Identifying hotspots of seizure-related hospital admissions in Australia

This study aimed to: (1) map geographic trends of seizure-related hospitalizations across Australia, (2) identify hotspots in hospitalization rates, and (3) assess geographic inequities in access to specialized services for seizure disorders.

Standardized seizure admission ratios (SR) were calculated using publicly available hospital admissions data incorporating the diagnoses of epilepsy, status epilepticus and convulsions for the year 2020-21 in Australia. Forward selection was used to ascertain optimal subset of covariates for spatial regression models. Model fitness was evaluated using Deviance Information Criterion and Watanabe-Akaike Information Criterion. Tiered hospital catchment maps and relationships between hospitals were generated based on proximity and available specialized services. A web-based application was created to view results and includes a search function to identify tiers of hospitals for Australian addresses (https://gntem3.shinyapps.io/epilepsyadmissions/).

Although the absolute number of hospitalizations was low, the Northern Territory had three local government areas (LGAs) with the highest SRs (e.g., MacDonnell LGA (SR 5.29, n = 50)). Hotspots were more frequently observed in regional and remote LGAs but were also present in urban areas (e.g., Geelong LGA (SR 1.24)). The bestperforming spatial regression model incorporated kidney disease, cancer, diabetes, mental health conditions, and the number of family physicians per 100,000 people as significant covariates.

Hotspots of seizure-related hospitalizations are often located in areas with limited access to specialized services, underscoring the geographic inequities in care delivery. Addressing these disparities through further modelling of spatial trends and targeted resource allocation is essential for improving equitable healthcare access for seizure disorders.