Electronic Thesis and Dissertation Repository


Doctor of Philosophy


Epidemiology and Biostatistics


Dr. Mark Speechley

2nd Supervisor

Dr. Robert Teasell

Joint Supervisor


Stroke is a leading cause of death and disability in Canada. As patients, their families, and their friends adjust to life after stroke, organized rehabilitation can play an important role in functional recovery and improving quality of life. Best-practice recommendations suggest that moderately-to-severely impaired patients receive care in an inpatient rehabilitation unit and more mildly impaired patients in out-of-hospital settings (outpatient clinics or in-home). However, data from Ontario (Canada’s most populous province) suggest that post-stroke rehabilitation resources in both settings may be lacking. This has led to concern that some patients may be receiving rehabilitation that is not appropriate for their needs, while others receive none at all. The objective of this thesis was to formally test the hypotheses that access to rehabilitation varies across the province and that this variation is due, in part, to limited availability of rehabilitation resources. An integrated article approach was adopted consisting of two literature reviews and two original research papers.

Literature reviews were performed to identify patient-level variables that can be used to 1) predict functional outcomes after inpatient rehabilitation and 2) infer suitability for early supported discharge to community-based rehabilitation. Findings from the first review were used to inform analyses testing variation in the proportion of patients discharged to inpatient rehabilitation across regions of Ontario, while adjusting for patient-level characteristics. Hierarchical logistic regression confirmed variability in referral patterns across the province, but mixed results in the association between resources and the adjusted probability of discharge to rehabilitation. Results from the second review were used to inform an ecological study of regional variation in the proportion of mild stroke patients unnecessarily admitted to inpatient rehabilitation after stroke across Ontario. This study also confirmed suspicions that variability exists across the province and suggested an association with the availability of in-home rehabilitation services. In combination, these articles offer Ontario’s policy makers confirmation of regional inequity in access to post-stroke rehabilitation and evidence to justify further exploration into the possibility that regional investment in rehabilitation may have a positive effect. The methods proposed here may also be useful in informing future health system evaluations.