|
Back to 2018 Posters
IS THE POSTOPERATIVE DAY 1 DRAIN-FLUID AMYLASE LEVEL OF VALUE AFTER DISTAL PANCREATECTOMY: A NSQIP DATABASE ANALYSIS
Essa M. Aleassa*1,2, Gautam Sharma1, Gareth Morris-Stiff1
1Section of HepatoPancreatoBiliary Surgery, Cleveland Clinic, Beachwood, OH; 2Department of Surgery, United Arab Emirates University, Al-Ain, Abu Dhabi, United Arab Emirates
Background and Aims: Unlike post pancreatoduodenectomy (PD), the role of postoperative day 1 drain fluid amylase (DFA-1) in predicting postoperative pancreatic fistulae (POPF) following distal pancreatectomy (DP) is less clear. We aim to assess the significance of the cutoff DFA-1 level of 5000 U/L suggested for PD and try to establish the optimal DFA-1 threshold best correlating with fistula formation.
Patients and Methods: The American College of Surgeons-National Surgical Quality Improvement Program (NSQIP) pancreatectomy targeted files from 2014 through 2016 were queried to identify patients who underwent DP. Only patients with a recorded DFA-1 level were included. We converted the drain fluid amylase levels variable to a dichotomous variable; less or greater than 5000 U/L. As for the fistula variable, we extrapolated two extra variables: presence or absence of fistula and grades B/C or not. These variables were then analyzed using chi square to determine significance of the cut off value 5000 U/L in determining the presence of a pancreatic fistula. A Receiver Operator Characteristic (ROC) curve was plotted to determine the optimal drain-fluid-amylase level to predict fistula formation.
Results: We identified a total of 5117 cases of DP, of which only 1016 (19.9%) had a postoperative DFA-1 level recorded. A postoperative DFA-1 level of 5000 U/L significantly correlated with development of a pancreatic fistula (P <0.001), regardless of its grade, with a specificity of 83% and sensitivity of 48.1% (positive predictive value of 42.4% and negative predictive value of 86%). However, it did not determine the presence of grade B/C fistulae (P=0.142). The ROC curve, with an area under the curve of 0.712 (P<0.001), demonstrated that a sensitivity of 70% is achieved at cut off value of 1870 U/L with a specificity of 62.5% for all grades of pancreatic fistulae.
Conclusion: The cutoff DFA-1 value of 5000 U/L used in PD should not be applied to cases of distal pancreatectomy due to its low sensitivity. Furthermore, clinically significant fistulae can not be determined by DFA-1 levels which questions its usefulness in the setting of DP.
Back to 2018 Posters
|
