IS THE MAYO POST-OPERATIVE MORTALITY RISK PREDICTION MODEL APPLICABLE TO ASIAN CIRRHOTIC PATIENTS?

초록

Background and Aim: Patients with liver cirrhosis are considered to be at increased risk of mortality and morbidity after surgery. Several factors associated with postoperative mortality have been identified, including increasing scores of Child-Pugh-Turcotte, model for end stage liver disease (MELD), and American Society of Anesthesiologists (ASA) class, but it is still difficult to predict postoperative outcomes in individual patient accurately. Most recently, a new model for the prediction of postoperative mortality risk in patients with cirrhosis was developed by Mayo clinic. The prediction model, which adopts variables such as ASA, MELD, and age, was based on mortality rate of cirrhotic patients who had undergone surgery under general anesthesia from a single hospital. Therefore it is unclear if the model could be applied to the cirrhotic patients at other centers, especially in Eastern countries which may have different etiology and prognosis of liver cirrhosis. To validate the usefulness of this model in Asian cirrhotic patients, we performed the present study. Methods: One hundred nine medical records from the patients who underwent surgery under general anesthesia at Korea University Ansan Hospital between January 1996 and December 2006 were retrospectively reviewed. Seventy three patients whose survival has been verified were included in this analysis. Results: A total of 17 patients (23.3%) died after surgery. MELD, ASA, and emergency surgery were proved to be independent risk factors for mortality by univariate analysis,but only ASA remained to be a risk factor by multivariable proportional hazards model. The areas under the receiver operating curve (AUC) of the prediction model were 0.988, 0.921,and 0.871 for 30 days, 90 days, and 1 year mortality, respectively. However, actual mortality were 5.4%, 10.9%, and 15.1% at 30 days, 90 days, and 1 year, respectively, which are lower than the predictive values of 7.2±5.9%, 11.2±8.2%,22.6±8.5%. The difference between actual and predictive mortality rate at 1 year was significantly different (p<0.001) Conclusions: The risk prediction model which was developed by Mayo clinic is also considered to a useful model in Asian cirrhotic patients. However, the model tends to overestimate the mortality, especially at the time of 1 year after surgery.