This proposal seeks to address two major problems in the delivery of care surrounding Total Knee Arthroplasty (TKA): 1) patients are under-informed regarding the trajectory and timing of postoperative recovery following this elective procedure, and 2) post-acute care is typically delivered according to one- size-fits-all protocols, which are inadequate for decision-making with individual patients. Our work explores a ?patients-like-me? approach to guide decision-making surrounding TKA. Briefly, for any new patient considering TKA, outcomes data from similar historical patients can be used to create a personalized reference chart (PRC) to describe the anticipated recovery profile of the new patient. The method for selecting patients-like-me is a novel extension of multiple imputation (predictive mean matching). With the PRC, clinicians and patients can more precisely judge whether recovery is proceeding better than, worse than, or just as expected, compared to a patient's peers. Deviations from the expected trajectory can be rapidly detected and addressed. Patients can be better informed regarding prognosis, and resources can be more efficiently allocated: more visits for those whose recovery is lagging; fewer visits for those excelling. With this proposal, we propose to bring the PRC innovation into practice.
First (Aim 1), we will develop procedures for optimizing and validating PRC algorithms for 3 important functional outcomes following TKA.
Second (Aim 2), we will incorporate these algorithms into a software application capable of producing PRCs at the point of care in routine practice. Finally (Aim 3), we will test preliminary efficacy of PRCs in improving functional outcomes following TKA as well as the quality of shared decision making and other outcomes such as post-acute care utilization. Following completion of this work, we will be positioned to 1) conduct a larger cluster-randomized trial to formally test the effectiveness of PRC-informed care pathway, and 2) make a web-based PRC application available for widespread use. Ultimately, we foresee PRC methodology as a means of advancing personalized medicine for a number of diverse patient populations. Our team includes clinical experts (Dr. Dawn Waugh, PT, and Dr. Michael Dayton, MD), experts in analytics (Dr. Kathryn Colborn, PhD and Dr. Stef van Buuren, PhD), an expert in shared decision making and implementation science (Dr. Daniel Matlock, MD, MPH) and experts in clinical research for TKA (Dr. Jennifer Stevens-Lapsley, PT, PhD and Dr. Andrew Kittelson, PT, PhD). This research responds to AHRQ priorities. It advances care for a major health condition (TKA), in a priority population (older adults), by providing a novel framework for shared decision-making. PRCs represent a paradigm shift from traditional one-size-fits-all approaches, with opportunities for higher quality, tailored care for individuals and more efficient resource allocation in a learning healthcare system.

Public Health Relevance

Statement The proposed research is relevant to public health because total knee arthroplasty (TKA) is the most common inpatient elective surgery, with approximately 700,000 procedures performed each year. We propose to develop and implement a clinical decision tool for shared decision making?Personalized Reference Charts (PRCS)?to facilitate personalized approaches to care. We believe our research will contribute to AHRQ's mission by supplying knowledge that will enhance individual clinical decision-making and thereby increase the efficiency of resource utilization and improve the quality of care surrounding TKA surgery in a learning health care system.

National Institute of Health (NIH)
Agency for Healthcare Research and Quality (AHRQ)
Research Project (R01)
Project #
Application #
Study Section
Healthcare Effectiveness and Outcomes Research (HEOR)
Program Officer
James, Marian
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Colorado Denver
Physical Medicine & Rehab
Schools of Medicine
United States
Zip Code