University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

Home > Global Health Literature Digest > Improved retention
Improved retention associated with community-based accompaniment for antiretroviral therapy delivery in rural Rwanda
Global Health Sciences Literature Digest
Published May 13, 2013
Journal Article

Franke MF, Kaigamba F, Socci AR, Hakizamungu M, Patel A, Bagiruwigize E, Niyigena P, Walker KD, Epino H, Binagwaho A, Mukherjee J, Farmer PE, Rich ML. Improved retention associated with community-based accompaniment for antiretroviral therapy delivery in rural Rwanda. Clin Infect Dis. 2013 May; 56(9):1319-26.

Objective

To determine whether the addition of community-based accompaniment to Rwanda's national model for antiretroviral therapy (ART) delivery would improve outcomes for HIV-infected patients on ART.

Setting

Two rural regions of Rwanda.

Study Design

Prospective observational double cohort study.

Population

HIV-infected adults initiating ART.

Intervention (Predictor Variable)

Addition of community-based accompaniment and a range of other community-based support services to standard care.

Main Outcome Measures

Attrition from treatment during the first year of ART; retention in care with viral load (VL) suppression at one year; and absolute change in CD4 cell count at one year. Attrition was defined as death, loss to follow-up, or default. An individual was defined as lost to follow-up if he or she had not returned to the clinic for at least 60 consecutive days. An individual was defined as having defaulted if he or she had stopped treatment for at least 60 consecutive days. Death was defined as death from any cause after program enrollment. Viral load suppression was defined at a threshold of <200 copies/mL.

Methods

The study was conducted in the Kayonza and Kirehe districts of southeastern Rwanda and in the Musanze district of northern Rwanda. In both study regions, ART was provided first through a district hospital and then decentralized to smaller satellite clinics. Each district hospital has a laboratory and is located 2 hours by car from the capital city of Kigali. Investigators enrolled patients for the community-based accompaniment model from five health centers in Kayonza and Kirehe. They enrolled patients for the clinic-based model from four health centers in Musanze.

Participants were eligible if they were residents of Kayonza, Kirehe or Musanze districts, planned to maintain residence there for at least one year, and had not previously initiated ART. Investigators consecutively enrolled consenting HIV-infected, treatment-naive adults (≥21 years), initiating ART with CD4 counts <350 cells/µL.

According to Rwanda's standard of care, first-line ART regimens were stavudine (d4T) or zidovudine (AZT), lamivudine (3TC), and nevirapine (NVP). Efavirenz (EFV) replaced NVP in patients receiving treatment for tuberculosis (TB). Cotrimoxazole was prescribed if patients had CD4 count <350 cells/µL or stage ≥3 World Health Organization (WHO) HIV disease. CD4 counts were routinely measured prior to ART initiation and every six months thereafter. Patients were encouraged to disclose their HIV status to family members or friends and to identify a "treatment buddy."

Participants enrolled in the clinic-based model were those residing in Musanze. They received treatment according to Rwanda's national standard of care. ART was provided without cost.

Participants enrolled in the community-based accompaniment model were those residing in Kayonza or Kirehe. They received treatment according to Rwanda's national standard of care. In addition to this, they received community-based accompaniment and a range of other community-based support services. They were visited daily in their homes by a community health worker (CHW) who provided social support, monitored for adverse events, identified potential barriers to adherence, and directly observed ingestion of all medications at least once per day. Patients received monthly food rations (based on a family of four) for their first 10 months on ART. They received transportation stipends for routine clinic visits. After regularly-scheduled visits, patients could pick up a food package. For the first four monthly clinic visits, CHWs accompanied patients, and after this accompanied them as needed. Social workers evaluated individual patients' needs for additional support (e.g. payment of school or health insurance fees, microfinance activities, employment assistance, or home repairs) and arranged for it if needed. CHWs completed an initial multi-day training on accompaniment for HIV care, and then participated in monthly refresher trainings and discussion sessions.

Results

Three hundred six participants were enrolled in the clinic-based cohort and 304 were enrolled in the community-based accompaniment cohort. In terms of demographics the cohorts were broadly similar, but patients in the clinic-based cohort were sicker, with more advanced WHO stage HIV disease and greater immune suppression. Patients in the community-based accompaniment cohort were generally worse off in terms of psychosocial, access-to-care, and socioeconomic variables.

