SHM Converge Scientific Abstract ePoster Gallery-Variables Predicting Discharge to Post-Acute Care for Hospitalized Adult Patients With and Without Cognitive Impairments

Pdf Summary

This cross-sectional study examined which factors predict discharge to post-acute care (PAC) versus home among 629 hospitalized adults at Boston Medical Center from February 2024 to January 2025. The study also tested whether these predictors differed for patients with and without cognitive impairment (CogImp). Data were collected prospectively by trained coordinators, with additional EHR data extracted from Epic. Eighteen candidate predictors were analyzed using stepwise logistic regression, including demographics, comorbidity, frailty, depression screening, mobility impairment, assistance needed, and length of stay.<br /><br />The sample had a mean age of 72.1 years, was 50.9% female, and was racially diverse. There were no significant differences in PAC discharge by age, sex, or race.<br /><br />Key findings showed that level of assistance needed was the strongest and most consistent predictor of PAC discharge across the total sample and both cognitive-status groups. Compared with patients needing no assistance, the odds of PAC discharge increased progressively for those needing minimal, moderate, maximum assistance, or being dependent. Length of stay was also a significant predictor in all groups. Positive depression screening predicted PAC discharge in the total sample and among patients without CogImp. Mobility impairment predicted PAC discharge in the total sample and especially among patients without CogImp; complete mobility impairment was significant only in the non-CogImp group.<br /><br />Stratified analyses suggested effect modification by cognitive status. In patients with CogImp, mobility impairment was less predictive overall, and those with CogImp and complete mobility impairment may have been less likely to be discharged to PAC because of existing home care arrangements.<br /><br />The authors conclude that assistance needs, mobility, LOS, and depression are important predictors of PAC discharge, and that cognitive status changes how these variables operate. Future work will examine predictors of specific PAC settings and additional factors such as employment and perceived readiness for discharge.

Asset Subtitle

Joo H. Lee

Meta Tag

Author List Andrea Cheville, Chanel Chung, Joo H. Lee, Mary D. Slavin, Maya Jones, Vatsal Mangal

Category Research

Concept Post-Acute Care Discharge

Concept Home Discharge

Concept Cognitive impairment

Concept Level of Assistance

Concept Mobility Impairment

Distinguished Non-Finalist

Presenter Organization Boston Medical Center

Presenting Author Joo H. Lee

Track Transitions of Care

Keywords

post-acute care

hospital discharge

cognitive impairment

logistic regression

mobility impairment

assistance needs

length of stay

depression screening

predictors

cross-sectional study

Post-Acute Care Discharge

Home Discharge

Cognitive impairment

Level of Assistance

Mobility Impairment