Identifying factors associated with high use of acute care in Canada: a population-based retrospective study

Objectives To determine demographic, socioeconomic, and clinical factors associated with being high-cost users (HCUs) in adult patients (≥ 18 years) who received acute care in Canada. Research design We conducted a retrospective study among adults who had at least one encounter with acute care facilities each year from 2011 to 2014 using national linked data sets. We defined HCUs as patients on the top 10% of the highest acute care cost users in the province, where the care was provided. Risk factors associated with being HCUs were identified using multilevel logistic regression. Provincial variations of identified risk factors were examined using logistic regression. Sensitivity analyses were also performed to investigate the influences of using different metrics and different thresholds to define high system users, missing data, and the inclusion of interaction terms on the study results. Results Between 2011 and 2014, a total of 3,891,410 patients with 6,017,430 hospitalizations were included. Patients who were male [odds ratio (OR), 1.60; 95% confidence interval (CI) 1.59–1.61], with low incomes [OR 1.42; 95% CI 1.41–1.43), with higher comorbidity score (OR 1.41; 95% CI 1.40–1.41] and older [OR 1.18; 95% CI 1.17–1.18] were more likely to be acute care HCUs. Significant interactions existed between comorbidity score and age/sex/income status. Across provinces, the associations between socioeconomic factors and being HCUs has the largest variation. When using various high system users (HSUs) definitions, the impacts of living in rural area and being visible minority on the odds of being HSUs differ. Conclusions A few demographic, socioeconomic, and clinical factors was associated with high acute care expenditures. The associations between included risk factors and being acute care HCUs vary across provinces and different definitions of high system users (HSUs).