Perioperative Predictors of Length of Stay After Total Hip Arthroplasty

• Published in: The Journal of arthroplasty 2016-07, Vol.31 (7), p.1427-1430
• Main Authors: Petis, Stephen M., MSc, MD, Howard, James L., MD, MSc, FRCSC, Lanting, Brent A., MD, MSc, FRCSC, Somerville, Lyndsay E., PhD, Vasarhelyi, Edward M., MD, MSc, FRCSC
• Format: Article
• Language: English
• Subjects: Aged; arthroplasty; Arthroplasty, Replacement, Hip; Body Mass Index; Female; hip; Humans; length of stay; Length of Stay - statistics & numerical data; Male; Middle Aged; Odds Ratio; Orthopedics; Patient Discharge; predictors; Prospective Studies; Regression Analysis; Time Factors; timed up-and-go
• ISSN: 0883-5403; 1532-8406
• DOI: 10.1016/j.arth.2016.01.005

Abstract Background Few studies had examined whether specific patient variables or performance on functional testing can predict length of stay (LOS) after total hip arthroplasty (THA). Such tools would enable providers to minimize prolonged LOS by planning appropriate discharge dispositions preoperatively. Methods We prospectively recruited 120 patients undergoing a THA through an anterior (n = 40), posterior (n = 40), or lateral (n = 40) approach. Patients performed a timed up-and-go (TUG) test preoperatively to determine if it was predictive of hospital LOS after THA. Other variables of interest included patient age, body mass index, age-adjusted Charlson Comorbidity Index, mean procedure time, and time spent in the postanesthetic care unit. A logistic regression analysis was performed to determine which variables predicted LOS greater than 48 hours, which is our institution’s target time to discharge. Results The TUG test was predictive of LOS beyond 48 hours. For every 5-second interval increase in TUG time, patients were twice as likely to stay in hospital beyond 48 hours (odds ratio [OR] = 2.02, 95% confidence interval [CI] = 1.02-4.01, P  = .043). Patient age (OR = 0.97, 95% CI = 0.90-1.05, P  = .46), body mass index (OR = 1.01, 95% CI = 0.86-1.18, P  = .90), Charlson Comorbidity Index (OR = 1.29, 95% CI = 0.68-2.44, P  = .44), mean procedure time (OR = 1.05, 95% CI = 0.97-1.14, P  = .27), and mean time in the postanesthetic care unit (OR = 1.00, 95% CI = 0.99-1.00, P  = .94) were not predictive of increased LOS. Conclusion The TUG test was predictive of hospital LOS after THA. It is a simple functional test that can be used to assist with discharge planning preoperatively to minimize extended hospital stays.