To predict the requirement for rehabilitation unit after acute care hospitalization for stroke: how many and which type of rehabilitation unit?

Data were obtained from the Rhône-Alpes inpatient database in Acute Care and Rehabilitation. All acute care hospitalization for stroke in Rhône-Alpes region where considered.

Five groups of acute stroke inpatient where determined according to the type of discharge required (rehabilitation, nursing home or home…) applying recommendation of the French Society of Physical Medicine and Rehabilitation (Sofmer). These groups were determined analyzing information contained in database (age, comorbidity, medical procedure, wards…). For each of the 5 groups, the type of discharge required was compared with the real discharge of the patient. When patients were admitted in rehabilitation units, logistic regression model was used to analyze effect of the type of rehabilitation (neurological unit or no) on dependence score improvement.

(1) Type of discharge required: among the 7511 discharges of surviving acute stroke, 858 (11%) had no indication for rehabilitation and should be supported in nursing home, 389 (5%) had very serious clinical conditions and required specialized post-acute care rehabilitation units, 1255 (17%) required an hospitalization in general or geriatric rehabilitation unit because of their bad prognosis factors of functional outcome and 1865 (25%) required a PMR unit. (2) When hospitalization in a PMR unit was required, 896 (48%) were actually admitted in rehabilitation unit, of which 703 in PMR. Those admitted in PMR had a greater probability of functional improvement compared with no neurological unit (aOR=1.64 [1.09–2.46]) after adjustment for age, comorbidity, initial level of dependence, and the duration of hospitalization.

Predicting and characterizing the requirement for rehabilitation center following acute stroke can help to optimize the orientation in the care pathway for better efficiency.