants were first invited to participate in the study by the healer. After observing each consultation the PI and research assistant invited only those patients who fit the criteria for the in-depth interview. l Interviews were conducted at baseline (after the clinical consultation) and one-month follow-up in a private room at the clinic or healer’s hut

Data Collected 254 individuals completed baseline and one-month follow-up l Data collected during the interviews included: – A socio-demographic questionnaire (5 min) – MOS-HIV (10 min) – In-depth Interview to elicit an Illness Narrative (20 min)* – WHO clinical guidelines for AIDS (5 min) – HAT-Qo. L (10 min) l *data will not be presented today

Analytic Plan l For continuous and categorical variables, bivariate two-group comparisons were performed using 2 -tailed t-tests and X 2 methods l A linear regression model was constructed for each QOL dimension with change-in-dimension-score as the dependent variable l Site-of-care (SOC) was entered into all models, along with covariates that exhibited a bivariate association (p≤ 0. 10) with SOC to adjust for potentially confounding l Baseline Qo. L scores were entered into each model to adjust for differences in perceived health status and well-being at study entry l Final alpha-level of 0. 01 for SOC differences was employed.

Results

Demographic Profile TAC WC N=155 N=99 13 -59 31. 9 (11. 2) 32. 2 (8. 4) NS Gender Female 60% 56% NS Marital status Married 58% 52% NS Employment No 14% 24% 0. 04 Education Partner Loss Child Loss O-Level > 25% 23% 40% 39% 34% 38 0. 03 0. 05 Mean Age (SD) Sig. NS • p-values from chi-square tests, except age, which is from t-test; • All percentages computed using denominator made up of all those who provided an answer; rounding may make addition of percentages for some characteristics greater or less than 100%; d O- and A-Levels in Zimbabwe correspond roughly to grades 8 -10 and 11 -12, respectively, in the United States.

Clinical Profile TAC n=155 WC n=99 (p-value) Weight loss 79% 88% NS Chronic diarrhea 67% 47% 0. 001 Prolonged fever 74% 82% NS Persistent cough 54% 80% <0. 001 Generalized rash 51% 41% NS Shingles 12% 19% NS Candida 27% 38% 0. 05 Cold sores 45% 44% NS Lymph nodes 68% 52% 0. 01 Meet AIDS definition TB (ever) 74% 9% 79% 66% NS <0. 001 STIs (ever) 62% 55% NS

Baseline Qo. L by Site-of-Care l The TAC subgroup had higher (thus, better) baseline scores in 16 out of 19 dimensions l On the HAT-Qo. L, the TAC group significantly exceeded the WC group on 4 out of 8 dimensions – overall function, disclosure worries, life satisfaction, and provider trust l On the MOS-HIV, the TAC subgroup significantly exceeded the WC group on 7 out of 11 dimensions – general health perceptions, physical function, role function, social function, pain, mental health, and health distress

Baseline HAT-Qo. L Scores by SOC Qo. L Dimension, mean (SD) Overall Function Disclosure Worries Health Worries Financial Worries Illness Mastery Life Satisfaction Medication Worries Provider Trust TACa WCa p-value 48. 49 (25. 91) 47. 32 (27. 98) 46. 09 (28. 85) 29. 09 (24. 77) 45. 56 (32. 16) 44. 40 (27. 52) 74. 26 (25. 97) 79. 25 (23. 67) 40. 24 (24. 35) 54. 69 (29. 16) 41. 86 (27. 10) 28. 45 (25. 06) 46. 09 (31. 93) 35. 35 (24. 99) 72. 71 (26. 38) 67. 09 (28. 08) . 012. 046 NS NS NS . 009 NS . 001

