RISK ASSESSMENT OF ACUTE-RENAL-FAILURE AFTER THORACOABDOMINAL AORTIC-ANEURYSM SURGERY

Objective The authors sought a statistical model that permits the calc ulation of the chance of postoperative dialysis in patients undergoing thoracoabdominal aortic aneurysm repair. The preoperative risk of eac h individual patient was assessed. Summary Background Data Acute renal failure requiring dialysis after thoracoabdominal aortic aneurysm sur gery is a common complication. Despite modern intensive care facilitie s, it continues to be associated with high morbidity and mortality. Me thods The authors studied the influence of preoperative and intraopera tive variables thought to be important clinically in the development o f serious acute renal failure in 88 consecutive patients. Simple cross -clamping was used with the Crawford's graft inclusion technique witho ut shunts or heparin. Statistical methods used include univariate and multivariate analysis (stepwise logistic regression analysis and Cox p roportional hazard model). Results The hospital mortality was 11.4% (n = 10). Long-term survival at 2 years was 78% (+/-4.4), and at 5 years was 54% (+/-5.3). A logistic regression model with only two explanato ry variables to predict postoperative dialysis showed an almost perfec t fit (chi(2) = 4.2, p = 0.84): the risk factors of postoperative dial ysis were age (p = 0.003) and the preoperative serum creatinine level (p = 0.04). As determined by stepwise Cox regression analysis, postope rative dialysis also was a significant risk factor for late death (p = 0.002). Conclusions Using this statistical model, the chances of post operative dialysis can be estimated exactly before operation; age and preoperative serum creatinine level are the only variables predictive for postoperative dialysis.