Complications and mortality in patients with chronic kidney disease undergoing total shoulder arthroplasty

Abstract

Shoulder arthroplasty (SA) provides pain relief and functional improvement in patients with shoulder arthritis. With increasing incidence of SA, there is a growing effort to risk-stratify these patients to diminish complication rates. The aim of this study is to quantify complication risk, resource utilization, and mortality of patients with chronic kidney disease (CKD) who are undergoing SA (both non-dialysis dependent and dialysis dependent) to identify any disease-severity relationship with delirious SA outcomes. We queried the Nationwide Inpatient Sample database for patients with and without diagnosis of CKD SA from 2007 to 2015. Additional patient comorbidities were stratified using the Elixhauser comorbidity index. Outcomes of interest include mortality, non-surgical and surgical complications, and resource utilization metrics. Multivariate logistical regression was run to determine odds ratios for development of these complications in the non-dialysis dependent (NDD) and dialysis-dependent cohort (DD). Between 2007 and 2015, the prevalence of patients with CKD undergoing SA more than doubled. On univariate analysis, mortality, non-surgical and surgical complications, and all resource-utilization metrics were increased in CKD patients, with a direct disease-severity-dependent relationship; the rate of inpatient mortality in the DD cohort after SA was 32 times higher than patients without CKD ( p <0.001). The association between CKD and complication rates, and the disease-severity relationship was conserved on multivariate analysis. The OR for mortality in the DD cohort was 10.6 ( p <0.001). Patients with CKD undergoing SA had a far greater risk of mortality, development of postoperative surgical and non-surgical complications, and resource utilization metrics during the postoperative inpatient stay after SA. CKD is an independent risk for a detrimental postoperative course after SA, and this relationship is worsened in patients requiring dialysis. The CKD population represents a potentially hazardous candidate for elective SA. Level III