Objective: To identify predictors of early and late outcomes of cardiac surgery in patients with chronic kidney disease.
Methods: Patients (n=545) with serum creatinine≥200 μmol/L or renal dialysis were identified from databases maintained by the largest Sydney cardiothoracic surgical units with data consistent with the Australian and New Zealand Society of Cardiothoracic Surgeons data definitions. The patient data were matched against the National Dialysis Database and the New South Wales Register of Births, Deaths, and Marriages. Statistical analysis was used to identify predictors of early and late outcomes.
Results: The Kaplan-Meier estimate of 1-, 5-, and 10-year survival for all patients was 78%, 56%, and 36%, respectively. The outcomes were similar after coronary bypass surgery and valve replacement and were also similar for dialysis and nondialysis patients. The odds ratios for the significant independent predictors of outcomes were, for perioperative death, age (1.4 per decade), emergency surgery (7.0), redo surgery (3.8), left ventricular impairment (moderate, 2.7; severe, 4.4); for new early postoperative dialysis, estimated glomerular filtration rate<20 mL/min (3.8), emergency surgery (2.7), tricuspid valve surgery (4.4); for new permanent dialysis within 6 months of surgery, serum estimated glomerular filtration rate<20 mL/min (odds ratio, 4.6). The hazard ratio for the independent predictors of late death in those alive 6 months after surgery was 1.4 per decade for age and 1.4 for moderate or severe left ventricular impairment.
Conclusions: Left ventricular impairment is a risk factor for perioperative and late death in patients with kidney disease. After cardiac surgery, preoperative dialysis-dependent and dialysis-free patients had similar long-term outcomes.
