Title Socio-Economic Impact of Reducing Premature Adult Mortality: The Case
of Antiretroviral Treatment for HIV/AIDS Patients
Abstract
The objective of this research is to measure the socio-economic impact of reductions in
adult mortality due to HIV/AIDS at the household level. To do this, this study will examine
the socio-economic benefits of anti-retroviral therapy. Over and above prolonging lives
which is the first-order benefit of treatment, the distribution of anti-retroviral therapy is
expected to produce a socioeconomic benefit for the entire household: both the patient and
the family members who were giving him care should be able to return to the labor market,
children might be able to return to school, the welfare of the family should increase. We will
also carefully analyze whether those impacts vary across genders. Our evaluation will
measure these effects, using a set of longitudinal household and health facility surveys being
conducted in several countries. It will also attempt to determine under which conditions
these socio-economic effects can be maximized, by comparing different delivery modes and
assessing which ones are more likely to not only promote adherence and postpone the
development of resistance, but also to maximize the socio-economic benefits of ART.
Contact information
Damien de Walque, ddewalque@worldbank.org
Lead institution
The World Bank.
Countries where the research will take place

Rwanda, Burkina Faso, Ghana, Mozambique, South Africa
How does the research describe the impact of population/reproductive
health on poverty reduction and/or economic growth?
HIV/AIDS is the leading cause of adult death on the African continent. Early sexual
initiation, early marriage, risky sexual practices, and commercial sex work have all
contributed to the transmission of the pandemic—with dire consequences for the wellbeing
not only of the person who has AIDS, but also for others in their household. This makes it
important to understand the socio-economic consequences of efforts to reduce premature
adult mortality through the use of anti-retroviral therapy.
How will the research address a policy need, and what kind of policy
lesson is expected?
1) We are measuring the impact of treatment on the welfare of patients and family
members, allowing to document and quantify the socio-economic benefits of ART (using
Datasets 1 and 2)
2) We are investigating what are the effects of ART on HIV transmission and prevention?
(using Datasets 1 and 2)
3) We are investigating what are the determinants of treatment success including
adherence? (using Datasets 1 and 3)
4) We are investigating how to encourage cost-effectiveness and capacity building to
reinforce the sustainability of ART delivery? (using Dataset 3)
Methods used
Analysis of longitudinal data sets
Data used
Datasets collected
1. Longitudinal household survey of HIV patients
 Rwanda, Burkina Faso, Ghana, Mozambique, South Africa
2. Household survey in general population : Ghana, Burkina Faso
3. Longitudinal health facility survey
 Burkina Faso, Rwanda, Mozambique, South Africa.
Research products and results
1) “HIV services delivery and overall quality of care and satisfaction in Burkina Faso:
are there privileged patients?” by Harounan Kazianga Seni Kouanda , Laetitia N.
Ouedraogo, Elisa Rothenbuhler , Mead Over & Damien de Walque.
Revise and resubmit at Health Policy and Planning. Resubmitted November 2010.
Abstract
Background: In Burkina Faso, the number of HIV-positive persons under
antiretroviral treatment (ART) quintupled from 2004 to 2007. The effect of
the provision of ART and other medical services related to HIV on the
quality of care for HIV and non-HIV patients is debated, but the evidence is
limited.
Methods: 43 health facilities delivering ART were surveyed in 2006. In each
facility, the manager and health providers were surveyed, one provider at
least selected from the HIV/AIDS department; 10 randomly selected
outpatients including 5 from the HIV/AIDS services were interviewed.
Healthcare quality was assessed by a quality index based on exit interviews.
Upfront costs and waiting time were evaluated from patients’ report. We
performed health facility fixed effects multivariate regressions of healthcare
quality, upfront costs and waiting time. We focused on patients’ wealth and
whether the purpose of the visit was related to HIV/AIDS.
Results: Consulting for HIV-related services, while not more costly to
patients significantly increases the quality of care received but also
increases substantially the time spent waiting for this upgraded service. The
wealth of patients does not affect care quality, but helps in reducing waiting
time, in particular for HIV patients.
Conclusion: Our finding that patients visiting for HIV services receive on
average better quality than others does not imply that HIV services have had
a negative impact on other health services, but further investigation using
panel data should be conducted to investigate this question. However, it
would be desirable for the quality of care in other services to reach at least
the level attained in HIV services and the upfront costs to be reduced.
2) “Antiretroviral Therapy Awareness and Risky Sexual Behaviors: Evidence from
Mozambique” by Damien de Walque, Harounan Kazianga and Mead Over.
World Bank Policy Research Working Paper # 5486, November 2010.
Under review
Presented at Symposium and Festschrift in Paul Schultz’s honor, Washington DC, May 2010.
Abstract:
This paper studies the effect of increased access to antiretroviral therapy on risky sexual
behavior, using data collected in Mozambique in 2007 and 2008. The survey sampled both
households of randomly selected HIV positive individuals and households from the general
population. Controlling for unobserved individual characteristics, the findings support the
hypothesis of disinhibition behaviors, whereby risky sexual behaviors increase in response to
the perceived changes in risk associated with increased access to antiretroviral therapy.
