Spatial proximity and childhood hospital admissions in a densely populated conurbation: Evidence from Hong Kong's 'Children of 1997' birth cohort

Objectives: Physical distance affects hospital use. In a densely populated city in China, we examined if child public hospital use was associated with individual-level proximity, and any differences by admission type or geo-spatially. Methods: We used negative binomial regression in a large, population-representative birth cohort to examine the adjusted associations of proximity to emergency facilities (A&E) with hospital admissions, bed-days and length of stay from 8 days to 8 years of age. We used geographically weighted regression to assess geo-spatial variation. Results: Proximity was positively associated with emergency admissions (incidence rate ratio (IRR) 1.21, 95% confidence interval (CI) 1.10 to 1.34 for <1. km compared to ≥2. km) and bed-days but not with length of stay, adjusted for parental education and mother's birthplace. There was no such association for other admissions (IRR 1.03, 95% CI 0.84 to 1.26). There was little geo-spatial variation. Conclusions: Proximity was associated with emergency admissions. Given the societal costs of such use and the risks of iatrogenesis, attention should focus on achieving a more effective use of scarce resources.