Rehabilitation after total knee arthroplasty (TKA) is typically generic and inefficient; most patients receive the same contents and dosage of rehabilitation regardless of their individual needs, preferences, or expectations. This generic treatment paradigm will be unsustainable in the Veterans Health Administration (VHA) as the demand for TKA surgery and postoperative care increase exponentially in the near future. Without new strategies to improve the efficiency of TKA rehabilitation, organizations like the VHA will struggle to meet and pay for this surging demand, and Veterans may be denied timely access to the postoperative care they need for optimal recovery. We have developed an innovative new clinical decision support (CDS) tool to optimize efficiency in TKA rehabilitation. Using the actual recovery data of similar historical patients, the tool can predict the recovery trajectory for new patients after TKA. This allows clinicians to (1) allocate rehabilitation resources based upon individual need, (2) identify Veterans at risk for suboptimal outcomes early after surgery, (3) tailor treatment strategies to Veterans’ unique goals and clinical presentation, and (4) monitor Veterans’ recovery relative to expected throughout postoperative rehabilitation. In this project, we propose to expand the CDS tool’s capabilities by establishing utilization guidelines based upon individual Veteran’s predicted recovery (Aim 1). These guidelines will be established by expert consensus in a three round Delphi process. Subsequently, we will test the CDS tool’s impact on Veteran’s functional recovery and rehabilitation utilization in four VHA outpatient physical therapy clinics using a pre-post design (Aim 2). We will compare patient-reported function (Lower Extremity Functional Scale) and physical therapy visit utilization between cohorts of Veterans treated with and without the CDS tool. Additionally, we will gather data from participating VHA clinicians and Veterans regarding the translation of our tool into clinical practice to assess its readiness for dissemination throughout the VHA (Aim 3). This will include qualitative data from participant focus groups and quantitative process data regarding the tool’s utilization. Ultimately, we expect this study will serve as a template for expanding our tool’s capabilities into numerous VHA populations in rehabilitation and beyond.