BMC Musculoskeletal Disorders (Apr 2021)
Identifying factors predicting prolonged rehabilitation after simultaneous bilateral total knee arthroplasty: a retrospective observational study
Abstract
Abstract Background Rehabilitation is an effective procedure for promoting functional recovery after simultaneous bilateral total knee arthroplasty (TKA); however, it has been cited as a significant economic burden of medical care. We hypothesized that preoperative factors, including age, sex, body mass index, living alone, the knee society function score (KSS), the American society of anesthesiologists (ASA) class, hemoglobin (Hb), albumin level, mean range of motion, and the Kellgren–Lawrence grade, would predict prolonged rehabilitation utilization. Methods In total, 191 patients undergoing simultaneous bilateral TKA in a single hospital were enrolled. The successful compliance group included patients who completed their rehabilitation program and could return to their residence within 3 weeks after surgery (n = 132), whereas the delayed group included the remaining patients (n = 59). Logistic regression analysis was performed using preoperative factors. A prediction scoring system was created using the regression coefficients from the logistic regression model. Results Logistic regression analysis revealed that age (β = − 0.0870; P < 0.01) and Hb (β = 0.34; P < 0.05) were significantly associated with prolonged rehabilitation programs, whereas body mass index, living alone, KSS score, and ASA class were not significantly associated with successful completion of rehabilitation programs; however, these factors contributed to the prediction scoring formula, which was defined as follows: Score = 10 - 0.09 × age - 0.09 × body mass index - ( 0.56 × living alone [ alone : 1 , others : 0 ] ) + 0.03 × KSS stairs + 0.34 × Hb - 1.1 × ASA class . $$ {\displaystyle \begin{array}{l}\mathrm{Score}=10-\left(0.09\times \mathrm{age}\right)-\left(0.09\times \mathrm{body}\ \mathrm{mass}\ \mathrm{index}\right)-\left(0.56\times \mathrm{living}\ \mathrm{alone}\ \right[\mathrm{alone}:1,\\ {}\mathrm{others}:0\left]\right)+\left(0.03\times \mathrm{KSS}\ \mathrm{stairs}\right)+\left(0.34\times \mathrm{Hb}\right)-\left(1.1\times \mathrm{ASA}\ \mathrm{class}\right).\end{array}} $$ The C-statistic for the scoring system was 0.748 (95% confidence interval, 0.672–0.824). The positive and negative likelihood ratios were 2.228 (95% CI, 1.256–3.950) and 0.386 (95% CI, 0.263–0.566), respectively. These results showed an increase of 15–20% and a decrease of 20–25% in the risk of prolonged rehabilitation. The optimal cutoff point for balancing sensitivity and specificity was 3.5, with 66.6% sensitivity and 78.0% specificity. Conclusions Older age and lower preoperative Hb were significantly associated with prolonged rehabilitation programs. We defined a new scoring formula using preoperative patient factors to predict prolonged rehabilitation utilization in patients undergoing simultaneous bilateral TKA.
Keywords
