2006 Operations Research on AIDS Care and Treatment in Africa Grants

In 2006, nine teams received two-year grants of $200,000 each to support operations research on AIDS care and treatment in Africa. (Eleven additional teams received grants in 2005 as part of the same competition.)

2005 and 2006 ORACTA Grantees

The first ORACTA competition was held in 2005, which resulted in grants being awarded to 20 teams in late 2005 and early 2006.

ART Adherence Among People in Rural Zambian Clinics
Adherence Based Viral Load Triage in Botswana
Peer Educators Impact of HIV Medicine Adherence
Do Peer Counselors Promote Adherence to ARVs?
Evaluation of Antiretroviral Therapy Impact in Zambia
Models of Care for Antiretroviral Service Delivery
Impact of ART on Sexual Behavior in Kisumu, Kenya
HAART Delivery Models: A Quasi-Experimental Study
An Enhanced Adherence Support Program for HAART
Therapeutic Options for Women Exposed to Nevirapine
Cost-Effectiveness of Public-Private Partnerships
Establishing the Cost-Effectiveness of Different Models of Antiretroviral Treatment Programs Across Clinical Sites in Southern Africa in Urban and Rural Settings
Prevention of TB Mortality & HIV Related Hospitalization
Directly Observed, Community-Based Treatment in Nigeria
Understanding HAART Refusal and Piloting Targeted DOT to Improve HAART Adherence in an ARV Treatment Expansion Program in Kenya: Operational Evaluation and Cost-Analyses
Identifying Optimal HIV Care Approaches in Africa
Patient Level Sustainability of Antiretroviral Therapy for HIV/AIDS in South Africa
Decentralization of Pediatric HIV Care and Treatment in Kampala, Uganda
Pharmacy Based Monitoring of ART Programs
Extending HIV Care Beyond the Rural Health Center

ART Adherence Among People in Rural Zambian Clinics

Principal Investigator:
Gretchen L. Birbeck, M.D., M.P.H.,
Michigan State University

Co-Investigators:
Elwyn M. Chomba, M.Med., University Teaching Hospital Zambia
Alexis M. Sinyama, M.B.Ch.B., Zambia Sugar Plc.

Abstract:
More than 20 years into the AIDS epidemic in sub-Saharan Africa (SSA), antiretroviral therapy (ART) is finally becoming available to the general public. The initial roll out of ART occurred primarily at urban healthcare facilities, usually large teaching hospitals. However, a third to a half of the people living with AIDS in SSA reside in rural regions too distant from the cities to access AIDS-related care there. Now, ART is slowly becoming available through clinics located in rural regions of Zambia. The social, economic, medical, and cultural environment of rural regions of Africa differs greatly from the urban centers.

Unfortunately, little is known regarding the determinants of drug adherence in rural regions. To begin to explore patient and organizational-level determinants of ART adherence among patients attending rural clinics in Zambia, we propose to conduct a study at 3 rural ART clinics in the Southern Province of Zambia. By assembling a cohort of new attendees at these rural ART clinics, we will prospectively assess the impact of socioeconomic status, ART-toxicity, social roles/responsibilities, perceived levels of HIV-associated stigma, pre-treatment cognitive deficits, health systems limitations, and co-morbid depression upon ART adherence. Adherence will be quantified based upon pill counts and clinic attendance. Individuals who agree to community-based visits and fail to attend scheduled ART clinic visits will be sought at home to determine reasons for non-adherence. In addition to assessing potential factors associated with poorer adherence, the study team will work with non-adherent patients and ART clinic staff to overcome barriers to treatment adherence for individual patients as these patients are identified. Structured interviews of clinic staff and administration will also be conducted to assist us in understanding underlying causes for health systems limitations that negatively impact adherence.

Adherence Based Viral Load Triage in Botswana

Principal Investigator:
Gregory P. Bisson, M.D., University of Pennsylvania

Co-Investigators:
Ndwapi Ndwapi, M.D., Botswana Ministry of Health Department
Gaolathe Tendani, M.D., Botswana Ministry of Health Department
Robert Gross, M.D., University of Pennsylvania

Abstract:
Approximately 25 million of the estimated 40 million people living with HIV-1 infection reside in sub-Saharan Africa. Despite this, it is estimated that less than 8% of those who need treatment for HIV infection in this region are currently receiving it. The World Health Organization's "3 by 5" Initiative aims to treat 3 million HIV-infected individuals with highly active antiretroviral therapy (HAART) by the end of 2005, and ARV scale-up is now rapidly occurring in multiple countries in sub-Saharan Africa. The growing number of HIV-infected Africans on HAART is creating new challenges related to monitoring thousands of patients in areas where healthcare infrastructure is limited.

Virologic monitoring is critical to assessing response to HIV therapy, and decreasing the HIV viral load to levels below those of clinically used assays (i.e., to "undetectable" levels), is associated with reversal of HIV-related immune destruction. Viral load assays are, however, too expensive for routine use in the developing world, where the burden of HIV infection is greatest, and low cost approaches to virologic monitoring are needed. Soon after patients start HAART, however, changes such as increases in CD4 count (the main immune cell type affected by HIV) occur in those whose viral loads decrease. Moreover, CD4 cell assays are cheaper and more widely available than viral loads in the developing world. This creates an opportunity to use CD4 cell response as a marker for virologic response during treatment with HAART. Our previous work has shown that use of CD4 response threshold of 50 cells at the 6 month follow-up visit can identify over 50% of virologic failures while avoiding confirmatory viral loads in over 80% of those whose viral loads are undetectable. Another major predictor of virologic response is adherence to therapy, or, taking the medication as prescribed.

