Background Information on the burden of drowning in Africa is limited. Existing estimates of national drowning death rates are based on small regional studies and exclude intentional, water transport, and flood-related drowning. We established an approach to collect nationally representative data on fatal and non-fatal drowning in Ghana.

Methods To include the diversity of exposure to water bodies in Ghana, all 261 districts were categorized into three strata: coastal (districts with ocean coastline), inland water (districts that border or contain major lakes or rivers), or dry (districts that do not have a coastline or major lake/river but may contain small lakes/rivers, dams, wells, etc.). The three strata were sorted by urbanicity, ecological zone, and region. Districts were randomly selected from each strata using proportional to population size sampling. Fifty-two districts were randomly selected covering approximately 28% of the population. Trained data collectors identified all drowning cases in each district by contacting community health workers and gatekeepers, enabling a complete census of the included districts without visiting every household.

Results This approach resulted in increased case identification and more complete circumstance information than could be obtained from existing administrative sources, enabling the calculation of national estimates for drowning in Ghana.

Conclusion Unique approaches can be used to collect nationally representative drowning data in settings with limited existing data. Such data are critical for the prioritization of resources.

Learning outcomes Learn new approach for calculating national drowning estimates in low-resource settings. Identify characteristics of drowning in Ghana and implications for prevention.