Does computed tomography–based local soft tissue composition more accurately predict outcomes following anatomic total shoulder arthroplasty compared with body mass index?: a pilot study

Background

Elevated body mass index (BMI) has been variably associated with increased surgical time, hospital length of stay (LOS), perioperative complications, postoperative emergency department return, and readmissions following primary anatomic total shoulder arthroplasty (aTSA). Computed tomography–based assessment of local soft tissue composition may better predict these outcome measures.

Methods

A retrospective review of consecutive primary stemless aTSA undertaken at a single academic institution was performed. Patient demographics and surgical data, including age, BMI, age-adjusted Charlson Comorbidity Index, operative time, and LOS, were collected. Local adipose and muscle area, in addition to proximal humeral bone area at three standardized axial levels, were measured using open-source medical imaging software. Regression analysis was utilized to determine the correlation between BMI, adipose-to-muscle ratio (AMR), and soft tissue-to-bone ratio with postoperative outcomes.

Results

There were 115 total cases included in this study. The median age of the cohort was 62 years (interquartile range 55, 69). The median BMI was 33.4 (interquartile range 28.8, 37.3). BMI correlated with AMR and soft tissue-to-bone ratio (r = 0.564 and r = 0.640, respectively, P < .001). Following regression analysis, only AMR was significantly correlated with increased LOS (incidence rate ratio 1.19, 95% confidence interval (CI) 1.03-1.37, P = .018), the risk for in-hospital medical complications (odds ratio 1.92, 95% CI 1.02-3.61, P = .045), and intraoperative total blood volume loss (r = −0.359, 95% CI −0.187 to −0.507, P = .034).

Conclusion

These findings suggest that local adipose tissue distribution about the shoulder may be more predictive than BMI with regards to perioperative clinical outcomes following primary aTSA.