The specific aim of this project is therefore to develop a clinical prediction rule for achievement of an undetectable plasma HIV-1 RNA level 4-6 months after initiation of ARV therapy in sub-Saharan Africa. We are conducting a cohort study in the largest free HIV care clinic in Botswana to categorize patients into various levels of adherence and CD4 response, and then evaluating the public health implications of triaging viral load testing based on these low-cost parameters. For example, it may be most efficient to perform viral loads only in those with adherence less than 80% and CD4 cell response of less than 100 cells from baseline. The project will therefore approach a "test some" rather than a "test none" or "test all" approach to virologic monitoring. Since viral load monitoring is both important and expensive, testing strategies that reserve virologic testing for those who are at highest risk of treatment failure is critically important to the sustainability of HIV treatment programs in the developing world.

Peer Educators Impact of HIV Medicine Adherence

Principal Investigator:
William A. Blattner, M.D.,
University of Maryland Biotechnology Institute

Co-Investigator:
Maria K.L. Eng, Dr.PH.,
University of Maryland Biotechnology Institute

Abstract:
Peer educators—those infected with HIV/AIDS and/or those who share the same socioeconomic or cultural background as the local AIDS populations— are effective advocates and partners for medication treatment adherence and represent a cost effective and proven labor source for supporting adherence programs. These community-based educators can also resolve barriers to medication adherence (such as stigma, low self-esteem, and adverse drug side effects) and improve pill taking behaviors since they are familiar with the disease and local culture and attitudes towards ARV treatment.

This project explores the effectiveness and sustainability of two community-based peer educator interventions compared to hospital-based standard care practices. The hospital services a largely Muslim population in northern Nigeria and is a collaborator with the President's Emergency Plan for AIDS in Africa (PEPfAR).

We have two aims. The first aim is to compare HIV medication adherence between three different interventions: (1) typical HIV care, support and education as offered by the hospital's adherence staff; (2) daily medication reminders via text messaging from peer educators and encouragement and observation of daily adherence by a home-based treatment coach; or (3) activities as mentioned in 2 plus home visits by peer educators via motorcycle.

The second aim is to determine how sustainable these interventions are by evaluating the patient's continuity in the program and their clinical, virologic and immunologic improvement. Since excellent adherence is important from the beginning of treatment, we will target those HIV-infected adults who have been newly prescribed HIV medications.

This study is relevant because aside from anecdotal evidence, few studies have demonstrated that peer educators are effective. We attempt to demonstrate that (1) peer educators who support and guide home-based adherence can significantly improve HIV medication adherence when compared to a facility-based adherence support service, and (2) community-based peer educator outreach models are sustainable in the Nigerian context of low education, high levels of stigma, and limited resources and infrastructure. This will hopefully provide evidence-based data for adjusting HIV treatment adherence standards of care.

Do Peer Counselors Promote Adherence to ARVs?

Principal Investigator:
Patricia Bright, R.N., M.S.P.H., Ph.D., Johns Hopkins School of Medicine

Co-Investigators:
Francis Mmiro, M.B.C.H.B., M.R.C.O.G., Makerere University Medical School, Uganda
Philippa Musoke, M.B.C.H.B., M.M.E.D., Makerere University Medical School, Uganda
Laura Guay, M.D., Johns Hopkins School of Medicine

Abstract:
This research project is a randomized intervention study that is comparing two HIV/AIDS care program models and involves 300 HIV-infected women recruited from the Prevention of Mother-to-Child Transmission (PMTCT) program at Mulago Hospital, Kampala, Uganda. The study is evaluating compliance with anti-retroviral therapy (ARVs) recommendations by measuring drug adherence, regular clinic attendance at scheduled visits, and laboratory/clinical outcomes. The same standard of HIV/AIDS care is being provided to participants in both models including free access to ARVs, medical consultation, standard Ministry of Health approved laboratory monitoring tests, and most essential opportunistic infection medicines. All participants are also being encouraged to attend peer psychosocial support group meetings. However, the intervention model (1) is evaluating the effect of peer counselors and home visiting in combination with a greater emphasis on nurses for routine visits and longer intervals between scheduled visits. The standard clinic-based model (2) will promote adherence through routine counseling at each scheduled visit, care provided by a medical officer at each visit, and shorter intervals between visits. In addition, the use of cell/mobile phones will be assessed on the frequency and immediacy of provider-patient contact within both models, using a 2x2 factorial design. All participants will be followed for 12 months or until the end of the study period. The goal is to identify low-cost programmatic elements which directly affect optimal adherence to ARV therapy among women participating in a PMTCT program at this large public hospital in Uganda. 

Evaluation of Antiretroviral Therapy Impact in Zambia

Principal Investigator:
Benjamin H. Chi, M.D.,
University of Alabama at Birmingham

Co-Investigators:
Jeffrey J. S. Stringer, M.D., M.P.H., 
University of Alabama at Birmingham

Moses Sinkala, M.P.H., Lusaka District Health
Management Team, Zambian Ministry of Health

Abstract:
Over the past 2 years, the number of people starting antiretroviral therapy (ART) has grown rapidly. Much of this has occurred in sub-Saharan Africa, where more than half-a-million HIV-infected individuals have initiated therapy. While this is extremely encouraging from a humanitarian perspective, the effects of rapid ART expansion are yet unknown. It is possible that therapies developed in places like the United States and Europe may not lead to the best course of HIV/AIDS treatment in Africa.

