Geographical disparities in intentional self-harm hospitalization risks in Florida: a retrospective ecological study.

Abstract

Background: Over the last decade, intentional self-harm hospitalization risks have increased in the United States (US). Although there is evidence of geographic disparities in intentional self-harm hospitalization risks, with some areas having significantly higher risks, little is known about these disparities in Florida. Identifying these disparities is useful for guiding prevention programs and tailoring efforts to mitigate the problem. Therefore, the objective of this study was to investigate geographic disparities of intentional self-harm hospitalization risks in Florida.

Methods: Retrospective county-level intentional self-harm hospitalization data for the years 2016-2023 were obtained from Florida Health Charts, a publicly available data dashboard maintained by the Florida Department of Health. County-level age-standardized and spatial empirical Bayesian (SEB) smoothed risks were calculated for four time periods: 2016-2017, 2018-2019, 2020-2021, and 2022-2023. Significantly high-risk clusters were identified using Tango's flexible spatial scan statistics. Two-sample tests of proportions were used to compare proportions of counties and populations in high-risk clusters between rural and urban areas.

Results: A total of 66,308 intentional self-harm hospitalizations were reported during the study period. Statewide SEB smoothed risks varied geographically, ranging from 25.3 to 1,683.1 hospitalizations per 100,000 persons. The highest risks tended to occur in the northwest, northeast, central, and southwest Florida. A total of four statistically significant high-risk clusters were identified in 2016-2017 (relative risks (RR) = 1.44-15.85), six in 2018-2019 (RR = 1.22-12.79), six in 2020-2021 (RR = 1.21-7.25), and eight in 2022-2023 (RR = 1.20-3.69). The cluster with the highest risks was observed in a single rural northeastern county (Union). Additional high-risk clusters were observed in predominantly urban counties. The proportion of Florida counties included in significant high-risk clusters ranged from 22.4% in 2016-2017 to 38.8% in 2022-2023. Urban counties were significantly more likely than rural counties to be part of high-risk clusters in 2016-2017 (p = 0.002) and 2020-2021 (p = 0.019), but not in 2018-2019 (p = 0.11) or 2022-2023 (p = 0.477). The highest proportion of the population in high-risk counties alternated between urban and rural areas, with significantly (p < 0.0001) higher urban proportions in 2016-2017 and 2020-2021, and significantly (p < 0.0001) higher rural proportions in 2018-2019 and 2022-2023.

Conclusion: Geographic disparities in intentional self-harm hospitalization risks exist in Florida, with high-risk spikes observed in rural counties and widespread risks in urban counties. Study findings are useful for guiding targeted intentional self-harm prevention strategies and needs-based resource allocation geared at reducing/eliminating disparities and improving population health for all.