Efficacy of Pre-Operative Antimicrobial Prophylaxis in Patients Undergoing Pancreatoduodenectomy: A Multi-Center Retrospective Analysis

Abstract

Background:

The most common complications after pancreaticoduodenectomy (PD) are infectious, despite the standard use of cefazolin and metronidazole prophylaxis. Pre-operative biliary drainage (PBD) is a well-known risk factor for infectious complications. The objective was to identify the pathogens in intra-operative bile cultures in patients undergoing PD—with and without PBD—to determine the optimal antimicrobial prophylaxis regimen.

Patients and Methods:

Patients who underwent PD between 2009 and 2016 were identified retrospectively in three major teaching hospitals in The Netherlands. Organisms isolated from intra-operative bile cultures were studied. If pathogen coverage by standard prophylaxis was incomplete, the most appropriate alternative regimen was determined.

Results:

Of this large cohort of 352 patients, 56% underwent PBD and 44% did not. Positive bile cultures were found in 87.9% in the PBD group, compared with 31.8% in the non-PBD group. The micro-organisms isolated most commonly were Enterococcus, Streptococcus, and Klebsiella species. Cefazolin and metronidazole were appropriate in only 71% of patients. Adding gentamicin would provide complete coverage in 99% of PBD and 100% of non-PBD patients.

Conclusions:

Our data confirm that PBD prior to PD leads to microbial colonization and antibiotic resistance. To potentially prevent infectious complications, gentamicin may be added to the standard antimicrobial prophylaxis.

Pancreaticoduodenectomy (PD) is performed to resect a variety of malignant and benign diseases of the pancreas and peri-ampullary region [1]. The most common complications after PD are infectious and occur in approximately 30% of patients [2 –7]. Pre-operative biliary drainage (PBD) is performed in the majority of patients by endoscopic stenting of the common bile duct or by percutaneous transhepatic drainage [8 –10].

Several studies have shown that the microbial colonization of the biliary system after PBD is frequent and leads to an increased risk of post-operative surgical site infections [7,10 –15], in particular when bacterial flora resistant to antibiotic agents is involved [7,16 –18]. Although the use of PBD in patients with peri-ampullary cancer prior to PD remains debated, it cannot be avoided in patients presenting with cholangitis or severe symptomatic jaundice. Additionally, patients who have an indication for neoadjuvant chemotherapy for peri-ampullary cancer require a PBD in most cases. As evidence for the benefit of neoadjuvant chemotherapy is increasing, it is expected that more patients will need a PBD prior to surgery.

For these reasons it is important to assess whether the most widely used standard pre-operative antimicrobial prophylaxis in these patients is adequate and how the efficacy of prophylaxis can be optimized based on intra-operative bile cultures. The objective of this study was to identify the most common pathogens in intra-operative bile cultures in patients undergoing PD, who did or did not undergo PBD, to determine the optimal antimicrobial prophylaxis.

Patients and Methods

Study design and study population

We performed a retrospective multicenter cohort study in patients who underwent PD. Data were collected from patients admitted between 2009 and 2016 in three major teaching hospitals in The Netherlands: Isala Hospital (Zwolle), Tjongerschans Hospital (Heerenveen) and Amphia Hospital (Breda). For each hospital, the inclusion started as of the date when routine intra-operative bile culture collection was implemented. Patients for whom no intra-operative bile culture was performed were excluded. Patients were divided into two groups: those who had undergone PBD (drainage group) and those who had not undergone PBD (non-drainage group).

Data collection

In all participating centers the required data were collected from digital patient records, including basic patient characteristics such as gender, age, height, weight, body mass index (BMI) and American Society of Anesthesiologists (ASA) classification. We also registered the use of pre-operative PBD, the type of pre-operative antimicrobial prophylaxis, and the intra-operative bile culture results.