In support of the Zambian national program for HIV/AIDS, we created a comprehensive individual-level database for patient follow-up and program reporting. Now implemented in over 18 sites across the country, this database contains medical information for over 30,000 HIV-infected patients. Because of its original design and objectives, however, we face problems with data completeness and patient outcomes ascertainment. Improvements in both areas are needed to perform meaningful longitudinal analyses.

The project will convert our current program database to one suitable for broad epidemiological research. Activities supported by the grant include: (1) improving patient follow-up and outcome ascertainment, (2) enhancing the quality of clinical care, (3) monitoring the quality of data entry, and (4) training young Zambian investigators in medical informatics and data management. Using this enhanced database, we will be able to investigate many relevant issues for HIV/AIDS treatment in Africa, including:

• Incidence of and predictors for antiretroviral drug toxicity in African settings
• Comparative effectiveness of brand-name versus generic antiretroviral formulations
• Population-level effectiveness of nevirapine or efavirenz-containing HAART among women with prior single-dose nevirapine exposure to prevent mother-to-child HIV transmission

As donor and host nations look to assess the impact of HIV treatment programs, we believe this type of population-level information will be critical to the continued support of ART in many parts of Africa.

Models of Care for Antiretroviral Service Delivery

Principal Investigator:
David Coetzee, M.B.B.S., M.S., D.T.M. & H., F.F.C.H. (SA), University of Cape Town

Co-Investigators:
Andrew M. Boulle, M.D., University of Cape Town
Landon Myer, Ph.D. University of Cape Town
Susan M. Cleary, M.A., University of Cape Town

Abstract:
HIV care in South Africa, including antiretroviral therapy (ART), is delivered at different levels around the country. With the large number of persons requiring antiretroviral therapy, a major challenge now is rapid extension and the identification of ways to best expand the service platform for HIV. This project will identify the models of care in existence in three provinces in South Africa and identify the advantages and disadvantages of each. Issues will include the location of care, whether ART services are best delivered centralized or decentralized, whether ART should be hospital initiated with clinic referral or clinic initiated, the specialization of the service (ART only or comprehensive HIV care), vertical or integrated with TB or with all primary health care services, doctor or nurse driven, the role of nurses (whether or not they provide clinical care), the role of community-based adherence supporters, the different mechanisms for referral and access to higher levels of care.

Impact of ART on Sexual Behavior in Kisumu, Kenya

Principal Investigator:
Craig R. Cohen, M.D., A.B. M.P.H.,
University of California, San Francisco

Co-Investigators:
Stephan C. Shiboski, Ph.D., M.S., B.S.
University of California, San Francisco

Eliza A. Bukusi, M.P.H., M.MED., M.B.Ch.B.
Center for Microbiology Research, Kenya Medical Research Institute

Anjali Sharma, Sc.D., B.A.,
Center for Microbiology Research, Kenya Medical Research Institute

Abstract:
The effect of antiretroviral therapy (ART) scale-up in Africa on sexual risk behaviors and the prevention of new HIV infections is unknown. Models demonstrating that ART can reduce HIV incidence also show that small increases in risky sexual behaviors can mask and reverse gains achieved by therapy. Studies from the developed world suggest that prevalence of unprotected sex and incidence of sexually transmitted infections have increased since the introduction of ART. It is possible that misperceptions about ART or reduced concern about HIV because of ART availability lead to more permissive sexual behavior in both HIV-infected persons and the general population.

The purpose of this study is to examine the association of ART expansion and ART-related beliefs with risky sexual behaviors and prevalence of sexually transmitted infections (STI) and HIV in the general population of Kisumu, the municipality with the highest HIV seroprevalence in the country. Similarly, the association of ART expansion and ART-related beliefs with risky sexual behaviors in HIV-infected persons receiving ART will also be explored. We will conduct cross-sectional surveys at baseline and one-year later to assess changes in knowledge and beliefs about ART and the impact of these changes on sexual risk behaviors. Surveys of the general population will be conducted by sampling 20 sentinel clusters within the Municipality of Kisumu. Surveys of HIV positive patients on ART will be conducted at our PEPFAR-funded ART rollout program in Kenya, entitled Family AIDS Care and Education Services (FACES). FACES began delivering comprehensive care to the HIV-affected population in Kisumu in March 2005. Understanding public perceptions of ART and their effects on sexual risk behaviors that lead to an increased risk of STI and HIV is essential for designing prevention interventions that address the changing landscape of HIV with the availability of ART. The results of this study will be widely disseminated throughout Kenya, as well as in international publications, so that our results may guide treatment programs in their efforts to improve incorporation of prevention services into HIV care programs.

HAART Delivery Models: A Quasi-Experimental Study

Principal Investigator:
Stephen S. Gloyd, M.D., M.P.H.,
Health Alliance International

Co-Investigators:
Kenneth H. Gimbel-Sherr, M.P.H.,
University of Washington/ Health Alliance International

James T. Pfeiffer, Ph.D., M.P.H.,
University of Washington/ Health Alliance International

Abstract:
A primary bottleneck encountered across all large-scale AIDS treatment initiatives is the lack of trained human resources. Mozambique has one of the lowest ratios of health providers to population in the world, with 2.4 medical doctors and 48 nurses per 100,000 people. The intractability of the human resource shortage highlights the need to define the most efficient and effective models for health care teams responsible for delivering AIDS treatment (HAART).

