OBJECTIVES. This study examines the poststroke rehabilitation experien
ce for a 20% sample of Medicare patients age 65 years and older admitt
ed to an acute-tare hospital with a stroke diagnosis during the 6-mont
h interval, January 1, 1991 to June 30, 1991. Their Medicare claims da
ta are used for two purposes: to identify current poststroke rehabilit
ation practice in the US population age 65 years and older, and to eva
luate the importance of practice variation within this population. MET
HODS. Regarding the first objective, the authors develop estimates for
many poststroke rehabilitation use and cost parameters that formerly
were unmeasured. With respect to the second objective, the authors con
struct and compare average service use rates across ail stroke patient
s in a census division and across all stroke patients residing in the
30 largest metropolitan statistical areas. RESULTS. The authors' Medic
are claims analysis indicates that 73% of stroke survivors received ei
ther postacute institutional or ambulatory rehabilitation care during
the first 6 months poststroke. The published stroke literature, on the
other hand, focuses on the minority of stroke survivors, 16.5% in the
Medicare data, who were admitted to an inpatient rehabilitation hospi
tal. Regarding the second study objective, the Medicare analysis provi
des graphic evidence that poststroke rehabilitation practice varies su
bstantially from one geographic area to another and that practice diff
erences translate into large geographic-related differences in the cos
t of poststroke rehabilitation. CONCLUSIONS. The authors believe the f
indings demonstrate a problem with inconsistent poststroke rehabilitat
ion practice.