Predicting Total Knee Arthroplasty Outpatient Discharge: Surgeons versus Insurance Companies

Background: Insurance companies are increasingly making unilateral determinations of admission status for primary total knee arthroplasty (TKA). These determinations may differ from those based on surgeonderived criteria for outpatient knee replacement. The goal of this study is to determine if insurance company determinations of outpatient status are as reliable as surgeon-derived criteria in predicting outpatient discharge after TKA. Methods: We retrospectively reviewed 709 patients who were preoperatively authorized for outpatient TKA. Patients were stratified into 2 groups: "outpatient per surgeon" (appropriate for outpatient surgery per institutional protocols) or "outpatient per insurance" (appropriate for inpatient surgery per institutional protocols but denied inpatient status by insurance). The primary endpoint of this study was the conversion rate of outpatient to inpatient stay. Univariate logistic regression was performed to compare the odds of conversion to inpatient stay between outpatient per surgeon and outpatient per insurance procedures and other covariates. Results: The cohort included 434 outpatient per insurance (61.2%) and 275 outpatient per surgeon (38.8%) patients. Surgeons accurately predicted outpatients' discharge 92.0% of the time, while insurance companies did so 81.3% of time (P <.001). Outpatient per insurance procedures (odds ratio [OR] 2.20, P = .003) and body mass index >35 kg/m(2) (OR 1.82, P = .026) had higher odds of being converted to inpatient. Males had higher odds (OR 1.52, P <.001) of being discharged as outpatient. Conclusion: Determining inpatient versus outpatient status is a complex decision involving both clinical and social factors. Surgeons accurately predicted outpatient discharge 92% of the time. Moreover, outpatient per insurance procedures were twice as likely to be converted to inpatient status. Therefore, insurance companies should leave deciding admission status up to both the patient and surgeon. (C) 2022 Elsevier Inc. All rights reserved.