The International Study Group for Pancreatic Surgery (ISGPS) defined criteria to objectively standardize delayed gastric emptying (DGE) after pancreaticoduodenectomy (PD). These criteria are inclusive by design, and may overestimate actual DGE incidence. This study critically examined individual DGE cases after PD to determine which patients are misclassified by these criteria, and for what reasons. Exclusion criteria designed to optimize the accepted DGE definition are proposed and evaluated.

We performed a retrospective review of prospectively collected data. We reviewed 357 consecutive patients undergoing PD by a single surgeon; included were 52 cases of ISGPS-defined DGE (14.6%). A detailed evaluation was conducted of cases using accepted and novel diagnostic criteria.

Of 52 ISGPS-defined DGE cases, 12 (23%) appeared not to represent genuine DGE on clinical review. Six required nasogastric tube placement for reoperation or management of emesis secondary to non-DGE conditions, 4 for reintubation without other evidence of DGE, and 2 remained NPO to treat non-DGE conditions. The proposed exclusion criteria use absence of gastric distention, passage of oral contrast, and presence of documented non-DGE conditions to determine genuine DGE. The incidence of true DGE was 11.2% in this cohort. The overall positive predictive value of the ISGPS criteria was 76.9%. Preoperative variables, DGE class, and incidence of disease-specific outcomes were similar with both definitions.

The ISGPS consensus guidelines promote a standardized, sensitive, and easily applicable definition of DGE, but may falsely classify DGE in approximately 23.1%. Introduction of the proposed exclusion criteria, which establish objective radiologic data as a component of the definition, could substantially limit this overestimation.