Background: Discharge planning is a major component of stroke care. Early prediction of hospital disposition may increase efficient delivery of health care by initiating appropriate discharge planning and shortening length of stay (LOS).
Purpose: To identify admission factors that are associated with longer LOS and discharge to non-home setting.
Methods: In this retrospective analysis of Get With The Guidelines-Stroke data from five stroke centers of the Mount Sinai Health System in New York City from 1/1/2010 to 12/31/2015, chi-square and logistic regression analyses were used to test for admission factors associated with discharge to non-home setting. Linear regression was used to test for factors associated with LOS. Admission NIH Stroke Scale Score (NIHSS) was analyzed as a continuous variable; other covariates included demographics, medical risk factors, and weekend admission.
Results: Among 7829 patients, mean age was 70.6 years (SD 14.8); 47.5% were male; 36.8% were non-Hispanic white, 28.1% non-Hispanic black, 10.9% Hispanic; 75.9% had hypertension, 35.6% diabetes, 37.4% dyslipidemia, 22.5% coronary artery disease, 16.6% atrial fibrillation, and 24.8% prior stroke. Factors independently associated with reduced odds of discharge home were higher admission NIHSS, per point increase (OR 0.83; 95% CI 0.82-0.85), age (OR 0.97 per year; 0.97-0.98), non-Hispanic black race (OR 0.74; 0.64-0.86), and diabetes (OR 0.75; 0.66-0.85). Factors associated with longer LOS were higher admission NIHSS (0.32 day longer LOS per point increase, 95% CI 0.29-0.35; p<0.0001), non-Hispanic black (1.43, 0.93-1.93; p<0.0001), Hispanic (1.06, 0.39-1.73; p<0.0001), and atrial fibrillation (0.55, .004-1.09; p=0.048) but not age (p=0.26) or diabetes (p=0.15).
Conclusions: Several admission factors (NIHSS, age, race-ethnicity, and diabetes) were associated with discharge to non-home setting. Early identification of these patients may help initiate proper discharge planning. Hispanic ethnicity and atrial fibrillation, but not age or diabetes, were associated with longer LOS. Future research is required into why Hispanic patients, compared to non-Hispanic whites, have longer LOS despite similar discharge to home rates.