Short-stay total knee arthroplasty (TKA), defined as a 1-day length of stay (LOS), is feasible in many patients, yet variables identifying who are candidates for a short stay are not well described in literature. With an emphasis on cost-efficiency, we examined preoperative patient characteristics and perioperative hospital factors that correlated with a longer LOS.
A retrospective review of 381 primary TKAs was performed. Clinical measures differentiating a 1-day LOS group from that of a ≥2-day LOS group were identified.
Multiple logistic regression demonstrated older age (odds ratio [OR], 1.92; 95% confidence interval [CI], 1.34-2.77; P < .001), female gender (OR, 4.22; 95% CI, 2.35-7.57; P < .001), American Society of Anesthesiologists score 3 or 4 (OR, 2.00; 95% CI, 1.01-3.95; P = .046), atrial fibrillation (OR, 8.87; 95% CI, 1.81-43.47; P = .007), and prior TKA on the contralateral side (OR, 3.57; 95% CI, 1.27-10.05; P = .016) as significant preoperative characteristics correlating with the ≥2-day LOS group. The most significant hospital perioperative factor associated with longer stays was patients not ambulating on the day of surgery (OR, 4.09; 95% CI, 1.77-9.48; P = .001). Walking 150 ft (93% sensitive, 35% specific) on the day of surgery was predictive of patients in the 1-day LOS group. Hospital costs were US$1873 (P < .001) lower for patients in the 1-day group.
Shorter stays decrease costs associated with TKA, and more refined predictive models are needed to optimize discharge protocols. Preoperative data help allocate limited healthcare resources toward patients more likely to leave in 1 day, while perioperative data facilitate learning to create a more efficient hospital process.