Temporal Trends and Predictors of Thirty-Day Readmissions and Emergency Department Visits Following Total Knee Arthroplasty in Ontario Between 2003 and 2016

Abstract

Background: Total knee arthroplasty (TKA) is the second most common surgery performed in Canada. Understanding and improving quality metrics associated with such high-volume procedures is of utmost importance to maximize value within the healthcare system, which is a balance between cost and quality. Although rates and predictors of hospital readmission and emergency department (ED) visits following TKA have previously been described in privatized healthcare settings, few studies have evaluated trends in length of stay (LOS), hospital readmissions, and ED visits following TKA in a universal single-payer system.

Methods: Using data from a provincially held and validated registry, the Institute for Clinical and Evaluative Sciences, we undertook a review of all 205,152 TKAs performed in the province of Ontario, Canada, between 2003 and 2016. We determined temporal trends in utilization, LOS, readmissions, and ED visits and evaluated patient and provider predictors of hospital readmissions and ED visits using multivariate logistic regression modeling. We also grouped and described the most common reasons for readmission and ED visits based on the available International Classification of Diseases, Ninth Revision and Tenth Revision coding information.

Results: LOS decreased significantly over the study period (P < .0001), from a median of 5 days (10th percentile 3 days, 90th percentile 8 days) in 2003 to a median of 3 days (10th percentile 2 days, 90th percentile 4 days) in 2016. All-cause 30-day readmissions did not change significantly over the study period, but the rate of ED visits increased significantly over time. Predictors of 30-day readmission following TKA included older age, male gender, lower income quartile, not having a postoperative visit with a primary care physician (PCP), increased comorbidities, longer LOS, urgent or revision surgery, admission to a teaching hospital, and discharge to an inpatient rehabilitation facility. Variables that predicted increased odds of an ED visit included older age, male gender, lower income quartile, not having a postop visit with a PCP, increasing comorbidities, year of surgery, longer LOS, and revision surgery. Admission to a teaching hospital and discharge to an inpatient rehabilitation facility showed a trend toward increased odds of an ED visit.

Conclusions: We identified a significant increase in ED visits following TKA in Ontario between 2003 and 2016, with no corresponding increase in hospital readmissions despite a significant temporal trend toward shorter LOS. Predictors of ED visits and readmissions were similar, including male gender, lower income, higher comorbidities, and lacking a PCP visit postoperatively. Increased rates of ED visits following TKA in Ontario represent a quality problem, as they are associated with increased cost to the public healthcare system without any substantial benefit. Interventions aimed at redirecting patients from the ED for minor postoperative concerns should be investigated, as this is likely to improve care by reducing costs, improving efficiency, and enhancing patient experience.

Keywords: 30-day emergency department visits; 30-day readmissions; length of stay quality; outcomes; predictors; total knee arthroplasty.