PREDICTING LENGTH OF STAY AFTER HIP OR KNEE REPLACEMENT FOR RHEUMATOID-ARTHRITIS

Objective. To identify predictors of the postoperative length of stay (LOS) after hip or knee replacement for rheumatoid arthritis (RA). Met hods. Demographic and clinical characteristics, medications, and posto perative course were abstracted from medical records of patients with RA who underwent total arthroplasties of the knee (TKA) or hip (THA) a t our institution between 1987 and 1991. The relationship between thes e variables and the postoperative LOS was examined using life tables, linear regression, and multiple regression analyses. Results. During t he 5 years of the study, 137 patients with RA underwent 119 TKA and 10 5 THA. The average LOS was 16.9 +/- 8.7 days after THA and 19.5 +/- 11 .4 days after TKA (p = 0.08). Significantly longer LOS was associated with age greater than or equal to 55 years, female sex, non-white ethn icity, poor functional status, known positive rheumatoid factor, use o f hone cement, and operating room (OR) time longer than 6 hours. in a multivariate regression model, a preoperative Steinbrocker functional class 3 or 4 was associated with an increase in LOS of 3.98 days (95% confidence interval 0.78, 7.18) and 7.14 days (2.59, 11.69), respectiv ely, while a known positive rheumatoid factor predicted an increase in LOS of 2.76 days (0.17, 5.35). Among operative factors, the use of ho ne cement was associated With a LOS that was longer by 3.50 days (0.80 , 6.20), and each hour increase in OR time with a delay in discharge o f 1.75 days (1.18, 2.33). Major postoperative wound complications incr eased LOS by 16.46 days (11.40, 21.52). Conclusion. Preoperative funct ional status is an important determinant of thr rate of recovery of fu nctional independence after surgery. Strategies for decreasing LOS aft er hip or knee replacements include optimization of preoperative funct ional status, early surgical intervention, and prevention of wound com plications.