Clinical impact of the sequentially‐checked drain removal criteria on postoperative outcomes after pancreatectomy: a retrospective study

BACKGROUND: Strict criteria for impeccably safe drain management following pancreatectomy have not yet been developed. We evaluated the utility of the sequentially-checked drain removal criteria by comparison with conventional criteria. METHODS: Postoperative outcomes of 801 patients who underwent pancreatectomy, including 395 patients for whom drain fluid amylase (DFA) < 375U/l on postoperative day (POD) 3 (control group), were used and 406 patients for whom the sequentially-checked criteria of DFA <5,000 U/l on POD 1 and DFA <3,000 U/l on POD 3 (sequentially-checked group) were used and were retrospectively evaluated. RESULTS: DFA on POD 3 and fistula risk score did not differ between groups. Significantly more patients in the sequentially-checked group met the criteria (control, 63.8% vs. sequentially-checked, 76.1%, P < 0.001). The incidences of clinically relevant postoperative pancreatic fistula (CR-POPF) (17.0% vs. 11.1%), intra-abdominal abscess (21.0% vs. 9.1%) were significantly lower in the sequentially-checked group (all P < 0.05). Multivariate analysis revealed that use of the sequentially-checked criteria was significantly associated with CR-POPF (odds ratio 0.601, 95% confidence interval [CI] 0.389-0.929; P = 0.022). C-reactive protein <15 mg/dl at POD 3 was identified as an independent predictive factor for false positive CR-POPF results in the sequentially-checked group (odds ratio 0.872, 95% CI 0.811-0.939; P < 0.001); thus, this criterion was added to create the new triple-checked criteria. CONCLUSIONS: The sequentially-checked criteria can provide safe drain management and improve postoperative outcomes.