TYPE OF STUDY: Continuing

PROJECT TITLE: Economic and social outcomes of HIV and AIDS treatment.

NAME OF LOCAL CO-INVESTIGATOR: Dr. Ian Sanne, Clinical HIV Research Unit, University of the

Witwatersrand; Ms. Mpefe Ketlhapile, Health Economics Research Office, Wits Health Consortium.

TIME AND BUDGET SUMMARY:

Year of activity: Year 4

Year started: 2005 (PEPFAR funding since COP 07)

Expected year of completion: 2010

Budget received to date: $100,000 (in COP 07)

Additional budget requested in FY 2008: $150,000

PROJECT DESCRIPTION: Boston University (BU) and its local partner, the Health Economics Research

Office (HERO), will use FY 2008 funds to continue an ongoing evaluation of the social and economic

outcomes and sustainability of antiretroviral therapy (ART) in South Africa. Follow up of the existing patient

cohort will be extended to a minimum of three years, with some patients reaching four years of follow up.

Results will be used to improve treatment outcomes and enhance patient-level sustainability of treatment.

The target populations for this PHE are adults, people living with HIV, policy makers, program managers,

clinicians, treatment facilities, and USG staff.

EVALUATION QUESTION: The evaluation question for this PHE is "What are the social and economic

consequences of HIV and AIDS treatment for individual adult patients over the first four years of ART?"

Little is known about how treatment affects non-clinical indicators such as employment status, household

cohesion, quality of life, costs incurred to obtain treatment, and related impacts on patients and their

households. This information is needed to ensure that receiving ART is socially and economically, as well

as medically, sustainable for African patients and to estimate the overall benefits of treatment expansion.

PROGRAMMATIC IMPORTANCE: While the medical effectiveness of ART in suppressing viral replication

and restoring immune function is well established, little is known about the impact of treatment of HIV and

AIDS on the economic and social welfare of African patients. In particular, it is not known if treatment will

offset the impact of untreated AIDS on labor productivity, family stability, quality of life, and other indicators

of social and economic development and treatment sustainability. In FY 2005, BU and HERO launched an

evaluation of the economic and social outcomes of treatment for adult South Africans receiving care from

three PEPFAR-supported treatment sites. In FY 2008, this evaluation will be continued for a fourth year,

allowing examination of longer term outcomes essential to treatment sustainability.

The expected result of this activity is rigorous empirical information available about the non-clinical

outcomes of treatment for South African patients treated through PEPFAR and South African government

treatment initiatives. If patients are shown to be able to resume their normal activities, find and retain jobs,

maintain family stability, and improve quality of life, support for long-term provision of treatment and

expansion of current programs will be strengthened. Analysis of the characteristics of patients for whom

outcomes are less successful will also help improve treatment program design and patient support efforts.

PROJECT DESCRIPTION/ METHODS: This evaluation is a prospective cohort study using data from

patient interviews and medical records. Patients enrolled in the cohort are interviewed at intervals of 3-6

months, depending on the patient's status, during routine clinic visits using a questionnaire focusing on

family stability, ability to work and/or perform other normal activities, quality of life, adherence, costs of

obtaining treatment, and sources of income. This information is then linked to biomedical indicators of

treatment outcomes (e.g. CD4 count) from patients' medical records and analyzed longitudinally.

Over the course of FY 2005 and FY 2006, 618 ART patients and 451 pre-ART patients were enrolled in the

study and completed baseline and follow-up questionnaires. By the end of the FY 2007 funding period, all of

these patients will have been followed for a minimum of 2 years, and some for more than 3 years. In FY

2008 no new patients will be added, but because the impact of treatment on patients' welfare will change

over time, following the current patients for an additional year will generate valuable information about the

sustainability of treatment beyond the initial two years. Almost all of the pre-ART patients will have initiated

ART by FY 2008, allowing a pre- and post-treatment comparison for this group.

POPULATION OF INTEREST/GEOGRAPHIC AREA: The population of interest is all adult pre-ART and

ART patients in South Africa, as represented by the patients at the study sites. The study sites are located

in Gauteng and Mpumalanga Provinces.

STATUS OF STUDY/PROGRESS TO DATE: The study is progressing on schedule. Follow-up interviews

and data analysis are ongoing.

LESSONS LEARNED: Analysis of baseline and initial longitudinal data show significant improvements in

ART patients' ability to carry out their normal activities, quality of life, job performance, and other indicators.

The baseline results of this evaluation have been presented widely. Abstracts have been presented at the

2006 International AIDS Conference in Toronto and the 2006 HIV Implementers' Meeting in Durban. A

paper about the costs to patients of obtaining treatment has been published in the South African Medical

Journal, and two other manuscripts are currently under review by international journals. Data are now being

used for a pooled analysis of treatment outcomes being conducted by a collaboration of researchers from

throughout South Africa.