More patients died in the clinic-based cohort (n=22, 7.2%) compared to the community-based accompaniment cohort (n=13, 4.3%). Ten (3.3%) patients in the clinic-based cohort defaulted or were lost to follow-up before one year, compared to three (1%) in the community-based accompaniment cohort. Seven (2.3%) patients in the clinic-based cohort and eight (2.6%) in the community-based accompaniment cohort transferred out of the program before one year.

Community-based accompaniment in addition to standard care was associated with lower attrition rates and a higher probability of retention with a suppressed VL at one year. After adjustment for CD4 count, depression, physical health quality of life sub-score, and food insecurity score, community-based accompaniment was strongly associated with a lower hazard of attrition from care during the first year of ART (hazard ratio [HR] 0.17, 95% confidence interval [CI] .09 to .35, p<.0001). The association was also strong without adjustment (HR 0.49, 95% CI .27 to .89, p=.02).

Among the 530 patients who were still in care and had a VL measurement at 12 months, 43 (8.1%) had a VL of >200 copies/mL (clinic-based cohort, 23/43, 8.9%; community-based accompaniment cohort, 20/43, 7.4%). After adjusting for baseline CD4 count, depression, social support score, physical health sub-score, and travel time to the clinic, patients receiving community-based accompaniment were more likely to be retained in care with a suppressed viral load (relative risk [RR] 1.15, 95% CI 1.03 to 1.27, p=.01).

After excluding data from one patient with an "extreme outlying value for CD4 count change," 12-month CD4 count data were available for 527 patients. In the clinic-based care cohort, CD4 counts increased by a median of 141 cells/µL (95% CI, interquartile range [IQR] 79 to 211 cells/µL). The median CD4 count increase in the community-based accompaniment cohort was 188 cells/µL (95% CI, IQR 86 to 293 cells/µL).

In univariable analysis, community-based accompaniment was associated with a greater increase in CD4 count at one year (difference in CD4 count change +40.1 cells/µL, 95% CI +14.7 to +66.5, p=.002). This association was not significant after adjustment for covariates (21.7 cells/µL, 95% CI -16.9 to +60.3, p=.27).

Table: Association between community-based accompaniment and selected antiretroviral treatment outcomes

Table: Association between community-based accompaniment and selected antiretroviral treatment outcomes (from the article)

Conclusions

The authors conclude that community-based accompaniment is a successful model for ART delivery and suggest that it may improve treatment outcomes when added to a clinic-based program lacking in patient support mechanisms.

Risk of Bias

As a non-randomized observational double cohort study, the risk of bias in this study is moderate. The intervention was compared to standard care and investigators used statistical methods to adjust for potentially significant baseline health, psychosocial, and socioeconomic differences between the two populations. LTFU was accounted for appropriately.

In Context

This study's overall finding that community-based accompaniment was associated with a higher probability of retention rate with viral load suppression at one year is consistent with those of a similar program in Peru,(1) as well as with other studies(2, 3, 4) showing the benefit of community-based adherence and psychosocial support for patients on ART.

Programmatic Implications

Health ministries, district health authorities and community-based organizations may wish to consider implementing pilot programs of community-based accompaniment for patients on ART.

References

  1. Muñoz M, Finnegan K, Zeladita J, et al. Community-based DOT-HAART accompaniment in an urban resource-poor setting. AIDS Behav 2010; 14:721-30.
  2. Fatti G, Meintjes G, Shea J, Eley B, Grimwood A. Improved surviva and antiretroviral treatment outcomes in adults receiving community-based adherence support: five-year results from a multicentre cohort study in South Africa. J Acquir Immune Defic Syndr 2012; 61:e50-8.
  3. Grimwood A, Fatti G, Mothibi E, Malahlela M, Shea J, Eley B. Community adherence support improves programme retention in children on antiretroviral treatment: a multicentre cohort study in South Africa. J Int AIDS Soc 2012; 15:173-81.
  4. Pearson CR, Micek MA, Simoni JM, et al. Randomized control trial of peer-delivered, modified directly observed therapy for HAART in Mozambique. J Acquir Immune Defic Syndr 2007; 46:238-44