Accurate prediction of severe postoperative complications after pancreatic surgery: POSSUM vs E‐PASS

Background/Purpose Few reports have evaluated the differences in the predictive accuracy between the physiological and operative severity score for the enumeration of mortality and morbidity (POSSUM) and estimation of physiologic ability and surgical stress (E‐PASS) in pancreatic surgery. Thus, we evaluated the accuracy and similarity of POSSUM and E‐PASS for the prediction of severe postoperative complications (PCs) after pancreatic surgery. Methods We enrolled 343 consecutive patients who underwent pancreatic surgery in our department between April 2006 and September 2017. The difference in predictive values of POSSUM and E‐PASS for the occurrence of PCs ≥ Clavien‐Dindo grade IIIa (PCs‐CD ≥ IIIa) was nonparametrically compared. The predictive accuracy and similarity of each tool was examined using the receiver operating characteristic (ROC) curve and linear regression analyses. Results Forty‐five patients developed PCs‐CD ≥ IIIa. E‐PASS had a significantly higher predictive value for estimating PCs‐CD ≥ IIIa occurrence (P = .002) than did POSSUM. The area under the curve value in ROC analysis was significantly higher in E‐PASS than in POSSUM (0.643 vs 0.543, P = .014), with a weak positive correlation in the predictive value between E‐PASS and POSSUM (R2 = .333, P < .001). Conclusion Estimation of physiologic ability and surgical stress was useful for predicting severe PCs after pancreatic surgery and had a higher accuracy than POSSUM.