If pre-operative PBD was indicated as determined by the referring gastroenterologist, patients underwent PBD by internal drainage using 10F plastic stents or fully covered self-expandable metal stents. Occasionally, an additional drainage with stent exchange was performed if the patient developed signs pre-operatively of inadequate biliary drainage or cholangitis. Antibiotic prophylaxis was administered routinely15–30 minutes prior to the start of anesthesia. An intra-operative bile specimen was taken routinely immediately after transection of the common bile duct and processed by the hospital's microbiology laboratory. Bacterial culture, identification, and antimicrobial susceptibility testing of isolated micro-organisms were performed according to standard laboratory procedures. All organisms isolated from the bile cultures were registered, including their susceptibility pattern. For each individual patient we assessed whether the applied antimicrobial prophylactic therapy had been appropriate for the micro-organisms found. If at least one of the isolates was resistant to the administered antibiotic agents, the antibiotic coverage was rated incomplete. In that case the most appropriate alternative treatment was determined. Enterococci were excluded from this assessment because generally the pathogenic contribution of these bacteria in post-operative infectious complications is disputed [11]. The use of the obtained data was approved and judged to be exempt from ethical review procedures according to the Isala Hospital ethics review board.

Statistical analysis

Statistical analysis was performed using IBM SPSS Statistics for Windows version 24 (IBM Corp. Armonk, NY) at a significance of 5%. Demographic and clinical characteristics were described as frequencies and proportions or as median value with interquartile range. Statistical analysis of qualitative parameters was performed using the χ² test and the Fisher exact test, of quantitative parameters using the one-way analysis of variance (ANOVA) and Mann-Whitney tests for more than two and two independent samples, respectively.

Results

Patients

In total 373 patients underwent a PD for peri-ampullary cancer during the inclusion period in the three participating hospitals. In 352 patients intra-operative bile cultures were obtained (94%), of which 198 (56%) underwent PBD and 154 (44%) did not. The demographic and clinical characteristics of these patients are shown in Table 1. Fifty-eight percent of patients were male (n = 205), median age was 68 years (range, 61–74) and median body mass index (BMI) was 24 kg/m² (range, 22–27).

Table 1.

Patient Characteristics According to Pre-Operative Biliary Drainage or Not

	Drainage (n = 198)	Non-drainage (n = 154)	p
Gender (M/F)	119/79	86/68	0.422
Age (years, IQR)	69 (61–74)	67 (70–74)	0.246
Body mass index (Median kg/m², IQR)	24 (22–27)	24 (22–27)	0.797
ASA classification
I	12 (6.1)	27 (17.5)	0.006
II	133 (67.2)	97 (63.0)
III	52 (26.3)	29 (18.8)
IV	1 (0.5)	1 (0.6)

IQR = interguartile range; ASA = American Society of Anesthesiologists.

Microbiologic assessment

Three hundred fifty patients received pre-operative prophylactic intravenous antibiotic agents in accordance with the hospital's institutional guidelines on antimicrobial prophylaxis in surgery [15], of which 263 (75%) patients received 2 g of cefazolin and 0.5 g of metronidazole. In the remaining 25% of cases alternative antibiotic regimens were chosen based on patient factors (such as allergies or specific infections at the time of surgery) or individual surgeon's preferences. The different antibiotic prophylaxis regimens are listed in Table 2.

Table 2.

Antimicrobial Prophylaxis Used According to Pre-Operative Biliary Drainage or Not

	Overall (n = 352)	Drainage (n = 198)	Nondrainage (n = 154)	p
Cefazolin plus metronidazole	261 (74.1)	135 (68.2)	126 (81.8)	0.004
Piperacillin/tazobactam plus fluconazole	11 (<show [ensp]>3.1)	8 (4)	3 (1.9)	0.360
Piperacillin/tazobactam	9 (2.6)	7 (3.5)	2 (1.3)	0.309
Amoxicillin/clavulanic acid	1 (0.3)	1 (0.5)	0 (0.0)	1.000
Amoxicillin/clavulanic acid plus gentamicin	65 (18.5)	45 (22.7)	20 (13)	0.019
Cefazolin plus metronidazole plus amoxicillin/clavulanic acid	1 (0.3)	0 (0.0)	1 (0.6)	0.437
Gentamicin	2 (0.6)	1 (0.5)	1 (0.6)	1.000
No antimicrobial prophylaxis	2 (0.6)	1 (0.5)	1 (0.6)	1.000

Values in parentheses are percentages.

