Utility of preoperative scores for predicting morbidity after cholecystectomy in patients with cirrhosis

Abstract

Patients with cirrhosis have an increased risk for cholelithiasis but also have an increased risk for morbidity and mortality after cholecystectomy. Current preoperative assessment of surgical risk is imprecise. Our aims were to identify preoperative factors that would accurately predict the risk for cholecystectomy in patients with cirrhosis. Preoperative clinical or biochemical parameters were determined for 33 patients with cirrhosis and 31 age- and sex-matched patients without cirrhosis. The use of these parameters and of the Child-Pugh and model for end-stage liver disease (MELD) scores as preoperative predictors of outcome after surgery were assessed. There were 2 deaths, both in cirrhotic patients. The overall risk for morbidity or mortality was increased in cirrhotic patients compared with controls. Postoperative morbidity was significantly associated with preoperative increases of international normalized ratio >1.2, bilirubin >1.0 mg/dL, creatinine >1.4 mg/dL, and a decreased platelet count <150 x 10(3) /mL. The MELD and Child-Pugh scores accurately predicted postoperative morbidity, with an area under the curve of 0.938 and 0.839, respectively. A preoperative MELD score of > or =8 had a sensitivity of 91% and a specificity of 77% for predicting postoperative morbidity. Persons with a MELD score of > or =8 had increased 30- and 90-day global charges and increased blood product usage. Preoperative biochemical parameters, international normalized ratio, bilirubin, platelets, and creatinine can predict increased morbidity in cirrhotic patients. A MELD score of > or =8 identifies a group at high risk for postoperative morbidity after cholecystectomy.