Over 30% of identified ART-eligible patients fail to initiate (FTI) ART resulting in increased rates of early mortality, MTCT and AIDS illnesses. Failures in ART initiation are predominantly based in the systems conditions in resource limited settings: 1) delays in obtaining CD4 cell counts for ART eligibility, 2) provider misconceptions of the urgency of ART initiation among specific patient groups, and 3) multiple pre-ART adherence counseling sessions and support requirements. We have developed a multi-component Streamlined ART Start Strategy (START) based on an empirically validated model of change that deploys 1) portable point of care (POC) CD4 testing and adapted counseling to enable ART start, 2) dissemination of information through education that predisposes provider behavior and 3) feedback reporting on FTIs that reinforces uptake of ART. We now propose to test this intervention in a randomized, controlled stepped-wedge trial of 24 clinics in Uganda through our PEPFAR supported Mulago-Mbarara Joint AIDS Program (MMJAP).
Aim 1 : Evaluate the effect of START on ART initiation. Specifically, we will compare the overall time to and completeness of ART initiation among those randomized to immediate and delayed implementation of START. We will conduct subgroup analyses in patients with TB and a CD4 <50/?l, all WHO Stage 4 patients and - given Uganda's recent adoption full ART as a strategy for pMTCT - pregnant women with a CD4 <350/?l. In addition, we plan to measure and evaluate the specific sub-intervals (e.g. between patient enrollment, request for CD4 testing, procurement of the specimen, etc.) that comprise the """"""""anatomy"""""""" of the ART initiation process so that specific efficiencies and bottlenecks in the intervention can be identified.
Aim 2 : Evaluate the effect of START on mortality and other outcomes using supplemental outcome ascertainment through tracking patients who are lost to follow up in the community. Understanding the causal effect of a new implementation strategy on """"""""hard"""""""" endpoints such as survival in real-world settings is a key objective of impact evaluation. Under program settings in Africa, however, high loss to follow-up (i.e. unknown outcomes) leads to biased assessment of outcomes. Our group has developed an approach to manage loss to follow-up based on active ascertainment of outcomes through patient tracking in the community. We will adapt this method to evaluate HIV RNA suppression as well as mortality.
Aim 3 : Assess the cost and cost-effectiveness of START. Streamlined ART initiation may not only save lives but also increase efficiency, by decreasing resources per patient and increasing yield to ART. Using effectiveness obtained in Aims 1 and 2 and clinic and individual level costing data, we will estimate the cost and cost-effectiveness of START versus standard of care for ART initiation. Outcomes of interest will include cost per: confirmation of meeting ART criteria;ART initiation;virologic suppression;averted death;averted vertical infection;and averted Disability-Adjusted Life Year (DALY).

Public Health Relevance

ART initiation in routine care in Africa is often suboptimal: up to 20-30% of eligible patients fail to initiate (FTI) ART and those who do start frequently encounter delays of 2-3 months resulting in increased mortality. Given that medication adherence in Africa is in general excellent, identifying strategies to streamline ART initiation is therefore an urgent public health priority. We now propose to implement and test the Streamlined ART Start """"""""START"""""""" strategy to reduce FTI in a randomized controlled, stepped-wedge trial of 24 PEPFAR supported clinics in Uganda.

National Institute of Health (NIH)
National Institute of Allergy and Infectious Diseases (NIAID)
Research Project--Cooperative Agreements (U01)
Project #
Application #
Study Section
Special Emphasis Panel (ZAI1-BP-A (J2))
Program Officer
Bacon, Melanie C
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of California San Francisco
Internal Medicine/Medicine
Schools of Medicine
San Francisco
United States
Zip Code
Kwarisiima, Dalsone; Balzer, Laura; Heller, David et al. (2016) Population-Based Assessment of Hypertension Epidemiology and Risk Factors among HIV-Positive and General Populations in Rural Uganda. PLoS One 11:e0156309
Balzer, Laura B; Petersen, Maya L; van der Laan, Mark J et al. (2016) Targeted estimation and inference for the sample average treatment effect in trials with and without pair-matching. Stat Med 35:3717-32
Brown, Lillian B; Havlir, Diane V; Ayieko, James et al. (2016) High levels of retention in care with streamlined care and universal test and treat in East Africa. AIDS 30:2855-2864
Chamie, Gabriel; Clark, Tamara D; Kabami, Jane et al. (2016) A hybrid mobile approach for population-wide HIV testing in rural east Africa: an observational study. Lancet HIV 3:e111-9
Balzer, Laura B; van der Laan, Mark J; Petersen, Maya L et al. (2016) Adaptive pre-specification in randomized trials with and without pair-matching. Stat Med 35:4528-4545
Chang, Wei; Chamie, Gabriel; Mwai, Daniel et al. (2016) Cost and efficiency of a hybrid mobile multi-disease testing approach with high HIV testing coverage in East Africa. J Acquir Immune Defic Syndr :
Chang, Wei; Chamie, Gabriel; Mwai, Daniel et al. (2016) Implementation and Operational Research: Cost and Efficiency of a Hybrid Mobile Multidisease Testing Approach With High HIV Testing Coverage in East Africa. J Acquir Immune Defic Syndr 73:e39-e45
Katrak, Shereen; Day, Nathan; Ssemmondo, Emmanuel et al. (2016) Community-wide Prevalence of Malaria Parasitemia in HIV-Infected and Uninfected Populations in a High-Transmission Setting in Uganda. J Infect Dis 213:1971-8
Kadede, Kevin; Ruel, Theodore; Kabami, Jane et al. (2016) Increased adolescent HIV testing with a hybrid mobile strategy in Uganda and Kenya. AIDS 30:2121-6
Balzer, Laura B; Petersen, Maya L; van der Laan, Mark J et al. (2015) Adaptive pair-matching in randomized trials with unbiased and efficient effect estimation. Stat Med 34:999-1011

Showing the most recent 10 out of 20 publications