Josiah Brown Poster Abstract


Sergio J. Marquez
Mark D. Girgis, MD
Evan D. Adams, MD
Impact of Antibiotic Choice on Surgical Site Infections in Patients Following Pancreaticoduodenectomy

Background:Surgical site infections (SSIs) are one of the leading causes of morbidity following pancreaticoduodenectomy (PD), leading to prolonged hospital stay, invasive procedures (stenting/drainage), additional imaging, and higher mortality. Perioperative antibiotic management has been demonstrated to decrease the risk of infectious complications, especially with a bile culture characterization. Despite the various documented benefits of preoperative biliary stenting, they have been shown to increase the risk of SSIs. Due to the increase in stenting for cancer, recent focus has shifted towards the choice of antibiotic prophylaxis in the perioperative management of SSIs following PD. With limited research comparing the effectiveness of different antibiotic regimen in pancreatic surgery, especially in the setting of chemotherapy, our analysis seeks to elucidate the impact of antibiotic choice on infectious perioperative outcomes in patients following PD.

Methods: A retrospective analysis was performed on 3,252 patients undergoing a PD from the National Surgical Quality Improvement Program database from 2014-2015.

Outcomes and Measures: Primary outcome is a combined outcome of infections within 30 days of surgery (superficial skin infection, organ space infection, deep incisional infection, pancreatic fistula, Clostridium difficile infection, sepsis, urinary tract infection, and pneumonia). Secondary outcome is chemotherapy within 90 days of operation.

Results: With respect to our chi-square analyses of infectious outcomes, the stented group showed a lower rate of combined outcomes when broad spectrum antibiotics were administered (40% vs. 39% vs. 34%, p=0.065), compared to the non-stented group (37% vs. 35% vs. 38%, p=0.63). The stented group receiving broad spectrum antibiotics had the lowest rate of superficial skin infections (12% vs. 10% vs. 7.3%, p=0.03), compared to the no significant difference in the non-stented group (5.0% vs. 4.6% vs. 5.2%, p = 0.92). Moreover, stented patients demonstrated a lower rate of organ space infections with broad spectrum management (18% vs. 16% vs. 12%, p = 0.004). This was not observed in the non-stented patients (19% vs. 17% vs. 19%, p=0.66). The multivariate logistic regression analysis of the combined outcome variable revealed broad spectrum antibiotics were associated with the lowest rates of infectious complications (odds ratio[OR], 0.82; 95% CI 0.69-0.98; P = 0.03), compared to 2nd/3rd  generation cephalosporins (OR 0.92; 95% CI 0.76-1.11; P = .386). Stents were associated with increased rates of infectious outcomes (OR 1.31; 95% CI 1.11-1.55; P = 0.002). Specifically, broad spectrum antibiotics were associated with lower rates of superficial skin infections (OR 0.73; 95% CI 0.53-1.00; P = 0.05), this was not observed in 2nd/3rd generation cephalosporins (OR 0.88; 95% CI 0.63-1.24; P = 0.47). Organ space infection showed a similar trend with broad spectrum antibiotic treatment (OR 0.72; 95% CI 0.57-0.91; P < 0.01); 2nd/3rd generation cephalosporin treatment was not a significant factor for the reduction of organ space infections (OR 0.83; 95% CI 0.64-1.06; P = .14). Our analysis on choice of antibiotics and successful chemotherapy within 90 days after PD showed that both in stented and non-stented patients, the rate of chemotherapy within 90 days was higher in the 2nd/3rd generation cephalosporin and broad spectrum antibiotic group (17% vs. 21% vs. 21%, P < 0.01). This was most pronounced in the stented group receiving broad spectrum management (21% vs. 26% vs 27%, p=0.034). The multivariate regression analyses of chemotherapy within 90 days as an outcome revealed broad spectrum antibiotic treatment to be associated with an increased rate of successful receipt of chemotherapy (OR 1.35; 95% CI 1.08-1.70; P = 0.01).



  • Patients undergoing preoperative biliary stenting are at higher risk for developing postoperative infectious complications, specifically skin and organ space infections.
  • Broad spectrum antibiotics are protective against infectious complications, especially skin and organ space infections in stented patients.
  • The use of 2nd/3rd generation cephalosporins or broad spectrum antibiotics may be used in non-stented patients.
  • 1st generation cephalosporins are not recommended in stented patients
  • Broad spectrum coverage should be considered in stented patients undergoing PD, as it may decrease risk of infectious outcomes and promote candidacy for adjuvant chemotherapy, thus improving prognosis.