Physical Function in the Hospital Is Associated With Patient-Centered Outcomes in an Inpatient Rehabilitation Facility.

Abstract

Poor knowledge of the relationships between physical function (PF) in the hospital and patient outcomes in an inpatient rehabilitation facility (IRF) limits the identification of patients most appropriate for discharge to an IRF. This study aimed to test for independent associations between PF measured via the AM-PAC "6-clicks" basic mobility short form in the hospital and outcomes in an IRF. This was a retrospective cohort study. Primary data were collected from an acute hospital and IRF at 1 academic medical center. Associations were tested between PF at hospital admission or discharge and PF improvement in the IRF, discharge from the IRF to the community, and 30-day hospital events by estimating adjusted relative risk (aRR) using modified Poisson regression and the relative difference in IRF length of stay (LOS) using Gamma regression. A total of 1323 patients were included. Patients with moderately low, (aRR=1.50; 95% CI = 1.15-1.93), moderately high (aRR=1.52; 95% CI = 1.16-2.01), or high (aRR=1.37; 95% CI = 1.02-1.85) PF at hospital discharge were more likely than those with very low PF to improve their PF while in the IRF. These same patients were more likely to discharge from IRF to the community and had significantly shorter IRF LOS. Hospital-measured PF did not differentiate risk for 30-day hospital events. Patients with moderate-but not very low or very high-PF measured near the time of acute hospital discharge were likely to achieve meaningful PF improvement in an IRF. They also had a shorter IRF LOS so may be ideal candidates for discharge to IRF. Prospective studies with larger samples are necessary to test this assertion. Providers in the hospital should identify patients with moderate PF near the time of hospital discharge as those who may benefit most from post-acute rehabilitation in an IRF.