Baseline MOS-HIV Scores by SOC Qo. L Dimension, mean (SD) TAC WC p-value General Health Perceptions Physical Function Role Function Social Function Cognitive Function Pain Mental Health Energy/Fatigue Health Distress Overall Quality of Life Health Transition 30. 61 (17. 44) 62. 94 (26. 15) 50. 97 (39. 26) 59. 87 (33. 46) 49. 06 (23. 60) 32. 12 (23. 86) 54. 32 (23. 23) 47. 78 (25. 11) 48. 81 (27. 55) 33. 71 (25. 58) 61. 45 (30. 58) 26. 16 (13. 49) 54. 48 (22. 64) 40. 40 (38. 25) 48. 28 (36. 81) 47. 12 (23. 49) 23. 30 (23. 47) 47. 35 (21. 86) 41. 87 (23. 89) 38. 08 (25. 37) 32. 58 (24. 86) 62. 37 (28. 21) . 032. 007. 036. 012 NS . 004. 018 NS . 002 NS NS

Linear Regression Comparing Change in Qo. L by Site l At an alpha level of 0. 05, the TAC subgroup was found to have reported significantly greater adjusted Qo. L improvements over one month in 14 of 19 dimensions l 6 of 8 HAT-Qo. L dimensions included overall function, health worries, illness mastery, medication worries, financial worries, and provider trust l 8 of 11 MOS-HIV include general health perceptions, physical function, role function, pain, energy/fatigue, health distress, mental health, and overall Qo. L.

Adjusted HAT-Qo. L Change Over 1 Month Qo. L Dimension HAT-Qo. L, mean (SD) Overall Function Disclosure Worries Health Worries Financial Worries Illness Mastery Life Satisfaction Medication Worries Provider Trust TAC WC +11. 1 (26. 5) +1. 2 (25. 7) +13. 3 (32. 2) +3. 2 (22. 5) +5. 5 (30. 2) +10. 8 (32. 6) +7. 9 (29. 7) +0. 7 (24. 1) +1. 3 (21. 9) -2. 2 (24. 9) +0. 5 (28. 1) -4. 2 (26. 2) -3. 2 (26. 7) +6. 9 (28. 3) -0. 9 (27. 0) -1. 3 (25. 8) b(se)a p-value 11. 7 (3. 6) 0. 001 5. 3 (3. 6) NS 16. 2 (4. 2) <0. 001 7. 0 (3. 5) 0. 046 11. 4 (4. 2) 0. 008 NS 6. 8 (4. 3) 10. 9 (3. 7) 0. 004 11. 0 (3. 6) 0. 002 Unstandardized regression results for site-of-care (WC=0, TAC=1) after adjusting for the following: baseline quality of life score, educational attainment, employment status, loss of a spouse, cough, TB, chronic diarrhea, weight loss, swollen lymph nodes, and candidiasis. b Effect size for changes in HAT-Qo. L and MOS-HIV scores were arbitrarily interpreted in terms of small (-5. 0), moderate (6. 016. 0), or large ( 17. 0) changes in quality of life. a

Adjusted MOS-HIV Change Over 1 Month QOL Dimension TAC WC b(se)b p-value MOS-HIV , mean (SD) General Health Perceptions Physical Function Role Function Social Function Cognitive Function Pain Mental Health Energy/Fatigue Health Distress Overall Quality of Life Health Transition +11. 1 (24. 0) +10. 3 (28. 8) +11. 3 (44. 7) +2. 6 (37. 4) +2. 8 (25. 3) +17. 9 (28. 7) +2. 8 (26. 0) +8. 9 (26. 1) +7. 6 (30. 8) +15. 8 (32. 5) +14. 8 (31. 7) +0. 7 (15. 5) +0. 5 (22. 4) -0. 5 (42. 0) -2. 4 (41. 6) -0. 2 (23. 2) +8. 3 (25. 8) -1. 1 (21. 8) +0. 5 (25. 1) +2. 5 (26. 8) +2. 0 (32. 3) +6. 3 (32. 8) 10. 0 (3. 3) 0. 003 16. 0 (3. 7) <0. 001 18. 3 (6. 3) 0. 004 NS 9. 7 (5. 2) NS 4. 4 (3. 5) 12. 7 (3. 6) 0. 001 7. 9 (3. 3) 0. 018 10. 2 (3. 6) 0. 004 12. 8 (4. 2) 0. 003 10. 1 (4. 2) 0. 017 4. 6 (4. 0) NS

