Inpatient stroke rehabilitation in Ontario: are dedicated units better?


  • Conflict of interest: The opinions, results, and conclusions reported in this paper are those of the authors and are independent from the funding sources.
  • Funding: This study was funded by a grant from the Canadian Stroke Network and was supported by the Institute for Clinical Evaluative Sciences (ICES), which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (MOHLTC).

Correspondence: Norine Foley*, Aging, Rehabilitation & Geriatric Care Program, Lawson Health Research Institute, Parkwood Hospital, Rm B-3019b, 801 Commissioner's Road E., London, ON N6C 5J1, Canada.




The superiority of dedicated stroke rehabilitation over generalized rehabilitation services has been suggested by the literature; however, these models of service delivery have not been evaluated in terms of their relative effectiveness in situ.


A comparison of the process indicators associated with these two models of service provision was undertaken within the Ontario healthcare system.


All adults admitted with a diagnosis of stroke for inpatient rehabilitation in Ontario, Canada during the years 2006–2008 were identified from the National Rehabilitation Reporting System database. Each of the admitting institutions was classified as providing rehabilitation services on either a stroke dedicated or nondedicated unit. A dedicated unit was identified by the presence of a collection of geographically distinct, stroke-dedicated beds and dedicated therapists. Selected process indicators from the National Rehabilitation Reporting System database were compared between the two facility types.


Sixty-seven facilities provided stroke rehabilitation services to 6709 adult stroke patients during the years 2006–2008. Of the total number of patients who entered inpatient rehabilitation, 1725 (25·7%) received care in eight facilities that met basic criteria for a dedicated stroke rehabilitation unit. On average, these patients took significantly longer to arrive for inpatient rehabilitation (37·2 ± 155·5 vs. 22·8 ± 95·0 days, P < 0·001), were admitted with higher Functional Independence Measure scores (77·5 ± 22·5 vs. 74·8 ± 24·5, P < 0·001), had significantly longer lengths of stay (42·1 ± 25·9 vs. 35·4 ± 27·2 days, P < 0·001), and demonstrated significantly lower Functional Independence Measure efficiency scores (0·62 ± 0·47 vs. 0·88 ± 1·03, P > 0·001) compared with patients who were admitted to nondedicated units. The proportion of patients admitted to a dedicated unit and subsequently discharged home was similar to that of patients discharged from nondedicated units (70·5% vs. 68·8%, P = 0·206).


In Ontario, patients admitted to dedicated stroke rehabilitation units fared no better on commonly-used process metrics compared with patients admitted to nondedicated rehabilitation units.