In total 224 (64%) intra-operative bile cultures were pathogen-positive, 161 (72%) of which were polymicrobial. Micro-organisms isolated from bile cultures are shown in Table 3. A total of 492 organisms were isolated. The micro-organisms most commonly isolated in the drainage and non-drainage groups were Enterococcus species (42.4% and 3.9%, respectively), Streptococcus species (31.8% and 9.1%), Klebsiella species (56% and 14%), and Enterobacter species (26.3% and 5.8%). Almost 90% of patients who underwent PBD (n = 175) had positive bile cultures (87.9%) compared with only 31.8% of non-drained patients (p < 0.001).

Table 3.

Micro-Organisms Isolated from Intra-Operative Bile Cultures in Patients Undergoing Pancreaticoduodenectomy

	Drainage (n = 198)	Nondrainage (n = 154)	p
Positive bile culture	175 (87.9)	49 (31.8)	<0.001
Polymicrobial culture	144 (82.3)	17 (34.7)	<0.001
Gram-negative bacilli
Escherichia coli	41 (20.7)	17 (11.0)	0.015
Klebsiella/Raoultella spp.	56 (28.3)	13 (8.4)	<0.001
Enterobacter spp.	52 (26.3)	9 (5.8)	<0.001
Citrobacter spp.	16 (8.1)	1 (0.6)	0.001
Proteus spp.	11 (5.6)	1 (0.6)	0.015
Morganella spp.	6 (3.0)	0 (0.0)	0.037
Hafnia alvei	13 (6.6)	2 (1.3)	0.016
Serratia spp.	4 (2.0)	0 (0.0)	NS
Lecelercia adecarboxylata	1 (0.5)	0 (0.0)	NS
Pseudomonas aeruginosa	1 (0.5)	0 (0.0)	NS
Aeromonas spp.	2 (1.0)	2 (1.3)	NS
Stenotrophomonas maltophilia	1 (0.5)	0 (0.0)	NS
Haemophilus spp.	2 (1.0)	2 (1.3)	NS
Gram-positive cocci
Staphylococcus spp.	9 (4.5)	0 (0.0)	0.006
Enterococcus spp.	84 (42.4)	6 (3.9)	<0.001
Streptococcus spp.	63 (31.8)	14 (9.1)	<0.001
Other	56 (28.8)	7 (4.5)	<0.001

Values in parentheses are percentages.

NS = not significant.

The appropriate antimicrobial prophylaxes according to the pathogens found in the intra-operative bile cultures are shown in Table 4. Cefazolin and metronidazole was appropriate in half of the patients (n = 110, 55.6%) whereas this prophylactic regime was far more appropriate in the non-drainage group (88.3%).The most appropriate alternative treatment would consist of adding gentamicin to the standard antimicrobial prophylaxis, which would provide complete coverage in 99% of PBD patients and 100% of non-PBD patients.

Table 4.

The Appropriate Antimicrobial Prophylaxis According to the Pathogens in the Intra-Operative Bile Cultures

	Overall (n = 352)	Drainage (n = 198)	Non-drainage (n = 154)	p
Cefazolin plus metronidazole
Appropriate	250 (71.0)	110 (55.6)	136 (88.3)	<0.001
Not appropriate	102 (29.0)	88 (44.4)	18 (11.7)
Cefazolin plus metronidazole plus gentamicin
Appropriate	350 (99.4)	196 (99.0)	154 (100)	0.506
Not appropriate	2 (0.6)	2 (1.0)	0 (0.0)
Amoxicillin/clavulanic acid plus gentamicin
Appropriate	349 (99.1)	195 (98.5)	154 (100)	0.260
Not appropriate	3 (0.9)	3 (1.5)	0

Values in parentheses are percentages.