The overall goal of this study is to define an optimal structure and composition of HAART service delivery teams. The primary study aim is to determine which of three existing models for the structure and composition of service delivery teams results in the most efficient and effective delivery of HAART among Mozambican public-sector health facilities. The secondary study aims are to assess the cost-effectiveness of the different human resource models and establish a patient database for long-term follow-up of clinical response, mortality, and adherence to HAART regimens.

This study will compare three models for team composition and management, including 1) a medical-doctor focused model, 2) a mixed approach of medical doctors and mid-level cadres collectively seeing patients and making joint decisions, and 3) a mid-level cadre focused model working under the supervision of medical doctors. Outcomes include efficiency measures, quality of care measures, patient/provider satisfaction, and cost of services provided. The proposed study is expected to provide insight into optimal approaches for designing teams to deliver HAART, directly benefiting Mozambique and other countries working to rapidly, safely and effectively increase HAART coverage with limited human resources.

An Enhanced Adherence Support Program for HAART

Principal Investigator:
Salim S. Abdool Karim, M.B.Ch.B., Ph.D.,
Columbia University

Co-Investigators:
Francois van Loggerenberg, M.A.,
University of KwaZulu-Natal

Kogieleum Naidoo, M.B.Ch.B.,
University of KwaZulu-Natal

Abstract:
This randomized-controlled study aims to compare the effectiveness of two adherence support strategies for use in patients with HIV-related TB in the context of CAPRISA AIDS Treatment program (CAT) in KwaZulu-Natal, South Africa; the enhanced adherence support program (E-ASP) or the standard adherence support program (S-ASP) used as part of the National roll-out program. The E-ASP program has been designed to be practical within the resource-constrained setting for which this study is planned and uses concepts adapted from Fisher and Fisher's (1992) Information-Motivation-Behavioral Skills (IMB) model of HIV prevention behavior change. It utilizes limited but structured patient-health educator contact, and counseling techniques drawn from motivational interviewing. Adherence assessment at each visit by the study nurse and pharmacist using self-report and pill counts. This will be correlated with viral load and CD4+ T cell count at 2, 3, 12 and 18 months post initiation of ART. These measures are also used to determine treatment success. 

Therapeutic Options for Women Exposed to Nevirapine

Principal Investigator:
Louise Kuhn, Ph.D., Columbia University

Co-Investigator:
Wei-Yann Tsai, Ph.D., Columbia University

Abstract:
Single-dose nevirapine (sd-NVP) is a simple and effective intervention to prevent mother-to-child HIV transmission (PMTCT) in low resource settings. Unfortunately, several studies have observed evidence of drug resistance to nevirapine among HIV-infected women soon after sd-NVP for PMTCT. If this "resistance" persists long-term and has adverse clinical consequences, then treatment options for women may be curtailed. Nevirapine and other drugs in the same class are an important part of therapeutic antiretroviral drug regimens. In our study, we propose to investigate whether allowing a longer time since exposure to sd-NVP to elapse can reduce any subsequent consequences on a woman's response to a treatment regimen that contains nevirapine. We hypothesize that if more than 18 months has elapsed since sd-NVP then there will be no adverse effect on a woman's response to therapy. However if therapy is started sooner, then some sd-NVP-exposed women may not suppress virus as easily as women who have never previously been exposed to nevirapine when they are started on a nevirapine-containing regimen. We will investigate this question in the context of a study of HIV-infected women at two clinics in Lusaka, Zambia, who received sd-NVP for PMTCT and who now have access to antiretroviral treatment through the government of Zambia and President's Emergency Plan for AIDS Relief HIV care and treatment programs. If we could identify a length of time after exposure to sd-NVP that no longer conveys any measurable impact on response to therapy, then we could significantly reduce the numbers of women whose treatment options would need to be limited. As HIV treatment programs are put in place, sicker women should be placed on treatment before and during pregnancy leaving few HIV-positive women who will need drug treatment soon after delivery. 

Cost-Effectiveness of Public-Private Partnerships

Principal Investigator:
Gary Maartens, M.B.Ch.B., M.MED.,
University of Cape Town

Co-Investigator:
Susan M. Cleary, M.A., University of Cape Town

Abstract:
In sub-Saharan Africa there is an enormous burden of people with HIV/AIDS who currently require treatment with drugs directed against HIV (known as antiretrovirals). Furthermore, as HIV is a progressive disease, nearly all of those currently infected (as well as those who will become infected) will eventually require antiretroviral therapy. Recent estimates are that 10-15% of people in Africa with HIV/AIDS who currently qualify for antiretroviral treatment are receiving it. Infrastructure in the health care sector is often very poor in Africa. One potential solution to increase the number of people able to access care is to utilize the private sector to treat public sector patients, forming public-private partnerships. In many developing countries the private sector is better resourced than the public sector. For example, in South Africa two thirds of the registered medical practitioners work in the private sector. One barrier to the public-private partnership is concern from governments that treating patients in the private sector will be considerably more costly. This study will assess the cost-effectiveness of a President's Emergency Plan for AIDS Relief-funded public-private partnership in South Africa. We aim to compare this with two completed cost-effectiveness studies carried out in public-sector antiretroviral programmes, and to use this data to build a model for public-private partnerships. 

