Implementation evaluation of biometric fingerprint scanning for monitoring HIV care engagement: acceptability, feasibility, and fidelity among peripartum women living with HIV in Lilongwe, Malawi.

Abstract

Background: Sustained HIV care engagement is essential for pregnant women living with HIV (PWLHIV) on lifelong antiretroviral therapy (ART). However, transferring clinics, missed HIV visits, and disengagement from HIV care can make monitoring engagement challenging. This study evaluated the acceptability, feasibility, and fidelity of a biometric fingerprint scanning (BFS) system to monitor women's engagement in HIV care across a network of large, urban clinics in Malawi using a cloud-based BFS system. Differences in BFS coverage compared to the standard of care monitoring (SOC) system were also examined.

Methods: PWLHIV aged ≥18 years, attending their first antenatal visit, and residing in Lilongwe District, Malawi, were enrolled and followed through 9-months postpartum at 5 clinics. An implementation evaluation survey was administered at 9-month postpartum to evaluate BFS acceptability and feasibility. BFS recorded attendance at each study and routine HIV visit, with daily data uploads to facilitate participant identification across sites. SOC attendance data were abstracted for fidelity assessment and comparison of BFS coverage. Descriptive statistics assessed acceptability, feasibility, and fidelity. Using the Conceptual framework of Implementation Fidelity, the composite score for the four fidelity aspects (Coverage, Consistency, Duration, Content) was calculated and categorized as "high" (≥ 80%), "moderate" (50%-79%), and "low adherence" (<50%). Chi-square tests and Poisson regression with generalized estimating equations and robust variance were used to assess differences in BFS coverage, with statistical significance set at p < 0.05.

Results: Of 399 participants, 91% completed a 9-month postpartum visit. Acceptability was high, with 83% comfortable using BFS and 98% finding it easy to use. Feasibility was high, with 99% recommending it. Fidelity was moderate, with a composite score of 68%. BFS was operational for 88% of 612 working days, used consistently at a median of 50% (IQR 30, 70) of HIV visits, and 75% of participants reported using it as trained. BFS captured 52% of 3,554 visits compared to 48% with SOC (PR = 1.06, 95% CI: 1.04-1.09).

Conclusion: BFS was acceptable and feasible for monitoring HIV visits across clinics, showing better coverage compared to the SOC system despite moderate fidelity. Further research is needed to identify barriers and facilitators to BFS implementation.

Supplementary information: The online version contains supplementary material available at 10.1186/s44247-025-00162-8.