Infectious Disease · Medical Literature
Impact of motivational interviewing training for lay counsellors on HIV care outcomes in South Africa: a cluster-randomized pilot trial.
Abstract
Retention in HIV care remains challenging in sub-Saharan Africa. We evaluated whether training lay counsellors in Motivational Interviewing (MI) could improve outcomes among newly diagnosed people living with HIV (PLHIV) in South Africa. We conducted a cluster-randomized pilot trial in eight Johannesburg primary healthcare clinics, randomized 1:1 to Thusa-Thuso intervention (7-day MI training plus 12-month mentorship) or standard-of-care (SOC). Adults PLHIV (≥18 years) were enrolled and followed for 12 months. The primary outcome was 12-month retention. The secondary outcome was HIV viral suppression (VL <50 copies/mL). We assessed dose-response associations between counsellor MI skills and outcomes. Between March 2020 and August 2021, 548 PLHIV were enrolled (n = 291 intervention; n = 257 SOC). The intervention significantly increased 12-month viral suppression (39.86% vs 28.40%; risk differences (RD): 11.12%; P = .007). Improvements occurred for ART initiation (88.3% vs 74.7%; RD: P = .085) and 12-month retention (51.9% vs 40.5%; P = .067). Patient retention increased with counsellor proficiency in cultivating change talk (RR: 1.47; P = .016), showing empathy (RR: 1.30; P = .020) and partnership (RR: 1.37; P = .010). MI training and mentorship for lay counsellors improved 12-month viral suppression among newly diagnosed PLHIV. Training existing cadres in MI skills is a high-leverage strategy to achieve UNAIDS 95-95-95 targets.
Clinical relevance
Does training lay counsellors in Motivational Interviewing (the Thusa-Thuso program: 7-day training plus 12-month mentorship) improve 12-month retention in HIV care and viral suppression among newly diagnosed adults living with HIV in South African primary healthcare clinics, compared with standard-of-care counselling? __SNAPSHOT__ {"studyDesign":"Cluster-randomized pilot trial (1:1 allocation at the clinic level)","setting":"Eight public primary healthcare clinics in Johannesburg, South Africa; enrollment March 2020 – August 2021","sampleSize":"n = 548 PLHIV (intervention: n = 291; standard-of-care: n = 257) across 4 intervention and 4 control clusters","population":"Newly diagnosed adults living with HIV (≥18 years) enrolled at participating Johannesburg primary healthcare clinics","intervention":"Thusa-Thuso program: 7-day Motivational Interviewing (MI) training plus 12-month mentorship for lay counsellors, implemented prior to patient enrolment","comparator":"Standard-of-care (SOC) HIV counselling at control clinics","primaryOutcome":"12-month retention in HIV care","followUp":"12 months post-enrollment","fundingSource":"USAID (reported by lead author; potential competing interest declared)"}
Key findings
PRIMARY OUTCOME — 12-month retention in HIV care: The intervention arm achieved 51.9% retention vs. 40.5% in SOC (absolute risk difference [RD] not numerically reported in the abstract beyond the directional comparison; p = 0.067). This did not meet conventional thresholds for statistical significance. SECONDARY OUTCOME — 12-month viral suppression (VL <50 copies/mL): The intervention significantly increased viral suppression: 39.86% vs. 28.40% (RD: 11.12 percentage points; p = 0.007). This was the trial's statistically significant finding. NNT for one additional patient achieving viral suppression = approximately 9 (calculated: 1/0.1112 ≈ 9.0). ART INITIATION: Improved in the intervention arm (88.3% vs. 74.7%; p = 0.085); directionally favorable but did not reach statistical significance. DOSE-RESPONSE ANALYSIS — Counsellor MI skill and patient retention: Higher counsellor proficiency in cultivating change talk was associated with greater patient retention (RR: 1.47; p = 0.016). Higher counsellor empathy was also associated with improved retention (RR: 1.30; p-value reported as significant in abstract but exact p not fully transcribed in provided text). Confidence intervals for these dose-response associations were not reported in the provided abstract. Adverse events and safety data: Not reported in the abstract.
Practice implications
This pilot cluster-randomized trial provides the most direct available evidence that a structured, contextually adapted MI training program for lay counsellors — the Thusa-Thuso intervention — can significantly improve HIV viral suppression among newly diagnosed adults in a high-burden South African urban primary care setting. The secondary outcome of viral suppression (the more objective, patient-centered endpoint) reached statistical significance with a clinically meaningful absolute difference of ~11 percentage points (NNT ≈ 9, calculated). The primary endpoint of 12-month retention showed a consistent directional benefit (~11.4 percentage point absolute difference) but did not achieve statistical significance (p = 0.067), which is expected in a pilot trial that was not powered for definitive hypothesis testing. The dose-response relationship between counsellor MI fidelity (change talk cultivation and empathy) and patient retention strengthens the biological and behavioral plausibility of the mechanism. PATIENTS NOT ADEQUATELY REPRESENTED: The trial enrolled only at Johannesburg urban public clinics, limiting generalizability to rural South Africa, other sub-Saharan African countries, or settings where lay counsellors serve different roles. Adolescents (<18 years), key populations (men who have sex with men, sex workers), and patients with comorbidities were not specifically reported as included or excluded. The enrollment window (March 2020–August 2021) overlapped significantly with the COVID-19 pandemic, which likely disrupted both retention and standard care delivery in ways that may not reflect steady-state conditions. IMPORTANT CAVEAT: Only the abstract was available for review; full methods (randomization procedure, allocation concealment, blinding strategy, missing data handling, covariate adjustment) could not be independently assessed. Practice should not change on the basis of this pilot alone; a definitive, adequately powered RCT is warranted. __LIMITATIONS__ Pilot trial design: explicitly not powered for definitive hypothesis testing; the primary outcome (retention) did not achieve statistical significance,Only 8 clusters (4 per arm) severely limits the ability to adjust for cluster-level confounding and reduces statistical power for cluster-level analyses,COVID-19 pandemic overlap (March 2020–August 2021) is a major source of uncontrolled confounding that could differentially affect intervention and control clinics and limit generalizability,USAID funding declared by the lead author introduces potential industry/sponsor bias risk, though the direction of influence on MI training research is unclear,Viral suppression rates in both arms (~40% intervention, ~28% SOC at 12 months) are substantially lower than national targets, suggesting high background attrition that may obscure intervention effects and raise questions about the completeness of viral load ascertainment,Blinding: Not reported in the abstract; blinding of counsellors and patients to intervention assignment is likely not feasible in behavioral trials, introducing performance bias risk,Generalizability limited to urban Johannesburg public primary care; rural South Africa and other sub-Saharan African health systems may differ substantially in counsellor workforce capacity and patient population,Dose-response analyses are observational in nature (non-randomized within the trial) and subject to confounding by counsellor-level characteristics; confidence intervals for dose-response estimates not reported in the abstract,Long-term durability of MI skill acquisition among lay counsellors beyond 12 months is unknown,Full methods, allocation concealment, and data handling procedures not assessable from abstract alone __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ In this cluster-randomized pilot trial, the Thusa-Thuso MI training program for lay counsellors significantly improved 12-month HIV viral suppression by ~11 percentage points (NNT ≈ 9, calculated) among newly diagnosed adults in Johannesburg primary care, though the primary endpoint of retention narrowly missed significance. The dose-response relationship between counsellor MI skill and patient retention strengthens mechanistic plausibility. These promising pilot results require confirmation in a fully powered, definitive RCT — ideally conducted outside a pandemic period — before broader implementation can be recommended.