INFORMATION DISSEMINATION PLAN: The results of this evaluation will be disseminated as widely as

possible in South Africa and other resource-constrained countries, where information regarding non-clinical

outcomes of treatment are very limited or absent all together. Results will be presented as a priority to the

study sites and to the provincial and national Departments of Health in South Africa. Reports will be made

available for rapid dissemination on the BU and International AIDS Economics Association (IAEN) websites.

Finally, results will be presented at relevant conferences in South Africa and internationally, and journal

manuscripts will be submitted for publication as new findings become available.

Activity Narrative:

ACTIVITIES: In FY 2008, follow-up interviews representing data collection rounds 5-8 will be conducted with

the full cohort, medical record data will be extracted, and an analysis covering the period from pre-ART to 3

years on treatment will be completed.

BUDGET JUSTIFICATION FOR FY 2008 MONIES (USD):

Salaries/ fringe benefits: $ 112,000

Equipment: $ 3,000

Supplies: $ 5,000

Travel: $ 18,000

Participant Incentives: $ 2,000

Laboratory Testing: $ -

Other: $ 10,000

Total: $ 150,000

The majority of funds (75% of the total) will be used for salaries and benefits for study staff, including the

principal investigator/lead economist, field director, data manager, statistical analyst, and interviewers.

Travel (12% of the total) will include local transport for the study team and limited international and domestic

travel for Boston and Johannesburg based investigators. Approximately 2 computers will be purchased for

data entry and analysis (equipment, 3%). Supplies will include general office supplies, computer supplies,

and photocopying of data collection instruments (<1%). Small gifts (e.g. t-shirts) will be purchased as tokens

of appreciation for study subjects (1%). Finally, other expenses including office space and communications

will account for 7% of the total.

TYPE OF STUDY: Continuing.

PROJECT TITLE: Costs and Cost-Effectiveness of Adult Treatment Delivery. This evaluation is also

conducted with PEPFAR support in Kenya and Zambia.

NAME OF LOCAL CO-INVESTIGATOR: Dr. Ian Sanne, Clinical HIV Research Unit, University of the

Witwatersrand; Mr. Lawrence Long, Health Economics Research Office, Wits Health Consortium.

TIME AND BUDGET SUMMARY: This is the fourth year of this study. The evaluation started in FY 2006

and is expected to be completed in 2010. A total of $520,000 has been received to date.

PROJECT DESCRIPTION: Boston University (BU) and the Health Economics Research Office (HERO), will

use FY 2008 funds to expand an ongoing analysis of cost and cost-effectiveness of models of treatment

delivery in South Africa. The number of sites will be increased to strengthen understanding of differences

among delivery models, and an initial set of pediatric treatment sites will be analyzed. Results will be used

to inform planning by the USG and South African Government and improve treatment delivery.

EVALUATION QUESTION: The evaluation question is "What is the cost per adult and pediatric patient

treated and per patient in care and responding to ART 12 and 24 months after initiation under different

models of treatment delivery in South Africa? Little is known about the relative costs of different ART

delivery models in South Africa or the relationship between resource utilization and patient outcomes, and

there is no information about the costs of pediatric treatment. Such information required to make program

scale-up more efficient, estimate budgetary needs, and ensure sustainability.

PROGRAMMATIC IMPORTANCE: To achieve South Africa's goals for treatment of AIDS, ART must be

delivered in a wide range of settings and at multiple levels of the healthcare system. The characteristics of

the treatment facility (setting, type, sector, size, etc.) and of the patients treated (socioeconomic level,

condition at initial visit, etc.) are likely to affect the patient outcomes and costs incurred. BU was requested

in FY 2005-07 to examine cost and cost-effectiveness of alternative models of treatment delivery in South

Africa. The original methodology considered only the first 12 months following treatment eligibility and

included relatively small sample sizes. Initial results have raised new questions requiring larger samples

and longer periods of follow-up. Questions include the cost per successful outcome in years 3 and 4

following treatment eligibility; costs for subsets of patients, such as those who initiate treatment with very

low CD4 counts or who switch to second-line regimens during the first year; cost-effectiveness of treatment

delivery models launched after the original study sites were chosen; and costs of pediatric sites. In FY 2008,

the methodology will be amended and number of study sites and sample sizes increased to address these

issues and to generate more detailed information about the relationships between resource utilization

(costs) and patient outcomes. The methodology will be adapted to analyze pediatric treatment delivery

models.

The expected results of this activity are accurate and detailed estimates of the costs of delivering treatment

and achieving successful and sustainable outcomes across a wide range of settings and types of patients.

This information will assist the South African Government, PEPFAR, and other donors to estimate future

resource needs, increase efficiency among providers, and target future investments toward the most cost-

effective models of delivery.

