Mobile Heat Early Warning System for Pregnant and Postpartum Women in South Africa: Prospective Observational...

Early warning systems (EWS) have the potential to reduce heat-related health risks, yet few are designed specifically for vulnerable populations. Pregnant women and infants, often cared for by postpartum women, are particularly susceptible to heat-related illnesses. The MotherHeat Alert mobile app is a heat EWS specifically developed for pregnant and postpartum women.

This study aimed to assess the usability and effectiveness of the MotherHeat Alert in the Tshwane District, South Africa.

This prospective observational evaluation study included a cohort of 201 participants (100 pregnant and 101 postpartum) who were followed over a 6-month period. In addition, a questionnaire was administered at 3 time points (baseline, midpoint, and end point) to collect self-reported information on health, living situation, and heat exposure, among other parameters. Usability was evaluated in the end point questionnaire through a validated Post-Study System Usability Questionnaire (PSSUQ), using descriptive statistics. Effectiveness was assessed through changes in self-reported heat-protective behaviors between baseline, midpoint, and end point, as well as through end point assessments of heat-health knowledge increase and self-reported behavior change attributable to the intervention. Generalized estimating equations were used to analyze repeated measures of heat-protective behaviors, while descriptive statistics were used to summarize end point-reported knowledge gains and behavior change.

Usability scores indicated high overall usability, with mean PSSUQ scores of 1.76 (SD 0.60) among pregnant women and 1.37 (SD 0.45) among postpartum women. Similar patterns were observed across most PSSUQ subscales, including system usefulness, interface quality, and information quality. Most self-reported behavior change due to alerts received through the app was rated as large to very large by the pregnant women (47/70, 67%) and postpartum women (68/82, 83%). Nevertheless, comparing behavior change at baseline and end point mainly highlighted a decrease in protective behavior in both pregnant and postpartum cohorts. Self-reported knowledge increase among pregnant and postpartum women was reported as large to very large following the use of the EWS (55/70, 79% and 72/82, 88%, respectively). Importantly, however, attrition was substantial, particularly among pregnant participants, reaching 30% (30/100) in the pregnant cohort and 19% (19/101) in the postpartum cohort by end point.

The MotherHeat Alert app demonstrated positive outcomes in usability. Effectiveness showed conflicting results, with positive self-reports on behavior change and knowledge increase, yet a decrease in protective heat behaviors when comparing baseline and midpoint with end point results, likely due to seasonality effects. Nevertheless, these findings highlight the potential of tailored EWS to support heat adaptation among pregnant and postpartum populations. Successful implementation and scalability require careful consideration of contextual factors, further assessment of long-term usability and integration into existing health infrastructure.