Retaining HIV-infected patients in care is critical to a successful response to HIV/AIDS in Africa, but loss to follow-up after enrollment often reaches 20%-40% by two years, placing millions of patients at risk of poor outcomes. A strategy to optimize retention within resource constraints is urgently needed, but must first overcome two critical challenges. First, lapses in retention may be due to psychological (e.g., stigma), structural (e.g., transportation) or systems (e.g., long waiting times) barriers Therefore, althoug three interventions have shown some efficacy for retention in randomized trials (SMS text messages, transport vouchers and peer navigators), each acts only on some barriers and is therefore is ineffectual for patients facing other barriers and is of limited overall effectivenessin real-world patient populations. Second, although poor retention is a crisis in Africa, most patients remain in care with minimal support. As a result, even the most effective of these interventions if applied uniformly as a one-size-fits-all approach will squander resources on patients who do not need help while helping only some patients in need. Sequential adaptive strategies - a novel class of public health approaches - may offer a solution to these challenges and simultaneously optimize both the effectiveness and efficiency of retention efforts. A candidate sequential adaptive strategy would start with a less expensive intervention (e.g., SMS) in all patients and then apply a more costly and intensive one (e.g., navigator) only to patients who show early signs of poor retention. Altering the initial intervention in response to a individual's behavior minimizes expenditures for patients for whom the initial intervention is sufficient (optimizing efficiency), but intensifies services for those who need additional or alternative help (optimizing effectiveness). We propose a sequential multiple assignment randomized trial to evaluate a family of such strategies. We will randomize 2,500 adults at six HIV clinics in Nyanza, Kenya to (1) standard of care routine education and counseling (REC), (2) SMS text messages, or (3) transport vouchers. Patients with early signs of weakening retention (defined as the first time a patient is 14 days late for an appointment) will be re-randomized to (1) a single episode of outreach (standard of care), (2) SMS combined with vouchers, or (3) a peer navigator.
In Aim 1, we assess the comparative effectiveness of first-stage strategies (REC, SMS, voucher) to prevent lapses in retention.
In Aim 2, we assess the comparative effectiveness of second stage strategies (outreach, SMS + voucher, navigator) to re-engage patients. Because the initial intervention changes both the numbers and types of patients who lapse and thereby the effect of any second intervention, in Aim 3 we assess the joint effectiveness and cost-effectiveness of sequenced prevention and re-engagement strategies. At study conclusion, our primary output will be a menu of adaptive strategies for retention, each accompanied by estimates of cost and effectiveness, which policy makers in different settings can use to advance the impact of HIV care and treatment programs in Africa.

Public Health Relevance

Retaining HIV-infected patients in care is critical for the long term success of care and treatment for HIV in Africa. Although to date researchers have identified three interventions that help patients stay in care (SMS text messages to remind patients of visits, transportation vouchers to cover costs of clinic attendance, and peer navigators to engage and support patients), the barriers faced by patients are diverse and therefore no single intervention works for all patients all the time. We apply a novel experimental design - sequential multiple assignment randomized trial (SMART) - to find the best sequential application of these interventions in the real world to maximize the total number of patients helped at the lowest cost.

National Institute of Health (NIH)
National Institute of Mental Health (NIMH)
Research Project (R01)
Project #
Application #
Study Section
Behavioral and Social Consequences of HIV/AIDS Study Section (BSCH)
Program Officer
Gordon, Christopher M
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
Odeny, Thomas A; Petersen, Maya; Muga, Charles T et al. (2017) Rapid sociometric mapping of community health workers to identify opinion leaders using an SMS platform: a short report. Implement Sci 12:80