Background
Identifying individuals who require particular strategies or possess characteristics for which the interventions are most effective is critical for achieving personalized care and optimal impact. We sought to identify and assess moderators of a home-based couples intervention on uptake of couple HIV testing and counselling (CHTC) in southwestern Kenya.
Methods
In a randomized control trial, we enrolled 800 stable couples (2/3 of the women living with HIV, WLWH) between 2019-2023. Consenting couples were randomized to a home-based couples intervention (HV, n=267), HIV self-test kits for couples (HIVST, n=266), or standard care (SC, n=267). In the HV arm, pairs of lay health workers (one female, one male) conducted two home visits during pregnancy and up to 3 visits postpartum, providing health education, couple communication skills building, and offering CHTC. Surveys with both partners occurred at baseline, 3- and 12-months postpartum, and brief phone assessments at 6- and 18-months postpartum, collecting data on socio-demographics, healthcare utilization, couple dynamics, and HIV testing outcomes. A log-binomial generalized estimating equation model was used to assess the potential moderating effect of baseline measures; including previous couple testing, education, wealth, couple age disparity, stigma perceptions, couple communication, relationship satisfaction, woman’s HIV status, parity, and length of the relationship; on the effect of the intervention on CHTC uptake. Marginal probabilities and effects with plots were reported.
Results
Significant moderators of the intervention effect on CHTC were the HIV status of a woman at baseline (interaction p-value=0.041) and parity (p-values=0.018). The overall probability of CHTC was higher among HIV-negative women compared to WLWH, with marginal effects of 11% in HIVST, 27% in HV, and 12% in the SC arm (Figure 1). Results also showed a significantly lower CHTC probability among women with numerous prior live births, as compared to those with few live births (Figure 2).
Conclusions
The home-based intervention appeared to have a more beneficial impact on CHTC uptake in couples where the woman was HIV-negative at baseline, compared to couples where the woman was HIV-positive. In addition, the intervention had more impact on CHTC uptake in couples with lower parity. Future intervention roll-out aiming for higher CHTC might target HIV-negative women with lower parity.