Early detection of postoperative pancreatic fistula (POPF) may help to improve the outcome following pancreatic surgery, and exclusion of POPF may allow early drain removal which can accelerate recovery. The aim of this study was to evaluate the diagnostic accuracy of drain/plasma pancreatic amylase values on postoperative day 1 (DPA1/PPA1) in POPF by means of a systemic review and meta-analysis.
Online journal databases and a manual search up to March 2015 were used. Studies clearly documenting DPA1 or PPA1 in predicting overall POPF (Grade 0 vs A+B+C) or clinically relevant POPF (Grade 0+A vs B+C) following pancreatic surgery were selected. Pooled predictive parameters were performed using STATA 12.0.
Fifteen studies were finally identified with a total of 4331 patients. The pooled sensitivity and specificity of DPA1 were 0.92 (95% confidence interval (CI) 0.81–0.96) and 0.77 (95% CI 0.64–0.86) for predicting overall POPF and 0.79 (95% CI 0.61–0.90) and 0.83 (95% CI 0.74–0.89) for predicting clinically relevant POPF. The pooled sensitivity and specificity of PPA1 were 0.74 (95% CI 0.63–0.82) and 0.62 (95% CI 0.55–0.70) for overall POPF. After the DPA1 at/over cutoff values for overall POPF or clinically relevant POPF, corresponding post-test probability (Post-test (+)) (if pretest probability was 50%) was 80% and 82% respectively, while, if values were below the cutoff values, the post-test probability (Post-test (−)) was 10% and 20% respectively. Post-test (+) and Post-test (−) of PPA1 for overall POPF were 66% and 30% respectively. In subgroup analysis, the summary sensitivities of cutoff <1000 group and cutoff >1000 group were 0.96 (0.92–0.98) and 0.85 (0.64–0.95), respectively; the summary specificities were 0.59 (0.44–0.72) and 0.86 (0.80–0.91) respectively. Positive LR were 2.3 (1.7–3.3) and 6.2 (3.7–10.2) respectively. Negative LR were 0.06 (0.03–0.14) and 0.18 (0.07–0.47) respectively.
DPA1 is a useful predictive test for overall POPF and clinically relevant POPF which has good sensitivity and specificity based on the current studies. Meanwhile, it should be cautiously applied to clinical practice because cutoffs had a wide range between studies.