Objective: To examine the association between initial hematocrit level at the time of ischemic stroke, discharge destination, and resource utilization.
Design: Case series.
Setting: University hospital.
Participants: A total of 1012 consecutive patients with ischemic stroke admitted to a university health system between August 3, 1995, and June 24, 1999.
Interventions: Not applicable.
Main outcome measures: Length of stay, hospital cost, and discharge disposition.
Results: Of 1012 patients presenting with ischemic stroke, 58% were discharged home, 10% were discharged home with home care services, 15% were discharged to a rehabilitation hospital, 11% were discharged to a skilled or intermediate care facility, and 6% died. After adjusting for age, sex, race, and comorbidities, a significant association (P=.009) existed between discharge outcome and initial hematocrit level. The probability of achieving an equivalent or less favorable outcome increased at both high and low hematocrit levels, with a minimum probability at a hematocrit level of approximately 45%.
Conclusions: An association exists between hematocrit level at the time of ischemic stroke and discharge outcome. Midrange hematocrit levels appear to be associated with discharge to home rather than to an inpatient rehabilitation unit or to a nursing facility. Further study is indicated to examine the relationship among hematocrit level, stroke severity, and outcome.