Preoperative Risk Factor Optimization Lowers Hospital Length of Stay and Postoperative Emergency Department Visits in Primary Total Hip and Knee Arthroplasty Patients

Abstract

BackgroundThe evaluation and management of outcomes risk has become an essential element of a modern total joint replacement program. Our multidisciplinary team designed an evidence-based tool to address modifiable risk factors for adverse outcomes after primary hip and knee arthroplasty surgery. MethodsOur protocols were designed to identify, intervene, and mitigate risk through evidence-based patient optimization. Nurse navigators screened patients preoperatively, identified and treated risk factors, and followed patients for 90 days postoperatively. We compared patients participating in our optimization program (N = 104) to both a historical cohort (N = 193) and a contemporary cohort (N = 166). ResultsRisk factor identification and optimization resulted in lower hospital length of stay (LOS) and postoperative emergency department (ED) visits. Patients in the optimization cohort had a statistically significant decrease in mean LOS as compared to both the historical cohort (2.55 vs 1.81 days, P < .001) and contemporary cohort (2.56 vs 1.81 days, P < .001). Patients in the optimization cohort had a statistically significant decrease in 30- and 90-day ED visits compared to the historical cohort (P30-day = .042, P90-day = .003). When compared with the contemporary cohort, the optimization cohort had a statistically significant decrease in 90-day ED visits (21.08% vs 10.58%, P = .025). The optimization cohort had a statistically significant increase in the percentage of patients discharged home. We noted nonsignificant reductions in readmission rate, transfusion rate, and surgical site infections. ConclusionOptimization of patients before elective primary total hip arthroplasty and total knee arthroplasty reduced average LOS, ED visits, and drove telerehabilitation use. Our results add to the limited body of literature supporting this patient-centered approach.