Factors in total shoulder arthroplasty revision: insights from a national database analysis

Background

Total shoulder arthroplasty (TSA) is increasingly recognized as a therapeutic solution for various shoulder disorders. However, the correlation between preoperative determinants of health and postoperative TSA revisions remains underexplored. These correlations are scientifically important as they may be used to design higher-level clinical studies to confirm if they are in-fact predictive of revision. As such, this investigation aims to identify possible health and demographic disparities between primary TSA and revision TSA cohorts.

Methods

Data for this research was sourced from the American College of Surgeons National Surgical Quality Improvement Program database from 2013 to 2021. Patients were classified based on Current Procedural Terminology (CPT) codes into 2 groups: primary TSA (CPT: 23472) and revision TSA (CPT: 23473 or 23474). Nineteen demographic and preoperative variables were compared between the cohorts using multivariate logistic regression. Odds ratios (ORs) along with their 95% confidence intervals (CIs) were calculated and graphically displayed. To account for multiple exploratory comparisons, Bonferroni correction was utilized with an adjusted significance threshold of 0.0026.

Results

TSA revision patients were more likely to arrive from a nursing home (OR 1.87, 95% CI [1.50, 2.32], P < .001), have diabetes (OR 1.09, 95% CI [1.05, 1.14], P < .001), be diagnosed with Systemic inflammatory response syndrome prior to surgery (1.38, 95% CI [1.13, 1.69], P = .002) and less likely to have independent functional status prior to surgery (OR 0.69, 95% CI [0.63, 0.76], P < .001). Moreover, those in the revision group were predominantly younger, male, receiving immunosuppressive therapy, exhibited baseline dyspnea, and had elevated American Society of Anesthesiologists class. Notably, factors such as smoking, racial background, and hypertension showed no significant differences (P > .0026).

Conclusions

Contrary to contemporary literature, our findings suggest that, when demographic factors are controlled for, smoking is not correlated with revisions in TSA when compared to primary TSA. Furthermore, patients undergoing primary TSA are more likely to be functionally independent than patients undergoing revision TSA. While at the same time, a patient’s arrival at surgery from a nursing home gave the greatest likelihood of a revision.