Current comorbidity indices inadequately identify patients who experience early adverse outcomes following total shoulder arthroplasty

• Published in: Clinics in Shoulder and Elbow Vol. 28; no. 3; pp. 342 - 351
• Main Authors: Alexander R. Zhu, Eve R. Glenn, Laurence Okeke, James H. Padley, Edward G. McFarland
• Format: Journal Article
• Language: English
• Published: Korean Shoulder and Elbow Society 01.09.2025; 대한견주관절학회
• Subjects: 30-day outcomes; comorbidity index; mortality; risk stratification; total shoulder arthroplasty; 정형외과학
• ISSN: 2383-8337; 2288-8721
• DOI: 10.5397/cise.2025.00584

Background Patient-related outcomes are increasingly emphasized in anatomic (ATSA) or reverse total shoulder arthroplasty (RTSA). Identifying patients at risk of adverse outcomes is key to mitigating complications. Comorbidity indices such as the American Society of Anesthesiologists Physical Status (ASA), classification, modified Charlson Comorbidity Index (mCCI), Elixhauser Comorbidity Measure (ECM), and 5-Factor Modified Frailty Index (mFI-5) can predict postoperative complications but were developed primarily for inpatient populations. Whereas studies have assessed these indices for total joint arthroplasty, few have assessed their performance for complications following TSA. This study compared their prognostic values for 30-day adverse outcomes after TSA. Methods Using the National Surgical Quality Improvement Program database, 39,810 patients who underwent ATSA or RTSA (2011– 2022) were analyzed. Logistic regression evaluated six outcomes: non-home discharge, length of stay greater than 1 day, major or minor complications, readmission, and mortality. Predictive performance was assessed using the concordance statistic (C statistic), where values greater than 0.7 indicate good discrimination and those greater than 0.8 indicate excellent discrimination. Results Of all the indices, only the mCCI was a good predictor of non-home discharge (C statistic, 0.713; 95% CI, 0.705–0.722). The ASA, ECM, and mFI-5 demonstrated limited predictive value for all outcomes. Conclusions Among the indices analyzed, only the mCCI demonstrated acceptable predictive accuracy and only for discharge destination. None of the indices effectively identified patients at risk for adverse outcomes, underscoring the need for a TSA-specific scoring system. Level of evidence III.