Furthermore, men and women respond differently to the perceived changes in risk. In
particular, risky behaviors increase for men who believe, wrongly, that AIDS can be cured,
while risky behaviors increase for women who believe, correctly, that antiretroviral therapy
can treat AIDS but cannot cure it. The findings suggest that scaling up access to
antiretroviral therapy without prevention programs may not be optimal if the objective is to
contain the disease, since people would adjust their sexual behavior in response to the
perceived changes in risk. Therefore, prevention programs need to include educational
messages about antiretroviral therapy, and address the changing beliefs about HIV in the era
of increasing antiretroviral therapy availability.
3) “Food crisis, household welfare and HIV/AIDS treatment: Evidence from Mozambique” by
Damien de Walque, Harounan Kazianga, Mead Over and Julia Vaillant.
Preliminary versions of this paper have been presented in a GAP seminar at the World
Bank, at the AIDS Impact Conference in Gaborone, at the 4th Annual AIID workshop on “The
Economic Consequences of HIV/AIDS” in Amsterdam, at the IAEN Pre-Conference
Workshop and the XVIII International AIDS Conference in Vienna, and at the DIIS
Conference on “Impacts, Responses & Initial Lessons of the Financial Crisis for Low Income
Countries” in Copenhagen.
Abstract
Using panel data from Mozambique collected in 2007 and 2008, we explore the impact of
the food crisis on the welfare of households living with HIV/AIDS. We find that there has
been a real deterioration of welfare in terms of income, food consumption and nutritional
status in Mozambique between 2007 and 2008, among both HIV and comparison
households. However, HIV households have not suffered more from the crisis than others.
Results on the evolution of the labor force participation suggests that initiation of treatment
and better services in the health facilities have counter-balanced the effect of the crisis by
improving the health of patients and their labor force participation. In addition, we look at
the effect of the change in welfare on the frequency of visits to the health facility of patients
and on their treatment outcomes, measured by the change in the CD4 count. Both variables
can proxy for adherence to treatment. This is a particularly crucial issue as it affects both
the health of the patient and public health, because sub-optimal adherence leads to the
development of resistant forms of the virus We find no effect of the change in welfare on the
frequency of visits, but we do find that people who experienced a negative income shock
also experienced a reduction or a slower progression in their CD4 counts.
4) “How do the presence and extent of ARV services affect other health services?
Evidence from Mozambique” by Rita Costa, Damien de Walque, Harounan Kazianga
and Mead Over.
Preliminary versions of this paper have been presented as poster at the IAEN PreConference Workshop and the XVIII International AIDS Conference in Vienna, July 2010.
Abstract: Background: Large amounts of money have been spent to fight the
HIV/AIDS epidemic in Africa. Recently a strong effort has been done to scale up
Antiretroviral Therapy (ART) services. This has caused large debate on whether or not
local health systems are prepared for this extra effort and what will be its impact on other
health services. Methods: This study analyzes quantitatively this issue using data from
three waves of a health facility survey conducted in five provinces of Mozambique in
2007, 2008 and 2009. The survey collected information on the characteristics of the
health facilities as well as opinions of workers and patients. Multivariate regression
analysis was used to measure the impact of the presence and extension of ART services
on the other services provided by the health facility. Wait time and time spent with the
health provider were used as indicators of service quality and patient satisfaction.
Results: We find that waiting times and time spent with the provider are not significantly
affected by the existence of ART service in a given health unit, and do not vary
significantly with the number of ART patients in treatment. Conclusions: For the
Mozambican case it seems that the health system is dealing relatively well with the
increasing provision of HIV/AIDS related services. No negative consequences for other
patients were identified and the treatment of other conditions is taking place at the same
level in facilities with ART and in those who have not yet implemented this service.
However, there is no sign of positive spillovers either.
5) “Assessing the Quality Of Life of HIV/AIDS Patients And Their Families In Ghana
During The Scale-Up of Anti-Retroviral Treatment Delivery” by C. Machingauta, J.
Amponsah, B. Ocran, S. Ahmed, S. Sory, R.A. Selby, S. Ayisi Addo, B.T. Dornoo, H. Kazianga,
M. Over, N. Akwei Addo, J.O. Gyapong, D. de Walque
Preliminary versions of this paper have been presented as poster at the IAEN PreConference Workshop and the XVIII International AIDS Conference in Vienna, July 2010.
Abstract
Background: Existing research that documents the negative impact of HIV/AIDS on
patient physical and mental health, and suggests spillover effects to family is mainly
qualitative. While patient health is expected to improve with antiretroviral intake,
the largely qualitative research on family rarely explores their physical and mental
health in a dynamic perspective. This paper aims to address these gaps by
quantitatively assessing the evolution of quality of life of Persons Living with
HIV/AIDS and their families in Ghana.
Methods: We use the two waves of survey data collected between 2007 and 2008.
The Medical Outcomes Survey Instrument was administered twice, an average of six
months apart. Responses from 507 patients, 415 family members and 387
neighbors for reference, were used to generate physical and mental health indices
which were regressed on respondent HIV status, controlling for survey round,
gender, age and individual fixed effects.
Results: On average, patients’ mean physical and mental health indices increased
from baseline to follow up, but remained significantly lower compared to family and
neighbor indices. Family members appear mentally worse off than neighbors at
baseline with at least 5 of 8 mental health indices significantly lower at the 5%
significance level. Further to a general improvement at follow-up, notable
improvements occur in family members’ mental health.
Conclusions: Evidence suggests that families of PLHA suffer a psychological burden
that may erode over time as patients improve physically and mentally. Patients
improve mentally and physically over time but remain worse off compared to family
and neighbors.