Predictors of inpatient mortality following reverse shoulder arthroplasty

The number of reverse shoulder arthroplasties (RSA) performed each year is growing rapidly, especially in elderly patients and with expanded indications including geriatric proximal humerus fractures. As the elderly population grows and the number of RSA's annually continues to rise, there will be a proportionate number of adverse events and mortality. However, the rate of early mortality has consistently shown to be less than 1%, so a large-scale analysis of possible risk factors for post-operative mortality is warranted.

A retrospective multivariate analysis of 59,915 patients from the National Inpatient Sample database between 2016 and 2019 was performed. Patients who underwent RSA were identified based on ICD-10 code. Patients were divided into two groups, early mortality and no mortality. Early mortality was defined as those who died within the same admission. Patient demographics and medical comorbidities were evaluated. Hospital admission status was classified as elective or non-elective. Odds ratios for predictive variables were measured as a ratio of incidence between the early mortality and no mortality groups.

The overall incidence of inpatient mortality was 0.07%. The incidence of mortality for elective admissions was 0.04% and for non-elective admissions was 0.34%. On univariate analysis, age greater than 75 years (p < 0.001), octogenarians (p < 0.001), nonagenarians (p < 0.001), and non-elective admission (p < 0.001) were associated with early mortality following RSA. Upon multivariate analysis, age greater than 75 years old had 4 times the odds of early mortality following RSA (OR 4.20; 95%CI (1.67, 10.60); p < 0.001) while non-elective admission had about 5 times the odds (OR 5.38; 95%CI (2.75, 10.53); p < 0.001).

Age greater than 75 years old has 4-fold higher odds and non-elective admission has 5-fold higher odds of early mortality following RSA. Appropriate pre-operative counseling should be performed with elderly patients and those undergoing non-elective indications for RSA.