Discussion l Patients from the TAC versus the WC sites demonstrated significantly greater (moderate to large in size) improvement across the majority of QOL dimensions assessed over one month l Since baseline QOL was higher for the TAC site, this group had less “room” on the scale to improve and, thus, their significantly larger increase is more surprising l The differences in improvement were independent of baseline QOL differences between the sites

Discussion l At baseline there were some clinical parameters on which the sites differed that might have suggested differences in QOL change were in part due to greater illness l However, these differences sometimes indicated greater illness in the TAC group (i. e. , swollen lymph nodes, chronic diarrhea) and sometimes indicated greater illness in the WC group (i. e. , cough, thrush, TB, and past hospitalization) l We adjusted for significant symptomatic differences in these analyses – Qo. L site differences were independent of these variables l Similarly, at baseline there were some sociodemographic parameters on which the sites differed. – In adjusted analyses, site QOL change differences were independent of these differences

Discussion l Our findings may reflect the fact that TAC sites, particularly, provide psychosocial support and a familiar cultural context for health care l As the primary system of health care for more than 80% of Africa, TAC provides patients culturally-valued understandings of HIV/AIDS l As such, TAC sites may prove to be the most accessible and cost-effective way to bolster access to health care for resourcepoor areas and improve the QOL for those living and dying with HIV/AIDS.

Role of Traditional Healers l Thus far, traditional healers remain underutilized partners in the delivery of HIV care l In the era of ART rollout, they could be an invaluable, community-based resource for linkage to HIV testing, education and prevention, and treatment for OIs such as TB l As community healthcare providers who are considered local authorities on issues of health and healing, they could be invaluable in implementing directly delivered therapy to improve ART adherence in rural areas where there are not enough WC professionals to monitor patients adherence.

Collaboration l Many traditional healers in Zimbabwe readily express their limitations in treating HIV/AIDS and are eager to collaborate with WC providers l The Zimbabwe National Traditional Healers Association (ZINATHA) has actively worked with the Zimbabwean Medical Research Council to establish protocols for testing traditional remedies thought to be efficacious in treating OIs associated with HIV l The time appears right, then, for a productive engagement and study of collaborative care with traditional African healers.

Limitations l Participants were not randomized to the two subgroups l Participants were a consecutively recruited, rural sample, so the representativeness of our findings to urban populations, remains unclear l Information about those who we missed in the study is not known, limiting our ability to clarify potential biases l The sample was a highly symptomatic one (76% met WHO definition for AIDS and 89% of the WC group had TB) l The sample also has an over representation of women l Finally, QOL changes were only assessed for one month of follow-up.

Conclusion l Our finding suggest that confirmatory studies are needed in larger, more diverse African populations l If confirmed, these findings will add more weight to recommendations to develop, pilot test, and evaluate culturally -tailored interventions to better engage African PLWHA in HIV care using traditional healers

Acknowledgments The National Institute of Mental Health (1 R 03 MH 62250 -01) l The Social Science Research Council l l – l Zimbabwe’s National Association of Traditional Healers Zimbabwe’s Medical Research Council The People of Chipinge Drs. Susan Tross, Curtis Dolezal & William C. Holmes HIV Center for Clinical and Behavioral Studies Dr. Taylor is a postdoctoral fellow supported by a training grant from the National Institute of Mental Health (T 32 MH 19139, Behavioral Sciences Research in HIV Infection; Principal Investigator: Anke A. Ehrhardt, Ph. D. ) at the HIV Center for Clinical and Behavioral Studies at NY State Psychiatric Institute and Columbia University (P 30 -MH 43520; Principal Investigator: Anke A. Ehrhardt, Ph. D. )

My Team (Back row) Fungai, Simon, Faith, Tonya, Tsaka (the cat) & Carrinah, (Front row) Bernard, Shumba (the dog) and Ines. (Missing) Mzilikazi (our other useless cat)

Ma. Shuma Chaivzo!

Medical Pluralism is the Norm Chi-Square for Treatment Site vs. Treatment Preference Site WC only TAC only WC 35. 5% TAC . 0% 91. 0% Total 17. 8% 4. 8% 77. 5% Chi. Square 95. 617 a df=2 . 5% Combined 64. 0% Sig=. 000