PROJECT DESCRIPTION/METHODS:

Summary: BU and HERO developed the methodology used in this evaluation. The aim of the study is an

estimate of the average cost to place a patient in care and response to ART 12 or 24 months after initiation

at each study site. At each site, the study team selects a random sample of patients and conducts a

retrospective medical record review. Unit cost estimates provided by site management are used to estimate

the average cost per patient treated and average cost to produce a patient who remains in care and

responds to treatment.

Data collection: The evaluation relies on retrospective data routinely collected by treatment programs to

generate information about models of treatment delivery that are successfully treating the largest number of

patients at the lowest cost, important characteristics of delivery systems, and whether patient medical

outcomes are affected by the model and cost of treatment delivery. Data collected from medical records

include all resources used by the site to treat the sampled patients during the specified study period (drugs,

lab tests, outpatient visits, inpatient days, support services, infrastructure and other fixed costs) and patient

status (in care or not), available laboratory test results (CD4, viral load, etc.), and clinical condition

(presence or absence of any AIDS-defining conditions) at the end of the study period.

Data analysis: Each patient is assigned to one of three outcome categories: (1) in care and responding; (2)

in care but not responding; and (3) no longer in care (died or lost to follow up). The specific measure of cost

-effectiveness being used is "cost to retain a patient in care and responding to therapy," with "responding"

defined as an undetectable viral load, incremental increase in CD4 count, and/or absence of serious clinical

conditions. Both costs and outcomes are estimated at the 12-month and 24-month points following medical

eligibility for treatment under South African national guidelines. The algorithm used for assigning outcome

categories takes into account the varying types of information available for each patient, making the

approach widely applicable to diverse treatment sites.

Sites and samples: Between 2005 and 2007, analysis was completed at approximately 10 ART sites

representing various models of delivery. The models and sites represent the most promising or common

approaches to large-scale treatment delivery in urban and rural areas, and in the public and private sectors.

In FY 2008, approximately 5 adult and 3 pediatric treatment sites will be added to the study, and cost

estimates for up to 48 months of follow up will be made. The current sample size for the study is 200

patients per site and treatment duration cohort; this will be adjusted as needed to provide information about

sub-populations.

POPULATION OF INTEREST/GEOGRAPHIC AREA: The population of interest is all adult and pediatric

ART patients in South Africa. The study will be national in geographic scope, with a nationally

representative sample of sites selected.

Activity Narrative:

STATUS OF STUDY/PROGRESS TO DATE: The study is progressing on schedule. The protocol for the

study has been approved by the ethics committees of the University of the Witwatersrand and Boston

University, data collection instruments and analysis models have been developed and refined, and the

study team has been trained. Analysis has been completed for five sites and is underway at the sixth;

discussions with the next two proposed sites are also underway.

LESSONS LEARNED: Results from the first five study sites have shown that both costs and outcomes vary

among treatment sites and delivery models, but the magnitude of the differences is generally modest. When

costs and outcomes are taken into account, the differences are magnified, leading to quite different cost-

effectiveness estimates. The initial results have been presented widely and a manuscript is under review for

journal publication. Abstracts have been accepted and oral presentations made at the 2007 World Congress

of the International Health Economics Association (IHEA) in Copenhagen; the 2007 HIV Implementers'

Meeting in Kigali; and the 2006 HIV Implementers' Meeting in Durban.

INFORMATION DISSEMINATION PLAN: The results of this evaluation will be disseminated widely in South

Africa and other resource-constrained countries, where information regarding treatment delivery model

costs and outcomes are very limited or absent. After internal data/results verification and review, results will

be presented to provincial and national Departments of Health in South Africa. Reports will posted on the

BU and International AIDS Economics Association (IAEN) websites. Results will be presented at

conferences in South Africa and internationally, and journal manuscripts will be submitted for publication as

new findings become available. This evaluation is conducted with PEPFAR support in Kenya and Zambia,

and results from South Africa will be included in a multi-country comparison of the costs, outcomes, and

cost-effectiveness of treatment delivery models.

FY 2008 ACTIVITIES: In FY 2008, the methodology for analysis of pediatric sites will be developed and

methodology for adult sites will be amended. Analysis will be completed for approximately five additional

adult and three pediatric treatment sites, and cost estimates for up to 48 months of follow up will be made.

In addition, results from South Africa will be compared to those from Zambia and Kenya.

BUDGET JUSTIFICATION FOR FY 2008 MONIES (USD):

Salaries/ fringe benefits: $203,898

Equipment: $5,230

Supplies: $1,221

Travel: $63,727

Participant Incentives: $ -

Laboratory Testing: $ -

Other: $25,924

Total: $300,000

Sixty-eight percent of funds will be used for staff salaries. Staff consists of PI/lead economist, field director,

data manager, statistical analyst data collectors. Travel will include local transport for the study team and

international and domestic travel for Boston- and Johannesburg-based investigators. Three laptop

computers will be purchased for field data entry and analysis. Office supplies and photocopying of data

collection instruments will comprise less than 1% of the budget. Other expenses including office space and

communications will account for 9% of the total.