Forecasting indoor air temperatures in tropical hospitals to improve patient health and preparedness for extreme heat

Extreme heat is a growing public health threat in tropical Africa, especially in Ghana where outdoor maximum daily temperatures (Tx) regularly exceed 35 °C. Vulnerability to heat is socially and economically differentiated, with pregnant women, young children, elderly people, and people with chronic health conditions most at risk. High proportions of these groups are found in hospitals, yet little is known about extreme temperatures inside wards. This paper addresses these gaps by developing proof-of-concept forecasting models for indoor air temperatures in Ghanaian hospital wards. A significant contribution of this work is the novel, high-resolution, empirical indoor environment dataset collected from four hospitals capturing a variety of building types in two distinct climate zones. This dataset enabled the development and testing of heat-health early warning systems which significantly contributes to a scalable, generalisable framework for climate adaptation in healthcare buildings across the tropical Global South. Measured indoor Tx were statistically related to numerical weather predictions of outdoor temperatures to create models capable of explaining up to 75% of the variance in daily Tx with accuracies of ∼1 °C at 24 h forecasting times. Such heat alerts could be used by hospital authorities to adjust staffing levels for expected surges in patient volumes, and for rescheduling of elective non-urgent outpatient treatments to free up capacity during heatwaves. We assert, however, that heat-health early warning systems should be integrated alongside other measures to adapt to rising indoor temperatures, such as improving hospital design, formulating heat-health action plans, and raising awareness of heat-related health issues.