Dialysis Dependence Predicts Complications, Intensive Care Unit Care, Length of Stay, and Skilled Nursing Needs in Elective Primary Total Knee and Hip Arthroplasty

Background: Limited data describe risks and perioperative resource needs of total joint arthroplasty (TJA) in dialysis-dependent patients.

Methods: Retrospective multiple cohort analysis of dialysis-dependent American College of Surgeons National Surgical Quality Improvement Program patients undergoing primary elective total hip and knee arthroplasty compared to non-dialysis-dependent controls from 2005 to 2015. Relative risks (RRs) of 30-day adverse events were determined by multivariate regression adjusting for baseline differences.

Results: Six hundred forty-five (0.2%) dialysis-dependent patients of 342,730 TJA patients were dialysis-dependent and more likely to be dependent, under weight, anemic, hypoalbuminemic, and have cardiopulmonary disease. In total hip arthroplasty patients, dialysis was associated with greater risk of any adverse event (RR = 1.1, P < .001), mortality (RR = 2.8, P = .012), intensive care unit (ICU) care (RR = 9.8, P < .001), discharge to facility (RR = 1.3, P < .001), and longer admission (1.5×, P < .001). In total knee arthroplasty patients, dialysis conferred greater risk of any adverse event (RR = 1.1, P < .001), ICU care (RR = 6.0, P < .001), stroke (RR = 7.6, P < .001), cardiac arrest (RR = 4.8, P = .014), discharge to facility (RR = 1.5, P < .001), readmission (RR = 1.8, P = .002), and longer admission (1.3×, P < .001).

Conclusion: Dialysis-dependence is an independent risk factor for 30-day adverse events, ICU care, longer admission, and rehabilitation needs in TJA patients. Thirty days is not sufficient to detect infectious complications among these patients. These findings inform shared decision-making, perioperative resource planning, and risk adjustment under alternative reimbursement models.