Does community health reduce hospital utilisation? Evidence from Australia
Melbourne Institute Working Paper No. 12/25
Date: October 2025
Author(s):
Abstract
Background – International evidence on the impact of community health on health outcomes is generally positive; however, the effect on reducing hospitalisations is mixed and often depends on the specific settings and institutions. Evidence from Australia is limited. Objective – This study assesses how community health affect hospital care utilisation under the Victorian Community Health Program (CHP), which provides a range of services to priority populations who are at risk of poorer health under a social model of care. International evidence on the impact of community health on health outcomes is generally positive; however, the effect on reducing hospitalisations is mixed and often depends on the specific settings and institutions. Evidence from Australia has been limited. Data and methods – To account for differences between community health users and non-users, this study used Propensity Score Matching to address the non-random assignment of individuals into CHP. The data were extracted from three administrative data sources: (i) the Community Health Minimum Dataset (CHMDS), (ii) the Victorian Admitted Episodes Dataset (VAED), and (iii) the Victorian Emergency Minimum Dataset (VEMD). The data were linked across datasets using patient identification information. Five outcomes were evaluated in relation to hospital admissions and two outcomes for emergency department presentations. Results – The findings were mixed—on some outcomes, access to CHP was found to reduce healthcare utilisation, but for other outcomes, users exposed to CHP were found to use more healthcare than comparable non-CHP users. Compared with non-users, users of CHP were found to have fewer unplanned hospital readmissions. In contrast, CHP users were found to have a longer length of stay on average, more potentially preventable admissions, more admissions with hospital acquired complications, and more emergency department presentations. No significant difference was found between users of CHP and non-users on the total number of hospital admissions. Further analyses on subgroups of patients by age also produced mixed results that vary by outcome and age. Conclusions – Access to CHP was found to have mixed effects on hospital utilisation, which may reflect previously unmet health needs among users. Further research is needed to examine potential factors that explain the variation in outcomes. A key limitation of the study is the lack of data on primary care use, health improvements allowing hospital avoidance, and changes in well-being—important dimensions when evaluating the impact of CHP. More robust testing and access to richer data are necessary before these results can meaningfully inform policy decisions.
To obtain a copy of this paper, please email the author.
Request Paper