Skip to main content
Fig. 1 | BMC Surgery

Fig. 1

From: Effective prediction of postoperative complications for patients after open hepatectomy: a simplified scoring system based on perioperative parameters

Fig. 1

Predictive efficacy of the three variables and the combined variable. Note: With these identified independent risk factors, the ROC curves were independently graphed (a, b, c, and d). The AUCs for diagnoses of postoperative complication were 0.568, 0.688, 0.687, and 0.688 for the variables: diabetes, scale of surgery, serum potassium, and blood loss versus body weight, respectively. Furthermore, the AUC of the four combined variables was 0.804 (e). After assigning scores 0 or 1 to each variable, the AUC of the newly constructed and simplified scoring system was 0.803 (f). In addition, the score of the group with no complications was much lower than that in the group with complications (P < 0.0001; g). With the multivariate analysis model, Hosmer-Lemeshow test showed the expected value and observed value were positively correlated (Pearson r = 0.942, P < 0.0001; h); AUC = area under curve; ROC = receiver operating characteristic

Back to article page