Establishing the Cost-Effectiveness of Different Models of Antiretroviral Treatment Programs Across Clinical Sites in Southern Africa in Urban and Rural Settings

Principal Investigator:
Richard G. Marlink, M.D.,
Elizabeth Glaser Pediatric AIDS Foundation

Co-Investigators:
Robert A. Pawinski, M.B.Ch.B.,
Nelson R. Mandela School of Medicine

Helga Holst, M.D., McCord Hospital

Abstract:
Sub-Saharan Africa is the area most affected by HIV/AIDS in the world. It is currently in a phase of rapid scale-up of the largest antiretroviral treatment (ART) programs worldwide. Little evidence exists to assist policy makers with optimally utilizing resources for maximal clinical outcomes.

This two-year operational research program is a collaboration between the Elizabeth Glaser Pediatric AIDS Foundation, the Enhancing Care Initiative, KwaZulu-Natal (KZN), Nelson R. Mandela School of Medicine and McCord Hospital. The overall goal of the project is to establish the cost-effectiveness of different models of antiretroviral treatment programs across clinical sites in southern Africa in both urban and rural settings.

The aims of this study include comparing the immunological, virological and clinical outcomes of ART under different models of care; estimating the financial and economic cost of providing a comprehensive model of ART at each site; and identifying the incremental costs and cost-effectiveness of differences within models of ART (e.g. doctor-centered approaches versus nurse-centered approaches or urban versus rural approaches to care).

To achieve these goals, we will examine the clinical outcomes, financial and economic costs and cost-effectiveness of ART programs at three research sites in KZN Province, South Africa. A secondary but important outcome of this study will be to build monitoring and evaluation capacity at each implementation site, as well as enhanced patient tracking systems.

The information generated by this study can contribute to answering questions about the optimal use of human resources and cost-effectiveness of different models of ART more generally, as well as provide opportunities to answer important clinical questions.
The results of this study will be useful in informing policy-making and HIV treatment guidelines in resource-limited settings, where ART needs to be scaled up. The investigators hope that the results of this study will assist with optimizing care to millions of patients infected with HIV who will be receiving HAART in the future.

Prevention of TB Mortality & HIV Related Hospitalization

Principal Investigator:
Neil A. Martinson, M.B.B.Ch., M.P.H., M.F.G.P.,
Johns Hopkins University School of Medicine

Co-Investigators:
Charles Holmes, M.D., M.P.H., Harvard Medical School
Richard E. Chaisson, M.D., Johns Hopkins University School of Medicine

Abstract:
Outpatient care for HIV-infected patients has expanded in South Africa, in part due to funding from PEPFAR and the Global Fund. At the Perinatal Health Unit (PHRU) affiliated with the University of the Witswatersrand, and Chris Hani Baragwanath Hospital (CHBH), a PEPFAR sponsored HIV wellness clinic provides primary HIV care to over 1,500 patients prior to qualification for ART. In addition, PEPFAR and Global fund-sponsored clinics at the site have enrolled nearly 2,000 patients on ART. To date, less attention has been paid to the integration of inpatient services for these patients, and to the causes and outcomes of inpatient utilization both prior to, and after the initiation of ART.

In the case of TB, it has been shown that up to 60% of patients with TB in the Gauteng region of South Africa pass through the inpatient wards of the CHBH at some point during their illness. Inpatient visits carry high costs to the health system, and for many opportunistic infections, high rates of death among inpatients. For instance, inpatient mortality rates in TB patients admitted to CHBH are approximately 20%. Despite a G8 global health target set in Okinawa in 2000 to reduce TB deaths by 50% by 2010, TB deaths are now increasing in South Africa, driven in part by high HIV prevalence. Development of methods to combat extremely high inpatient mortality rates is another important step in the process of improving HIV care in South Africa, and will allow high risk patients to benefit from efforts to improve HIV care and survive to access antiretroviral treatment.

In response to the needs identified above, we propose the following:

Aim 1: To examine utilization of inpatient healthcare services, causes and outcomes of hospitalization among patients followed in PEPFAR-funded wellness and antiretroviral therapy clinics in Soweto, South Africa.

Aim 2: To determine predictors and causes of in-hospital death among patients admitted to hospital with clinical TB and either with or without HIV co-infection.

Aim 3: To construct and validate a simple scoring system that allows for accurate prediction of in-hospital mortality in South African TB patients, with and without HIV co-infection.

The research proposed in this application stems from needs identified by providers and program managers of the ongoing AIDS care and treatment programs in Soweto. It targets the causes and rates of inpatient usage, and high inpatient mortality rates in TB patients, issues with high impact for HIV-infected patients in the area. Both of these themes seek to use the tools of operational research to gain a better understanding of the problems, such that interventions to address them can be developed and employed. The team is lead by a PI of the PEPFAR-funded HIV programs in Soweto, South Africa, and includes an experienced group of co-investigators and collaborators with clinical and methodological expertise from South Africa, Johns Hopkins, and Harvard, all committed to the success of this project.

