Skip to main content
Physiatry '26 Main banner
Final Program


Final Program

TBI

1029 - Receiver Operating Curve analysis of FIM scores for predicting discharge to the home/community following inpatient rehabilitation in a US national cohort of patients with traumatic brain injury

Thursday, February 19, 2026
5:00 PM - 6:30 PM AST

    Presenting Author(s)

  • DK

    David S. Kushner, MD

    Clinical Professor
    University of Miami Miller School of Medicine; Department of Physical Medicine & rehabilitation
    Miami Beach, Florida, United States

    • Co-Author(s)

  • DJ

    Doug Johnson-Greene, PhD, MPH, ABPP

    Professor and Vice-Chairman
    University of Miami Miller School of Medicine; Department of Physical Medicine & Rehabilitation
    Miami, Florida, United States

  • CM

    Cheryl Miller, DrOT

    Vice President, Therapy Operations
    Encompass Health Corporation
    Birmingham, Alabama, United States

    • Submitter(s)

  • DK

    David S. Kushner, MD

    University of Miami Miller School of Medicine; Department of Physical Medicine & rehabilitation
    Miami Beach, Florida, United States

Objectives: Effective and valid predictors of discharge to the community are needed in Inpatient-Rehabilitation (IR) for people who have Traumatic-Brain-Injury (TBI) to appropriately allocate resources and to aid in discharge planning, thus this study.

Design: US national retrospective-observational-cohort-study of 21,009 patients with TBI admitted to IR 1/1/2015-9/30/2019 comparing discharge destinations by demographic (Age, gender), length-of-stay (LOS) and Functional-Independence-Measure-[FIM] discharge scores.  Receiver-Operating-Curve (ROC) analyses were done comparing FIM discharge scores of walking/wheelchair mobility, bed/chair/wheelchair transfers, bladder management/sphincter control, and cognitive skills average scores-(comprehension, expression, social interaction, problem solving and memory) for prediction of discharge to community/home.

Results:

76% of the cohort of TBI patients were discharged to home/community, mean-age=71; average-LOS=13.3-days, average-total-FIM-gain=41. Of the remaining cohort 11% were discharged to SNFs and 13% were discharged back-to-acute-care hospitals.  Those discharged to the community were on average younger, had a shorter length of stay, and higher overall FIM gains.  ROC analyses found that walking/wheelchair discharge FIM included the greatest area under the curve (.897) with Bed/chair/wheelchair Transfers (.879) and bladder management/sphincter control discharge FIM (.774) being the second and third most significant predictors.  Interestingly, cognitive predictors were not as significant as the other predictors tested.  Patients that achieved a FIM Walking/wheelchair discharge score of 5 or greater showed reasonable sensitivity (54%) and optimal specificity (91%) for discharge to the community.

Conclusions:

In TBI IR Walking/Wheelchair mobility Discharge FIM is the most significant predictor for discharge to the community with scores of 5 or greater showing optimal sensitivity and specificity.