A national database was queried to evaluate the peri-operative outcomes of patients with chronic kidney disease undergoing total shoulder arthroplasty.
The association of chronic kidney disease on in-hospital outcomes after shoulder arthroplasty has not been well characterized. Our aim is to determine the effect of non-dialysis dependent chronic kidney disease (NDD-CKD) and dialysis dependent chronic kidney disease (DD-CKD) on complications, hospital cost, and length of stay (LOS) after shoulder arthroplasty. We hypothesize that NDD-CKD and DD-CKD will lead to significant increases in complications, cost, and LOS.
The National Inpatient Sample was queried from 2007 to 2015 for all patients undergoing elective total, reverse, or partial shoulder arthroplasty. Patients were identified as having NDD-CKD or DD-CKD through ICD-9 procedure and diagnosis codes. Data regarding patient demographic, hospital characteristics, comorbidities, in-hospital complications, and mortality were retrieved. Patient comorbidities were stratified using the Elixhauser comorbidity index. Separate multivariate regression models were run to assess the contribution of NDD-CKD and DD-CKD to in-hospital complications and mortality. Differences on cost and LOS were assessed on cohorts matched by patient and hospital specific factors.
490,868 patients underwent TSA in this study period. 25,254 patients (5.1%) had a diagnosis of NDD-CKD and 1,055 (0.2%) had a diagnosis of DD-CKD. NDD-CKD and DD-CKD was an independent predictor of mortality, surgical site infection (SSI), myocardial infarction, UTI, stroke, deep vein thrombosis, transfusion, hemorrhage, pneumonia and non-home discharge. For the NDD patients, the most striking differences were found in mortality (OR:1.80, p<0.001), SSI (OR: 1.95, p<0.001), hemorrhage (OR: 1.74, p<0.001), pneumonia (OR: 1.49, p<0.001), deep vein thrombosis (OR: 1.41, p=0.003), and non-home discharge (OR: 1.27, p<0.001). For the DD patients, the most striking differences were mortality (OR: 10.64, p<0.001), stroke (OR: 4.88, p<0.001), deep vein thrombosis (OR: 4.17, p<0.001), myocardial infarction (OR: 3.95, p<0.001), SSI (3.17, p<0.001), transfusion (OR: 2.98, p<0.001), hemorrhage (OR:2.00, p<0.001), and UTI (OR:1.617, p<0.001). Non-infectious wound complications and pulmonary embolism were not significantly associated with either NDD-CKD (p=0.120) or DD-CKD (p=0.123). Post-operative prosthesis complications were not associated with NDD-CKD (p=0.468) but DD patients had 3.98 times the odds (p<0.001). A patient with NDD-CKD cost an additional $1,610 and had a 0.67-day greater LOS than non-CKD patients (both p<0.001). Patients with DD-CKD cost an additional $8,940 and had a 3.49-day greater LOS compared to the non-CKD cohort (both p<0.001).
Mortality, complication rates, hospital cost, and LOS were higher in the NDD-CKD cohort, and most associations were uniformly worsened further by DD-CKD. With elective shoulder arthroplasty volume rising in the United States, enhanced perioperative care of patients with CKD is essential to reduce the risk and cost associated with CKD. These findings support optimization of reversible causes of CKD and may bring into question the elective surgical candidacy of these patients.