Effect of a digital training package on clinical outcomes in Malawi's index case testing programme: a cluster...

Training health workers is a common implementation strategy to expand evidence-based interventions. Digital training is a promising way of reaching more health workers, minimising clinical interruptions and lowering costs, but clinical outcomes are rarely evaluated in low-income settings. Clinical outcomes were evaluated in Malawi's HIV index case testing programme through the Package of Resources for Assisted Contact Tracing: Implementation, Costs, and Effectiveness study.

An unblinded cluster randomised controlled trial was conducted in 33 health facilities in two Malawian districts (2022-2023). Facilities were stratified by district and size and randomly assigned 2:1 to receive standard training or standard training plus a digital-based package. The package included asynchronous role-modelling, small-group practice, one-on-one feedback and tablet-guided quality improvement. Participants were lay health workers involved in index case testing. Index clients were people diagnosed with HIV. Contact clients were their partners, children and household members. Five coprimary outcomes, abstracted from programme registers over 1 year, were assessed at the cluster level: index client participation, contacts elicited, contacts tested, new HIV diagnoses and self-test kit provision. Impacts were estimated using negative binomial mixed-effects models (α=0.05).

Clusters were randomly assigned to enhanced (n=11) or standard (n=22) arms and analysed in four 3-month increments (calendar-quarters) over a 1-year period. In each calendar quarter, clusters had a median of 281 potential index clients (IQR 220-427). Significant effects were observed for three primary outcomes: contact client elicitation (RR=1.37, 95% CI 1.10 to 1.71, p=0.006), contact client testing (RR=1.45, CI 1.10 to 1.92, p=0.01) and self-test kit provision (RR=2.29, 95% CI 1.19 to 4.40, p=0.01). Positive, but non-significant effects were observed for index client participation (RR=1.22, CI 0.93 to 1.60, p=0.1) and new HIV diagnoses (RR: 1.28, CI 0.94 to 1.76, p=0.1). No study-related adverse events occurred.

Enhanced digital training positively impacted meaningful clinical outcomes and could be replicated for expansion to other evidence-based interventions.

NCT05343390.