Transcript Microeconomic impact of HIV disease among female bar/hotel

Microeconomic impact of
HIV disease among female
bar/hotel workers in northern
Tanzania:
methodological considerations
Tony Ao
Advisor: Dr. Saidi Kapiga
Harvard School of Public Health
Population Impacts on Economic Development Research Conference
03 NOV 2006
Background

HIV disproportionately affects women

59% of infections are women in SSA (UNAIDS 2005)
 Male: 6.4%
 Female: 7.7%
At-risk populations in Tanzania

Women working in bars/hotels have highest
risk:



Arusha: 75% (Nkya 1991)
Moshi: 26% (Kapiga 2002)
Mbeya: 68% (Reidner 2006)
Macroeconomics & HIV
No clear link between HIV and economic growth
Negative effect:






Kambou et al (1992)
Cuddington (1993)
Cuddington and Hancock (1994)
Bonnel (2000)
Papageorgiou and Stoytcheva (2004)
Corrigan, Gloom, Mendez (2005)
No effect:


Bloom and Mahal (1997)
Werker, Ahuja, Wendell (2006)
Microeconomics & HIV
Examples:




Household verbal autopsies (Ngalula et al 2002)
Kenyan tea plantation workers (Fox et al 2004)
Household surveys in Kenya and Rwanda (UNAIDS 2004)
Elderly health and AIDS death (Dayton & Ainsworth 2004)
Microeconomic impact of HIV



Mostly assessed within formal sector or households
No study with female bar/hotel workers
Important for intervention and policy implications
Proposed Framework
Clinical
Factors
Clinical signs &
symptoms
Behavioral
factors
Health seeking
behavior
Environmental
factors
HIV Infection
Microeconomic
impact
Objective and hypotheses
Objective:

To investigate the microeconomic impact of HIV
disease among female bar/hotel workers
Hypotheses:
Compared to HIV negative women, HIV positive
women are expected to:




Report lower monthly income
Report higher health care expenditure
Report higher health seeking behavior
Report lower level of savings
Possible Approaches



Randomized controlled trial
Longitudinal study
Cross sectional
Instrumental variable (IV)
 Propensity score matching (PSM)

Method



Study design: cross sectional with retrospective
questionnaire (adapted LSMS)
Study population: bar/hotel workers presenting
for screening for existing CHAVI study at
clinic
Outcomes:




Monthly income
Health care utilization in past 3 months
Health care spending in past 3 months
Household savings
Propensity Score Matching

Propensity score matching


Uses predicted probability of HIV status based on
observed predictors from logistic regression to create
counterfactual group for comparison
Advantages:



Improves causal inference
Ethically appropriate
Logistically feasible
Analysis
Propensity score matching

Step 1: Run Multivariate Logistic Regression




Step 2: Match each HIV+ to one HIV- woman based on PS 
New sample of “randomized” individuals






Dependent variable: Y=1 if HIV+; Y = 0, otherwise
Include all observed characteristics except outcomes
Obtain PS: predicted probability (p) or log[p/(1-p)] for each woman
Nearest neighbor matching
Caliper matching
Mahalanobis metric matching in conjunction with PSM
Stratification matching
Difference-in-differences matching (kernel & local linear weights)
Step 3: Run multivariate analyses using newly matched sample
Data collection
Issues to consider:
 Reliability of self-report of income and
sexual behavior
 Recall bias
 Income not a sufficient variable
Data collection
ACASI
(audio computer-assisted self-interviewing)
Source: Waruru et al. 2005
Data collection
Advantages of ACASI






Using tablets vs. conventional laptops
Local written and spoken language
Accurate reporting of sensitive data
Accurate data entry
Validated in Zimbabwe1 and Kenya2
Builds local research capacity
1van
de Wijgert, J., N. Padian, et al. 2000
2Waruru
et al. 2005
Ethical considerations

