Background: Physical therapy (PT) in the inpatient setting is a limited and valuable resource. Inappropriate PT consultation is costly and can lead to delays in care. Overutilization of inpatient PT services is an increasingly recognized problem. Patients with minimal or no functional limitations frequently receive PT evaluation, diverting resources and delaying care for the patients with the greatest need. Delays in discharge to patients who require post-acute rehabilitation may worsen hospital-associated disability and is associated with a prolonged functional recovery. These delays in discharge are thought to represent significant cost to the medical system. Additionally, delays in care, high patient volumes, and time wasted screening inappropriate consults may contribute to reduced job satisfaction for physical therapists and other team members. Various nursing tools exist to better triage PT resources, including the Activity Measure-Post Acute Care (AM-PAC) score. Baseline data at an academic hospital revealed that approximately one in four PT consults were inappropriate (N=29,230) across all services, as defined by AM-PAC score of >22.

Methods: To understand the extent of physical therapy (PT) overutilization at our institution, we analyzed a retrospective cohort of patients who received PT consults over a two-year period across all services including ED, medicine, surgery, and ICU patients at the University of Colorado Hospital via EHR data abstraction. Analysis demonstrated that 23.9% of PT consults were inappropriate hospital-wide. A multidisciplinary team comprised of nurses, physical therapists, care managers, informaticists, and physicians convened to construct a process map of PT ordering and performed a root cause analysis. Recurring themes included 1) providers were often unfamiliar with AM-PAC and the implications regarding need for physical therapy, 2) Typical day-to-day interactions between providers and patients often fail to capture a patient’s functional mobility, and 3) providers often order physical therapy consults at the request of the bedside nurse. Based on the root-cause analysis, two interventions were designed. The first was a modified EHR order designed with the assistance of a clinical informaticist that incorporated the patient’s AM-PAC score to provide clinical decision support suggesting an alternative to PT if AM-PAC was >22. The second was redistribution of PT ordering responsibility primarily to nursing staff, with the ability for other providers to order if needed. Interventions were rolled out in a stepwise fashion with the modified epic order going live in April 2021 and the ordering responsibility redistribution going live in 2021.

Results: Post-intervention, the average number of inappropriate PT consults was obtained via EHR data extraction. The average monthly percentage fell from an average of 23.9% pre-intervention to less than 10%, sustained for five months during the post-intervention period (Figure 1).

Conclusions: Our research demonstrates the power of utilizing traditional quality improvement tools to mitigate waste in the healthcare system. The rate of inappropriate PT consults fell from a baseline of approximately 23.9%% to less than 10% sustained in the post-intervention period. Our multifaceted interventions likely reduced the number of unnecessary PT consults and can be used to facilitate more expedient evaluation of patients truly needing PT.

IMAGE 1: Figure 1