Background: Hospitalists rely on recommendations from occupational and physical therapists (OT and PT) to determine safe discharge (DC) location and level of care for hospitalized patients. During the COVID-19 pandemic, social distancing limited interactions between patients and providers, including PTs and OTs. As a result, occupational and physical therapy (OT and PT) evaluations of patients at some institutions transitioned to a telehealth model. Telehealth has been used by OTs and PTs in the inpatient and ambulatory setting to improve outcomes and help patients achieve functional goals1-4. However, whether telehealth can be used by OT’s and PT’s to make DC recommendations for hospitalized patients is unknown. The purpose of this study was to assess the reliability of telehealth therapy assessments for determining DC recommendations for patients on hospital medicine services.
Methods: We conducted a retrospective analysis of all patients under investigation for COVID-19 diagnosis on hospital medicine services at the University of Chicago medical center. Patients who had orders for referral to PT or OT first received a telehealth evaluation comprised of chart review (age, reason for admission, lab results, vital signs, major in-hospital medical events, AM-PAC mobility score), consultation with bedside nurse, and patient/representative interview via telephone. DC recommendations were determined based on those components. After COVID tests were confirmed positive or negative (24-48 hours), an in-person assessment was completed and a DC recommendation was again determined. In-person assessments were not required to be performed by the same therapist performing the virtual assessment, were not blinded, and were conducted independently. Patients were required to have ≥ 2 evaluations by a physical or occupational therapist as this was an indicator of having both a virtual and in-person visit. Agreement between virtual and in-person DC recommendations was assessed using Cohen’s kappa statistic. Exact discharge recommendations (no therapy, outpatient therapy, home health therapy, acute or subacute rehabilitation, long-term acute care) and general discharge recommendations (home vs. post-acute care) were both analyzed. P-values < 0.05 were considered statistically significant.
Results: 195 patients received therapy assessments via telehealth during the study period. Of these, 104 had ≥ 2 visits from OT, PT, or both. Compared to in-person discharge recommendations, telehealth discharge recommendations made by physical therapists demonstrated 42.3% agreement [kappa statistic 0.24; p < 0.001] for exact level of care and 70.2% agreement [kappa statistic 0.25; p < 0.01] for general discharge location. For occupational therapists, agreement was 44.2% [kappa statistic 0.21; p < 0.001] and 61.5% [kappa statistic 0.22; p < 0.05], respectively. Therapists ranged in experience from 1 – 38 years indicating minimal guessing during evaluations.
Conclusions: Our results show a strong agreement between in-person and telehealth evaluations perfromed by expereinced PTs and OTs for general discharge recommendations (home vs. post-acute care). Agreement between recommendations on exact level of care was moderate. This suggests that inpatient telehealth therapy assessment may be a feasible means of discharge location determination. Given that therapists are a constrained resource on many hospital medicine services, this strategy could be used outside beyond the pandemic to improve hospital throughput.