Prediction of surgical outcomes of laparoscopic liver resections for hepatocellular carcinoma by defining surgical difficulty

Abstract

Background

Several classification systems for assessing the surgical difficulty of laparoscopic liver resection (LLR) have been proposed. We evaluated three current classification systems, including traditional Major/Minor Classification, Complexity Classification, and the Difficulty Scoring System for predicting the surgical outcomes after LLR.

Methods

We reviewed the clinical data of 301 patients who underwent LLR for hepatocellular carcinoma between March 1, 2004 and June 30, 2015. We compared the intraoperative, pathologic, and postoperative outcomes according to the three classifications. We also compared the prognostic value of the three classifications using receiver operating characteristic (ROC) curves.

Results

The Major/Minor Classification, Complexity Classification, and the Difficulty Scoring System efficiently differentiated surgical difficulty in terms of blood loss (P = 0.001, P = 0.009, and P < 0.001, respectively) and operation time (all P < 0.001). Regarding intraoperative outcomes, the Difficulty Scoring System and Complexity Classification successfully differentiated the transfusion rate (P = 0.001 and P < 0.001, respectively). However, only the Complexity Classification adequately predicted severe postoperative complications (P = 0.032), the severity of complications (P < 0.001), and the length of hospital stay (P = 0.005). In ROC curve analysis, the Complexity Classification (area under the curve [AUC] = 0.611) outperformed the Major/Minor Classification (AUC = 0.544) and the Difficulty Scoring System (AUC = 0.530) for predicting severe postoperative complications. None of the classification systems predicted recurrence or patient survival.

Conclusion

The Complexity Classification was superior to the other methods for assessing surgical difficulty and predicting complications after LLR for hepatocellular carcinoma.

from Endoscopy via xlomafota13 on Inoreader http://ift.tt/2r1JN3l
via IFTTT