Predictors of hospitalization in a Canadian MS population: A matched cohort study

Hospitalizations are the most costly component of healthcare in Canada, and hospitalization rates are higher in the multiple sclerosis (MS) population compared to the general population. This study aimed to examine predictors of hospitalizations in the MS population in Saskatchewan, Canada.

This retrospective cohort study used population-based health administrative data from Saskatchewan, Canada from 1996 to 2016. Subjects with MS were identified using a validated definition (≥3 hospital, physician, or drug claims for MS). Up to five general population controls were identified for each MS case and matched on sex, age, and geographical location. The rate of hospitalizations and reason for admission were determined for each case and control. Negative binomial (hospitalization rate) and binary logistic (reason for admission) regression models fitted with generalized estimating equations were used to test the following potential predictors: sex, age, median household income, calendar year, prior hospitalizations, and comorbidity status.

We identified 4,878 MS cases (11,744 hospitalizations), and 23,662 matched controls (32,541 hospitalizations). Higher comorbidity burden, older age, and prior hospital admissions were associated with an increased rate of all-cause hospitalizations for both cohorts. Males were more likely to be hospitalized than females for all-cause (adjusted rate ratio: 1.20; 95% CI: 1.07 - 1.34) and MS-specific (adjusted odds ratio: 1.34; 95% CI 1.15 - 1.55) hospitalizations. The rate of MS-specific hospitalizations decreased with age, and there was no association with comorbidity or prior hospitalizations. A diagnosis of MS was associated with decreased odds of hospitalization due to neoplasms, diseases of the circulatory system, and mental health and behavioural disorders.

Increased age, comorbidity, and prior hospital admissions are predictors of all-cause hospitalizations. Conversely, MS-related hospitalizations decreased as subjects aged, and there was no association with comorbidity. Our results highlight that reasons for hospitalizations can differ by age, and clinicians should consider this when managing patients, as they make efforts to reduce hospitalizations in the MS population.