Back to 2011 Program

Objective: To evaluate multiple variables hypothesized to contribute to delayed gastric emptying(DGE) development in patients undergoing pancreaticoduodenectomy(PD), thereby identifying high-risk patients for DGE and allowing optimization of therapy.
Design: Retrospective cohort
Setting: Academic tertiary center
Patients: 174 patients underwent pancreaticoduodenectomy between 1/03 and 4/10 by a single surgical oncologist. The most common indication for PD were benign and malignant neoplasms of the pancreas. The average patient age was 62.5 years. Preoperative medical history was assessed for correlation with DGE.
Interventions: N/A
Main Outcome Measures: The primary outcome was development and severity of DGE as defined by the International Study Group for Pancreatic Surgery.
Results: Overall incidence of DGE was 13.9% with 3.0% having Grade C(most severe) disease. Comorbidities, including diabetes mellitus, did not correlate with DGE. Retrocolic gastrojejunostomy correlated with increased DGE(p<0.01), as did placement of intra-abdominal drains(p<0.01) while pylorus-preserving procedures were not correlated when compared to classic PD(p=0.77). Presence of perioperative complications such as abscess or fistula greatly increased DGE incidence(p<0.01). Patients without postoperative nasogastric decompression did not have a higher incidence of DGE(p=0.253).
Conclusions: DGE is a frequent complication of PD that is most closely associated with procedural variables such as anastomotic location. Incidence of DGE does not correlate with patient comorbidities. Furthermore, decompression of the stomach post-operatively does not reduce DGE incidence, despite its routine employment. Finally, the milieu of the operative bed is a key factor in DGE development with tight correlation between DGE and other complications. These data suggest an optimal approach to care during and after PD may minimize DGE incidence.