Multi-level Analysis for Geographical Inequalities on Ambulatory Care Sensitive Hospitalizations
Abstract
Ambulatory care sensitive conditions (ACSC) hospitalizations are potentially preventable events and considered as indicator of the efficiency of the primary healthcare system. Therefore, a high level of geographic variation in ACSC hospitalizations warrants more research. The objective of current research was to assess the variation in odds of ACSC-related hospitalizations across Canadian communities and health regions. To do so, the Discharge Abstract Database (DAD) from the Canadian Institute of Health Information (CIHI), was linked to the long-form census by Statistics Canada. Data from three fiscal years (FY), (2006 to 2009), were pooled. Statistical analysis included hierarchical three-level mix modeling. Results of my study showed that between 2006 and 2009, out of 4305400 Canadian population aged below 75 years age, 29130 individuals were hospitalized because of ACSC diseases. This study indicates that up to 14.62 % of variation in the odds of ACSC-related hospitalization was attributable to general contextual factors at the Census Subdivision (CSD)-level, 1.13% was accounted by health regions and the remaining 84% was related to individual-level variations. In summary, results suggest high geographic variation in the odds of ACSC hospitalization across CSDs and health regions. Beyond urbanicity characteristics, the place of residence (CSDs) appeared as a more influential attribute for the odds of ACSC compared to the place within which primary or acute healthcare services were received (health regions).