Directly Observed, Community-Based Treatment in Nigeria

Principal Investigator:
Robert L. Murphy, M.D., Northwestern University

Co-Investigators:
John A. Idoko, M.D., Jos University Teaching Hospital
Babefemi O. Taiwo, M.D., Northwestern University

Abstract:
Adherence to the prescribed antiretroviral regimen is the cornerstone of successful HIV/AIDS treatment, and evidence from clinical experience and research is that optimal adherence requires the collaboration of patients, healthcare personnel, and the healthcare system. It appears that this collaboration can be further strengthened, and adherence enhanced, if the latent support mechanisms that are within the patient's social network are activated and utilized.

This research project follows our previous study at Jos University Teaching Hospital (JUTH), in Jos, Nigeria, the biggest of the 6 Harvard Track 1.0 PEPFAR sites in Nigeria. At JUTH, we have already preliminarily demonstrated that taking directly observed, community-based antiretroviral therapy appeared to improve adherence and virologic outcomes. In our proposed study, 500 HIV-infected patients at JUTH who initiate combination antiretroviral therapy will be randomized into two adherence strategy intervention groups: 1) community-based, directly observed therapy (D-CART), and 2) patient administered therapy (PAT). The interventions in both groups will be similar, except that patients in the D-CART arm will select a treatment partner (spouse, sibling, other relative or friend) who will directly observe the ingestion of at least one of the daily doses of the antiretroviral regimen, and assist the patient in other adherence promotion activities.

The virologic and immunologic responses of patients in both groups will be measured, and the resistance mutations that develop in those who fail treatment will be assessed. This is important because the risk of virologic rebound with resistance appears to be significant among patients who take < 90%="" but=""> 70% of their antiretroviral drugs 1. Thus, it is uncertain which adherence strategy will lead to more resistance as either strategy may lead to adherence rates that fall within the danger zone for resistance development. Most initial regimens in resource-limited settings are based on non-nucleoside reverse transcriptase inhibitors and lamivudine, which have low genetic barriers to resistance.

Increasing the pool of persons who are informed about HIV treatment, and who are involved in the care of HIV-infected persons, has the potential to greatly benefit society as well as the patients. We will assess whether D-CART has the potential to add to this pool by evaluating changes in the treatment partners' perception of HIV infection, HIV-infected persons, and their potential role in mitigating the epidemic.

The specific aims of this study are to:

1. Scale-up a D-CART model among HIV-infected patients who are receiving care in the PEPFAR program at JUTH, Nigeria, and determine the effectiveness and durability of the approach by comparing the virologic and immunologic responses of patients enrolled in D-CART to those of patients who self-administer their ART (PAT).
2. Characterize the resistance mutations in patients failing D-CART and those failing PAT.
3. Assess the impact of D-CART on patients and the community-based treatment partners.

Understanding HAART Refusal and Piloting Targeted DOT to Improve HAART Adherence in an ARV Treatment Expansion Program in Kenya: Operational Evaluation and Cost-Analyses

Principal Investigator:
Mkaya Mwamburi, M.D.,
Tufts University School of Medicine

Co-Investigators:
Christine A. Wanke, M.D., Tufts University School of Medicine
Ira Wilson, M.D., M.S., Tufts-New England Medical Center
David M. Kent, M.D., M.S. Tufts-New England Medical Center
Omu A. Anzala, Ph.D., M.B.Ch.B., University of Nairobi

Abstract:
Kenya has received international funding to expand HIV treatment in the public sector. HIV treatment significantly improves health and reduces transmission of HIV. However, for the treatment to be effective, patients must accept to start the treatment when it is offered and, when they do start the treatment, adhere strictly to the treatment protocols. Early reports suggest that refusal of HIV treatment and poor adherence may become major long-term problems. In a collaborative effort with the University of Nairobi Institute of Tropical and Infectious Diseases, Tufts University, Nutrition/Infection Unit this project aims to establish why some patients who have been identified as needing HIV treatment refuse to start the treatment when offered and to understand what influences the patients' decision to start the HIV treatment. For those who accept HIV treatment, we hope to assess the adherence to HIV medications and to evaluate the feasibility, acceptability and sustainability of targeted directly observed therapy (TDOT) for treatment. In TDOT, severely ill patients are identified and offered additional support through daily monitoring in order to reduce the risk of death and severe complication from HIV medications. We also plan to assess the economic implications of the HIV treatment program in Kenya as well as the implications of refusing to start HIV treatment when it is first offered. Addressing treatment refusal and poor adherence early in the treatment campaign is important for the success of HIV treatment provision in Africa. 

Identifying Optimal HIV Care Approaches in Africa

Principal Investigator:
Denis Nash, Ph.D., M.P.H., Columbia University Mailman School of Public Health

Co-Investigators:
Batya Elul, Ph.D., M.Sc., Columbia University Mailman School of Public Health
Wafaa El-Sadr, M.D., M.P.H., Columbia University Mailman School of Public Health

Abstract:
This project endeavors to identify optimal approaches to HIV care and treatment in Africa by conducting a multi-site and multi-level operational research study of the characteristics of HIV care and treatment programs and outcomes among individuals receiving care at these sites. This research is unique in that it examines site-level and contextual factors associated with patient outcomes for patients in advanced stages of HIV disease receiving ART (both adult and pediatric) as well as patients in earlier stages of HIV receiving non-ART care. The specific aims are as follows:

1. To assess the degree of variation in patient outcomes (e.g., adherence to ART, retention in care, immunological response to ART, and survival) across HIV care and treatment delivery sites, independent of the differences in the characteristics of patient-level factors (e.g., age, sex, stage of disease) across sites.
2. To identify site-level and contextual factors that are associated with HIV care and treatment outcomes, after adjusting for patient-level factors.
3. To assess the costs and clinical benefits of modifying site-level factors that appear to influence HIV care and treatment outcomes and quality adjusted life years (QALYs).

