Short-term effectiveness of a community health worker intervention for HIV-infected pregnant women in Tanzania to improve treatment adherence and retention in care: A cluster-randomized trial

Abstract

IntroductionCommunity health workers (CHWs) are lay workers who have the potential to enhance services to prevent mother-to-child HIV transmission (PMTCT) and improve the health of women living with HIV infection. We conducted a cluster-randomized trial of an intervention to integrate CHWs with ‘Option B+’ PMTCT services in Shinyanga Region, Tanzania.MethodsThe intervention was implemented for 11 months and included four integrated components: 1) formal linkage of CHWs to health facilities; 2) CHW-led antiretroviral therapy (ART) adherence counseling; 3) loss to follow-up tracing by CHWs; and 4) distribution of Action Birth Cards (ABCs), a birth planning tool. We cluster-randomized 32 facilities offering PMTCT services, within strata of size, to the intervention (n = 15) or comparison (standard of care, n = 17) groups. Intervention effectiveness was determined with a difference-in-differences strategy based on clinical and pharmacy data from HIV-infected postpartum women at baseline (births in 2014) and endline (births April-Oct 2015). The primary outcome was retention in care between 60 and 120 days postpartum. Secondary outcomes included ART initiation, timing of ART initiation (as measured by week of gestation), and ART adherence 90 days postpartum, measured using the medication possession ratio (MPR≥95%).ResultsIntervention and comparison facilities were similar at baseline. Data were collected from 1,152 and 678 mother-infant pairs at baseline and endline, respectively. There were no significant differences in retention in care, ART initiation, or timing of ART initiation between the intervention and control groups. Adherence (MPR≥95%) at 90 days postpartum was 11.3 percentage points higher in the intervention group in ITT analyses (95% CI: -0.7, 23.3, p = 0.06), though this effect was attenuated after adjusting for baseline imbalance (9.5 percentage points, 95% CI: -2.9, 22.0, p = 0.13). Among only sites that had the greatest fidelity to the intervention, however, we found a stronger effect on adherence (13.6 percentage points, 95% CI: 2.5, 24.6).ConclusionsDespite being feasible and acceptable, the CHW-based intervention did not have strong effects on most PMTCT indicators. CHW involvement in PMTCT programs may improve ART adherence among HIV-infected postpartum women, however, and success appears heavily dependent on program implementation.Trial registrationRegistry for International Development Impact Evaluations (RIDIE, ID 552553838b402) and ClinicalTrials.gov (NCT03058484)

A meta-analysis of effectiveness of interventions to improve adherence in pregnant women receiving antiretroviral therapy in sub-Saharan Africa

“…In the first study, in addition to standard care provided for women in the health facility, community health workers (lay health workers) supported patients to overcome barriers to ART adherence. Using a defaulter's tracking tool, community health workers in intervention facilities worked with healthcare workers to identify and locate pregnant and postpartum HIV-positive women in catchment areas who had defaulted on their antenatal or HIV care, with the goal of re-linking them back to care and treatment (Nance et al, 2017). In the second study, in addition to standard care, the intervention involved peer mentoring services provided by women through 'mentor mothers' who were also HIV positive (Richter et al, 2014).…”

“…The measures of adherence used in the 19 studies were selfreporting of medication utilization (n = 11) (Aluisio et al, 2011;Turan et al, 2015;Kim et al, 2015;Stinson et al, 2013;Dillabaugh et al, 2012;Weiss et al, 2013;Herlihy et al, 2015;Nance et al, 2017;Richter et al, 2014;Le Roux et al, 2013;Finocchario-Kesslera et al, 2014), pill counts (n = 2) (Mepham et al, 2011;Yotebieng et al, 2016), pharmacy refill/dispensary data (n = 1) (Kirsten et al, 2011), documentation of adherence in labor and delivery prevention of MTCT of HIV register (n = 1) (Kalembo et al, 2013), liquid chromatographic assay for nevirapine of cord blood (n = 1) (Kieffer et al, 2011), and the use of medication event monitoring systems (MEMS) (n = 1) (Matthews et al, 2016). Among the 19 studies, one used a composite (multi-method) measure to assess adherence comprising a pill count, drug calendar and self-report (Okonji et al, 2012).…”

“…As shown in the meta-analysis, some investigators sought to address methodological concerns by using cluster randomization (Turan et al, 2015;Nance et al, 2017;Richter et al, 2014;Le Roux et al, 2013). This approach is valuable in terms of removing the potential for contamination of an intervention, which can occur if randomization takes place at the patient level rather than at the clinic level.…”

A Cascade of Interventions to Promote Adherence to Antiretroviral Therapy in African Countries

“…In Tanzania, community health workers (CHWs) supported linkage to healthcare facilities, provided adherence counselling, traced those LTFU and distributed a birth planning and appointment reminder tool. A non-significant 8% point increase was seen in ART initiation compared to controls [ 40 ]. Community-based ‘Expert Mothers’ (women living with HIV who recently underwent PMTCT and were on ART) providing 1:1 support, support groups and follow-up for missed appointment in Malawi resulted in higher ART initiation compared with facility-based models (86% versus 81%), although the difference was not statistically significant; retention at 12 months was also higher [ 41 ].…”

Missed Prevention of Mother-to-Child Transmission of HIV (PMTCT) Visits and Associated Programmatic Predictors: A Pilot Study

“…For instance, use of community health workers, counselling in the community and HIV disclosure [14] [15]. This may be possible because the interventions tended to give women in PMTCT a passive role in the programme.…”