Geospatial modelling reveals flood-driven inequities in nutrition-sensitive healthcare access across Kampala, Uganda

Background Urban floods slow or sever travel to healthcare facilities, reducing timely access to maternal and child healthcare in Kampala. We estimated flood-related changes in walking access to all public and private-not-for-profit facilities and to hospitals. Methods We modelled walking travel times at ~ 10 m resolution using land cover, roads, hydrography, and elevation, treating water bodies as barriers except at mapped crossings. Speeds reflected adults with young children, with wet-surface penalties and depth rules. We simulated 15 rainfall scenarios, 20–100 mm over 1, 3, or 6 hours, and computed anisotropic travel time to the nearest facility. Outcomes were parish-level, population-weighted changes from baseline, stratified by a maternal, child, and socioeconomic vulnerability index. Results Here we show travel time increased across all scenarios. City-wide means to the nearest facility ranged from 11.2 minutes at 20 mm over 6 hours to 19.1 minutes at 100 mm over 1 hour. For hospitals, means ranged from 23.0 to 55.8 minutes. At fixed rainfall, shorter storms produced larger increases than longer storms with the same depth. Under the 100 mm, 1 hour scenario, variation across vulnerability levels was modest for all facilities, 23.5 versus 21.8 minutes between highest and lowest quartiles, but substantial for hospitals, 97.6 versus 43.1 minutes. Conclusions Floods limit walking access in Kampala, with the largest penalties for hospital care and in high-vulnerability parishes. These delays risk undermining timely antenatal and postnatal care, child illness assessment, and immunisation contacts, routine moments through which supplementation, growth monitoring, and caregiver counselling are delivered.