Correlations Between a Dedicated Orthopaedic Complications Grading System and Early Adverse Outcomes in Joint ArthroplastyHarris, Dorothy, Y., MD1,a; McAngus, Jillian, K., MS1; Kuo, Yong-Fang, PhD1; Lindsey, Ronald, W., MD1
Background Reliable classification of postoperative complications is important for quality improvement efforts. In 2014, The Knee Society proposed a grading system for complications after TKA, but to our knowledge, a relationship between complication grades and surgical outcomes has not yet been established.
Questions/purposes We attempted to determine (1) whether an association exists between complication grade and early adverse outcomes after TKA and THA, and (2) what proportion of the variability in complications could be associated with the classification grade (a metric of potential predictive value of the grading schema).
Methods A total of 210 primary THAs and TKAs in 201 patients performed at one center from January 1, 2011 to December 31, 2011 were reviewed; of those, 188 patients (94%; 197 procedures) had complete 90-day postoperative data and were evaluated retrospectively for postoperative complications. We defined and graded complications according to the classification system proposed by Iorio et al. and The Knee Society. Early adverse outcomes assessed included length of hospital stay and unplanned readmissions or reoperations. A total of 254 complications were documented in 135 patients (137 procedures); 53 patients (60 procedures) had no complications. Bivariate analyses were conducted to identify associations between complication grade and early adverse outcomes and patient variables; analyses considered patient variables including age, sex, status as a state prisoner (yes or no), American Society of Anesthesiologists score, BMI, and procedure (TKA or THA). Multiple regression and logistic regression analyses were conducted to determine the association between complication grade and early adverse outcomes (length of stay [LOS] and unplanned readmission or reoperations) adjusted for confounding patient variables. Alpha was set at 0.05 for two-sided tests.
Results Maximum complication grade (range, from 0-4) was associated with a longer LOS (for each point increase of maximum grade, LOS increased 0.105 ± 0.024 days, p < 0.001) and more readmissions or reoperations (odds ratio [OR], 3.79; 95% CI, 1.91-7.54; p < 0.001). Total grade (range, 0-22) also was associated with increased LOS (for each point increase of total grade, LOS increased 0.032 ± 0.006 days, p < 0.001) and increased readmissions or reoperations (OR, 1.34; 95% CI, 1.18-1.53; p < 0.001). Total grade could account for 38% of the variation in LOS and readmissions or reoperations (C-statistic = 0.94; 95% CI, 0.90-0.98); whereas maximum complication grade could account for 35% of the variation in LOS and readmissions or reoperations (C-statistic = 0.35; 95% CI, 0.88-0.96). Thus, we found total grade to be a slightly better predictor of LOS and readmissions or reoperations than maximum grade.
Conclusions We found that the proposed grading system is applicable to TKA and THA in terms of documentation of complication severity and as an indicator of increased LOS and increased unplanned readmissions or reoperation rates. That total complication grade was a better predictor of LOS than maximum grade suggests that multiple complications of a lesser grade can be just as important as a single higher grade complication in terms of effect on outcomes.