Background: Physical therapy (PT) is an invaluable tool to prevent and treat the functional decline that hospitalized patients experience. Although it is a constrained resource in most inpatient settings, PT is often requested for patients with high mobility and independence who may not require skilled therapy. This can waste hundreds of PT hours and decrease the amount of time that debilitated patients spend with PT. The Activity Measure Post-Acute Care (AM-PAC) score is a validated mobility measurement tool that has been used to predict discharge destination within 48 hours of admission and to define potential overutilization of PT(1-3). However, no studies have used mobility scores to optimize PT utilization. We aimed to optimize PT utilization on hospital medicine services by creating clinical decision support based on validated mobility measurement tools.
Methods: We conducted a prospective study of all patients admitted to the direct care hospital medicine services at the University of Chicago Medical Center. Patients who left AMA, died, or discharged to hospice were excluded. We identified patients who had a physical therapy consult at any time during their admission and obtained their admission AM-PAC score. A clinical decision support tool was designed using AM-PAC scores to recommend for or against PT and was then embedded in hospital medicine “History & Physical” and “Progress Note” templates (Figure 1). Providers were required to indicate status of PT referral based on current AM-PAC scores. Baseline data was collected for one year prior to implementation of the tool and one year after implementation. Due to implementation coinciding with the start of the COVID-19 pandemic, post-intervention time periods were divided based on theorized changes in PT referral practices during the pandemic (COVID unit wave, post-COVID unit wave, post-vaccination wave). Multivariate logistic regression was used to assess the association between the intervention and PT referrals for high mobility patients controlling for number of PT consults/month, wave, and # of admits/month. P values of < 0.05 were considered statistically significant.
Results: Between October 2018 and March 2021, 6,149 admissions were eligible for the study. Compared to the preintervention period (October 2018 – October 2019) there was a lower rate of referral to PT for patients with high AM-PAC scores in the post-intervention period (February 2020 – March 2021) [39.3% vs 35.8%; t(6147) = 29.88; p < 0.001]. Using multivariate logistic regression, the documentation tool was associated with significantly lower odds of PT referral for patients with high AM-PAC scores (OR 0.78; 95% CI 0.65 – 0.93; p = 0.006) controlling for COVID wave, admits/month, and PT consults/month.
Conclusions: Our results suggest that simple EMR documentation-based clinical decision support can decrease the rate of PT consults on high mobility patients. Hospitalized patients with high mobility may not require the skills of a physical therapist so decreasing therapy referrals for these patients may lead to increased therapy time for patients at risk of hospital-associated disability. Hospitalists should consider using mobility score-based interventions to prioritize PT for at-risk patients.
