Enhanced Selection of Candidates for Same-Day and Outpatient Total Knee Arthroplasty

Abstract

BackgroundMedicare removed total knee arthroplasty (TKA) from its inpatient-only list and private insurers created ambulatory surgical codes; these changes bring about logistical challenges for TKA episode planning. We identified preoperatively determined factors associated with hospital length of stay for (1) same-day discharge (SDD) and (2) inpatient TKA defined by Medicare’s 2-midnight rule benchmark. MethodsWe retrospectively reviewed 325 consecutive unilateral primary TKAs performed on patients completing the Perioperative Surgical Home preoperative optimization pathway within a single hospital system. Stepwise logistic regression modeling was performed to identify preoperatively determined factors associated with (1) SDD and (2) inpatient TKA. We compared these models’ ability to discern the length of stay category to the Risk Assessment and Prediction Tool (RAPT) score alone. ResultsThe cohort included 32 (10%) SDD, 189 (58%) next-day discharges, and 104 (32%) inpatients. Lower body mass index (BMI; odds ratio [OR], 0.92; 95% CI, 0.85-0.1.0; P = .04) and fewer self-reported allergies (OR, 0.66; 95% CI, 0.46-0.95; P = .03) were associated with SDD. The SDD model outperformed the RAPT alone (C-statistic, 0.73 vs 0.52; P < .01). Older age (OR, 0.96; P = .04), higher BMI (OR, 0.93; P 0.01), lower RAPT score (OR, 1.2; P = .04), and later surgery start time (OR, 0.80; P < .01) were associated with inpatient discharge. The inpatient model outperformed the RAPT alone (C-statistic, 0.74 vs 0.62; P < .01). ConclusionWe identified preoperatively determined factors associated with (1) SDD as BMI and allergies and (2) inpatient TKA as age, BMI, RAPT score, and surgery start time. Hospitals, providers, patients, families, and payers can use this information for TKA episode planning.