Preliminary Feasibility and Effectiveness of a New Clinical Decision Support Tool in Knee Arthroplasty Rehabilitation.

JEREMY GRABER, DPT
University of Colorado

Patients recovering from total knee arthroplasty (TKA) have unique postoperative needs, goals, and expectations.1,2 Unfortunately, most patients with TKA receive the same generic course of rehabilitation. To facilitate a more patient-centered approach, we developed a clinical decision support (CDS) tool that provides patients and clinicians with personalized recovery information following TKA.

This is a quality improvement initiative at ATI Physical Therapy in Greenville, SC. Clinicians at two ATI clinics are using the CDS tool to inform rehabilitation decisions for all patients with TKA. In this report, we analyzed the CDS tool’s preliminary feasibility and effectiveness using the RE-AIM framework.3

We compared data from the following sources to examine the CDS tool’s implementation: (1) data captured in real time by the web-based CDS tool, (2) data pulled monthly from ATI’s electronic medical record, and (3) fidelity assessment data captured during discrete patient-clinician encounters with the CDS tool. To examine the CDS tool’s preliminary effectiveness, we also compared outcomes between patients treated with the CDS tool (surgical dates from 2018-2021) and without the CDS tool (surgical dates from 2020-2022).

Reach: 63% of eligible patients were exposed to the CDS tool (Clinic 1 = 19/19, Clinic 2 = 12/30). Adoption: 100% of eligible clinicians used the CDS tool (Clinic 1 = 3/3, Clinic 2= 2/2). Implementation: Clinicians averaged 96% fidelity to the CDS tool’s core components on a per-encounter basis. Clinicians were instructed to use the CDS tool once every 3 weeks during each patient’s episode of care; clinicians used the CDS tool once every 29 days on average. (Clinic 1 = 22 days, Clinic 2 = 41 days). Effectiveness: Patients treated with the CDS tool (1) self-reported better physical function and knowledge regarding their recovery and (2) attended fewer visits compared to patients treated without the CDS tool.

This report suggests that clinicians from Clinic 1 are using the CDS tool with nearly all eligible patients and with high fidelity. However, clinicians at Clinic 2 are using the tool less frequently and with fewer patients. After discussing this data with stakeholders from Clinic 2, we discovered the COVID-19 pandemic has created challenges related to staffing and CDS tool prioritization. We are working closely with Clinic 2 stakeholders to address these challenges, and we will use the processes described in this report to provide both clinics with consistent performance feedback moving forward. Despite Clinic 2’s challenges, we are encouraged by Clinic 1’s strong implementation and by the preliminary data supporting the CDS tool’s effectiveness. We will use a mixed methods design to formally evaluate the CDS tool’s clinical impact at the end of this project. Our long-term goal is to implement the CDS tool throughout ATI’s nationwide clinic system.