Data will be collected from two sources. First, we will extend our existing routine monitoring and evaluation field methods and data collection efforts at Columbia-supported President's Emergency Plan for AIDS Relief (PEPfAR) sites (n=29) in order to combine core HIV care and treatment data on patients across sites, including Mother-to-Child Transmission (MCTC)-Plus sites (n=14). The combined data set will include data on a growing number of patients at 43 sites throughout sub-Saharan Africa (including 26,107 patients, 9,687 (37%) of whom are on ART [8,752 adults and 935 children] as of June, 2005). Second, systematic site assessments will be conducted at the 43 Columbia-supported PEPFAR and MTCT-Plus sites every six months in order to capture information about site-level and contextual factors that may influence HIV care and treatment program outcomes. Data from systematic site assessments will be merged with data on patients receiving care and treatment. Using the final merged dataset with patient-level and site-level/contextual information, we will conduct data analyses to determine the relations between individual-level factors, site-level/contextual factors, and HIV care and treatment program outcomes. Finally, we will conduct a cost-benefit analysis to compare the clinical benefits and costs of various care models using cost estimates collected from a sample of sites during the study period.

This work is vital to identifying important programmatic aspects of HIV care and treatment programs for use in planning future programs or improving existing ones. The information gathered as part of this activity will suggest novel interventions or policies, which take into account factors at multiple levels, and could thereby improve care and treatment program outcomes in resource limited settings in the immediate future.

Patient Level Sustainability of Antiretroviral Therapy for HIV/AIDS in South Africa

Principal Investigator:
Sydney B. Rosen, M.P.A., Boston University

Co-Investigators:
Mary Bachman, D.Sc., M.P.H., Boston University
Ian M. Sanne, M.B.B.C.H., F.C.P. (SA), DTM&H, University of Witwatersrand

Abstract:
In order for the HIV/AIDS treatment programs now being scaled up to be successful, they must be sustainable beyond the current period of substantial international funding and high levels of political interest. Antiretroviral therapy for HIV/AIDS requires a lifelong commitment from patients as well as providers. For individual patients, both adherence to a specific treatment regimen and retention in the treatment program are essential to success.

In this study, we will investigate the reasons for and determinants of long-term patient retention, including the willingness and ability of individual patients to remain in an antiretroviral treatment program and to come to the treatment clinic for scheduled visits beyond the initial 6-month period when the symptoms of HIV/AIDS are diminishing. The study will estimate the rate of ARV patient retention and attrition at public treatment sites in South Africa, describe the main reasons for attrition for patients with different characteristics and receiving care from different types of facilities, and determine whether the probability of dropping out varies with specific patient or clinic characteristics.

The study will recruit patients currently receiving antiretroviral therapy at an urban, a semi-urban, and a rural clinic in South Africa to participate in the study. Patients will be interviewed upon recruitment into the study (at baseline) and then followed for at least one year. The baseline questionnaire will focus on issues that are hypothesized to affect retention (e.g. costs, side effects, and stigma). Patients who have dropped out of the program will be located using the contact information collected at the baseline interview and interviewed again to determine reasons for discontinuation.

Results of the study will improve the information available for long-term ART program design, patient care, and public sector planning. It will identify the most common reasons for patient attrition and allow better targeting of interventions, indicate whether the therapeutic regimen itself can be improved to promote patient retention, and provide more accurate information on expected numbers of patients in later years of ART rollouts.

Decentralization of Pediatric HIV Care and Treatment in Kampala, Uganda

Principal Investigator:
Heidi Schwarzwald, M.D., M.P.H., Baylor College of Medicine

Co-Investigators:
Ezekiel Mupere, M.D., PIDC Mulago Hospital
Addy Kekitiinwa, M.D., PIDC Mulago Hospital

Abstract:
It is estimated that 1,750 children are newly infected with HIV every day. Although this figure represents 13 percent of all incident infections globally, only one percent of those receiving antiretroviral medications are children. According to recent estimates by UNICEF only 15-25,000 children are receiving antiretroviral therapy (ART) globally.

The Uganda-Baylor Children's Center of Excellence (COE) is located on the campus of Mulago Hospital in Kampala, Uganda, the result of collaboration between Mulago, the pediatric faculty of Makerere University, and Baylor College of Medicine. The center provides care for over 2,000 children, 600 of whom are on antiretroviral therapy. The center sees approximately 170 children each day for treatment, screening, and prevention of mother-to-child transmission (PMTCT) follow-up.

Because it is one of the few places where HIV-infected children can receive care and treatment in Uganda, the center is quickly approaching maximum capacity. As a result, the Uganda-Baylor Children's Center of Excellence has begun to develop linkages with other regional health care facilities in order to refer patients for routine care and follow up. This critically important move to decentralize care for HIV-infected children benefits families by providing care that is closer to the patients' home, builds capacity for care and treatment outside the principal urban clinic, and liberates staff at the Center in Kampala to treat critically ill patients and initiate greater numbers on ART.

