Suburban Blues: Urbanicity, healthcare service mix and spatial patterns of admissions for depression

OBJECTIVES: Medical practice variation is a key target for policy makers aiming to improve the efficiency of healthcare delivery. Studies point to strong regional variation in healthcare utilization in Austria which seems to be a by-product of institutional rules and healthcare service mix rather than variance in epidemiology. METHODS: We use a set of routine healthcare data on hospital admissions for depressive episodes (ICD-10 F320-F329) of adult Austrian patients from 2009 to 2014 on the municipality level to explore spatial patterns of supply-induced demand by nearby hospitals (“Roemer's law”) in hospital admission rates. The dataset covers 99,844 hospital episodes of 69,555 patients in 2,114 municipalities. We estimate a random effects spatial autoregressive model to regress log hospital admission rates on hospital supply and urbanicity as proxies for the municipalities’ healthcare service mix alongside a set of relevant socio-economic/-demographic characteristics. PRELIMINARY RESULTS: Hospital admission rates per 1,000 inhabitants in the population for depressive episodes vary strongly between municipalities (min=0, max=37.7, mean=1.9, SD = 1.8). We find higher admission rates in the population of suburban municipalities (0.10, 95%-CI: 0.58 - 0.13) and municipalities with hospitals (0.07, 95%-CI: 0.01 - 0.1), regardless of hospital type and size. The spatial structure suggests positive spill-over effects on the hospital admission rates between neighbouring municipalities. CONCLUSIONS: Our results suggest that high admission rates are a seemingly suburban phenomenon and that healthcare service mix is a driving factor for admission rates. Though it is not clear whether the regional variation in medical practice reflects under- or oversupply of hospital care, promoting timely access to primary care and early-stage treatments may reduce the burden of avoidable hospitalizations for patients and public budgets, and close a gap of unmet need for care of vulnerable populations. KEY MESSAGES: • Hospital admission rates for depression of adults per 1,000 inhabitants in the population for depressive episodes vary strongly between municipalities. • Hospital supply and urbanicity are major driving factors of the spatial patterns in regional varation of hospital admissions.