Screening study has been approved, no
additional specimen collection needed

Sensitive information will be obtained

Confidentiality and data management is
paramount
Limitations

PSM does not match on unobserved contextual
characteristics  matching might not be 100%
perfect

Retrospective data may not capture outcome
accurately

Generalizability

Acceptability of ACASI
Thank you
William & Flora Hewlett Foundation
Population Reference Bureau
David Canning
Ajay Mahal
Grace Wyshak
Saidi Kapiga
References
Bloom, David and Ajay Mahal. Does the AIDS Epidemic threaten Economic Growth? Journal of Econometrics.
1997. 77:105-124.
Bonnel, Rene. HIV/AIDS: Does it Increase or Decrease Growth in Africa? World Bank, mimeo (2000).
Corrigan, Paul & Glomm, Gerhard & Mendez, Fabio, 2005. "AIDS crisis and growth," Journal of Development
Economics. 77(1), pages 107-124, June
Cuddington, John T. and John D. Hancock (1994) ‘Assessing the Impact of AIDS on the Growth Path of the
Malawian Economy’, Journal of Development Economics 43: 363–68.
Dayton J and Martha Ainsworth. The elderly and AIDS: coping with the impact of adult death in Tanzania. Soc Sci
Med. 2004 Nov; 59(10):2161-72.
Fox, M. P., S. Rosen, et al. (2004). "The impact of HIV/AIDS on labour productivity in Kenya." Trop Med Int Health
9(3): 318-24.
KAMBOU, G., S. Devarajan and Mead Over (1992) ‘The Economic Impact of AIDS in an African Country:
Simulations with a General Equilibrium Model of Cameroon’, Journal of African Economies 1(1): 109–30.
Ngalula, J., M. Urassa, et al. (2002). "Health service use and household expenditure during terminal illness due to
AIDS in rural Tanzania." Trop Med Int Health 7(10): 873-7.
Nkya WM, Gillespie SH, Howlett W, et al. Sexually transmitted diseases in prostitutes in Moshi and Arusha,
Northern Tanzania. Int J STD AIDS 1991;2:432–5.
Riedner, G., M. Rusizoka, et al. (2003). "Baseline survey of sexually transmitted infections in a cohort of female bar
workers in Mbeya Region, Tanzania." Sex Transm Infect 79(5): 382-7
Tanzania Commission for AIDS (TACAIDS), National Bureau of Statistics (NBS), and ORC Macro. 2005. Tanzania
HIV/AIDS Indicator Survey 2003-04. Calverton, Maryland, USA: TACAIDS, NBS, and ORC Macro.
Over, Mead. The Macroeconomic Impact of AIDS in Sub-Saharan Africa. World Bank Working Paper 1992.
Papageorgiou, Chris and Petia Stoytcheva. What Do We Know About the Impact of AIDS on Cross-Country Income
So Far? LSU, mimeo (2004).
UNAIDS (2004). 2004 Report on the Global HIV/AIDS Epidemic: 4th Global Report. Geneva, Switzerland,
WHO/UNAIDS.
van de Wijgert, J., N. Padian, et al. (2000). "Is audio computer-assisted self-interviewing a feasible method of
surveying in Zimbabwe?" Int J Epidemiol 29(5): 885-90.
Waruru AK, NduatiR, Tylleskar T. Audio computer assisted self interviewing (ACASI) may avert socially desirable
responses about infant feeding in the context of HIV. BMC Med Inform Decis Mak. 2005 Aug 2; 5:24.
HIV in Tanzania
Men: 6.3%
Women: 7.7% (DHS 2005)
Age and sex-specific HIV prevalence, 2003
Source: Tanzania Commission for AIDS (TACAIDS), National Bureau of Statistics (NBS), and ORC Macro. 2005.
Tanzania HIV/AIDS Indicator Survey 2003-04. Calverton, Maryland, USA: TACAIDS, NBS, and ORC Macro.