The proposed study seeks to monitor and evaluate the process of decentralizing pediatric HIV/AIDS care and treatment by examining the impact on the quality of care, the total numbers of children on ART, and health care costs associated with a decentralized system of care. Moreover, this study will elucidate some of the important issues (such as policy, creating linkages, training, etc.) related to the process of decentralizing pediatric HIV/AIDS care.

Pharmacy Based Monitoring of ART Programs

Principal Investigator:
Robin Wood, M.D., University of Cape Town

Co-Investigators:
Ulrike K. Rivett, Ph.D., University of Cape Town
Linda-Gail Bekker, M.D., Ph.D, University of Cape Town

Abstract:
There are many constraints to the scaling up of access to antiretroviral treatment in resource-poor settings and it is important to efficiently utilize existing scarce resources. This project addresses the problems of maintaining a secure and well-monitored drug supply to outlying clinics where on-site pharmacist support is impracticable. In South Africa there are 11,145 registered pharmacists of whom only 1,740 (17%) are employed in the public sector and even fewer engaged in the national antiretroviral program. The proposed project will maximize the capacity of a pharmacist to rapidly increase patient numbers and maintain a drug supply chain from time of stock entry to the central pharmacy, through the dispensing process, up to drug delivery to any individual patient. The planned system is not dependent on maintaining sophisticated computer technologies at the peripheral clinic sites, which would be a significant constraint to the use of any system in rural and remote sites.

This project embodies a unique approach to facilitating the development of electronic assistance in health care driven by moving key information on an appropriate technology base and enhancing peripheral capacity development essential to properly manage HIV/AIDS. The system addresses the challenge of the remote tracking of drugs once dispensed from a central pharmacy using a simple bar-coding system and monitoring this information using the hub pharmacy as an information node. The peripheral activity data will be integrated with central stock and dispensing logs to maintain peripheral drug supply without the need for large peripheral drug stocks.

The proposed system will develop central pharmacy stock control, peripheral site stock control, generation of sponsor-required reports and programme evaluation information. Real-time programme evaluations would include dates and numbers of patients initiating therapy, dates and times of drug switches, retention of patient on treatment, adherence (missed prescriptions) and regimen switches.

Extending HIV Care Beyond the Rural Health Center

Principal Investigator:
Kara Wools-Kaloustian, M.D.,
Indiana University School of Medicine

Co-Investigators:
Sylvester Kimaiyo, M.B.Ch.B., M.MED., Moi University Faculty of Health Sciences
John E. Sidle, M.D., Indiana University School of Medicine
William M. Tierney, M.D., Indiana University School of Medicine

Abstract:
UNAIDS estimates that 2 million of Kenya's 29.5 million people are currently infected and that 1.5 million have already succumbed to HIV. It is known that antiretroviral therapy is effective in African populations; however resources for HIV care are limited. In particular, trained medical professionals are sorely lacking in most of sub-Saharan Africa. It will be impossible to meet the needs of HIV infected patients in sub-Saharan Africa if we continue to rely on a physician and clinical officer (CO) based model for care delivery. Yet, data on alternative care delivery systems are limited.

The long term goal of our group is the development and optimization of an HIV care delivery system that will serve as a model for sub-Saharan Africa. The objective of this application is to develop and assess a system which uses non-clinician extenders to provide selected aspects of HIV care in rural western Kenya. The plan is to train persons living with HIV/AIDS (PLWAs) to undertake this role. Our central hypothesis is that PLWAs can be effective members of the health care team and that their involvement in community-based HIV care will facilitate patient access to services and improve outcomes. We base this hypothesis on four years of experience providing HIV care in western Kenya as well as our familiarity with integrating PLWAs into our HIV care system. Our group is uniquely qualified to carry out this work as Dr. Kimaiyo (Co-PI) is the Clinical Director for the AMPATH HIV Clinics affiliated with the Moi Faculty of Health Sciences (MUFHS, Eldoret, Kenya). Dr. Wools-Kaloustian, the PI, has been responsible for clinical protocol development and currently practices in one of the rural clinics. As such, our aims are:

Specific Aim 1: To develop a sustainable system to extend HIV care into the community and to train the individuals necessary to support such a system.

• Hypothesis 1a: HIV infected individuals (Community Care Coordinators) can be trained to assess and support medication adherence, monitor for adverse drug effects, trouble shoot common health and social problems, and perform simple clinical assessments for opportunistic infections among HIV infected patients in a community setting.

Specific Aim 2: To perform a randomized controlled trial to determine the impact of Community Care Coordinators (CCC) on patient adherence (to drugs and to clinic visits), clinical outcomes (i.e. viral load responses, inter-current opportunistic infections, hospitalization, drop out, change to second line therapy and mortality) and patient perception of stigma.

• Hypothesis 2a: Adherence will be better in the CCC group than in the control group.
• Hypothesis 2b: Clinical outcomes will be better in the CCC group than in the control group.
• Hypothesis 2c: Patient perceived stigma will not be greater in the CCC group than in controls.

This study will provide invaluable data on the use of non-clinician care extenders for providing HIV care in resource poor settings. If this system is found to be feasible, successful and acceptable, the CCCs will be extended to the other eight AMPATH rural HIV clinics in western Kenya. Knowledge gained from this study will assist in developing a model for non-clinician centered HIV care systems elsewhere in sub-Saharan Africa.