Discussion

Principal findings

This study demonstrates that in patients undergoing PD 87.9% of patients had positive bile cultures in the PBD group compared with 31.8% in the non-PBD group. Pre-operative antibiotic prophylaxis with cefazolin and metronidazole was appropriate in only 55.6% of PBD patients and in 88.3% of non-PBD patients. Adding gentamicin to cefazolin and metronidazole was determined to provide complete coverage of pathogens in almost all patients after PBD (99%) and all non-PBD patients (100%).

One possible way to reduce the overall morbidity rate after PD is to reduce morbidity related to infectious events. With adequate antimicrobial regimens, adverse outcomes caused by serious infectious complications may sometimes be prevented. For adequate antibiotic prophylaxis and treatment, it is therefore essential to identify the patient-specific microbiologic spectrum of bile duct contamination as a source of possible infection.

Similar to other studies, we demonstrate that PBD increases biliary pathogen contamination. The utility of pre-operative PBD in patients with jaundice before undergoing PD is still debated. Sewnath et al. [19] state in their meta-analysis that PBD with current standards for patients with obstructive jaundice carries no benefit and should not be performed routinely. Infectious complications were shown to be increased after PBD [14,20]. However, PBD has the advantage of providing extra time to allow for further diagnostic investigations, referral to a tertiary center, or when performing early surgery is not feasible for practical reasons [20]. Also, PBD remains necessary in patients presenting with cholangitis or severe symptomatic jaundice or in patients receiving neoadjuvant systemic treatment, which is increasingly utilized. Given the frequently remaining indication for PBD, the question is how to optimize peri-operative antibiotic prophylaxis and treatment based on intra-operative bile cultures.

It has been well established by many studies that positive intra-operative bile culture is associated with a higher morbidity and mortality rate after PD [6,15,21 –25]. Many reports have found a strong correlation between organisms responsible for post-operative infectious complications and found in intra-operative bile cultures. The same organisms are often found both in bile and at the infection site [8,26 –29]. The bacteria isolated from abdominal abscesses were identical to those present in bile in 50%–100% of patients in previous studies [11,13,27,28]. Also an increased rate of SSI is described in PBD patients [14,26,29].

In our analysis, we found a higher rate of positive intra-operative bile cultures after PBD compared with patients without PBD (87.9% vs. 31.8%). Most other studies confirm a higher incidence of positive bile cultures in the PBD group in accordance with our data [6,8,9,15,19,22 –24,28 –31]. Recently published data by Scheufele et al. [10] described contamination of bile in 97.1% of PBD patients versus 18.6% in those without stenting.

Biliary tract infection can also occur as a consequence of biliary tract obstruction, inflammation, or stone formation. Therefore, most other studies also found a substantial number (range, 21%–55%) of positive bile duct cultures in patients without PBD [8,9,14,15,22,28,32]. Obstructive jaundice is therefore a risk factor for the development of bactibilia and subsequent cholangitis [8,33]. Stenting further increases this risk by providing an environment for the development of a bacterial biofilm [34].

In our data bile cultures were polymicrobial in 82.3% of PBD patients, whereas only 34.7% of patients without stent had more than one species cultured from their bile. Other studies showed comparable results [22,28,35]. A large variety of micro-organisms were identified in the bile cultures. The micro-organisms isolotated most commonly were Enterococcus species, Streptococcus species, Klebsiella species, and Enterobacter species, as reported in many other studies [15,22,26,28,29,31,35 –37].

In our study the widely used combination of cefazolin and metronidazole was appropriate in 78.4% (PBD 67.7% and non-PBD 89%) of patients. Sudo et al. [31] found that in patients with PBD prophylactic antibiotic agents (i.e., cefazolin or cefpirime) targeted approximately 30% of biliary micro-organisms effectively [31]. They therefore recommended the peri-operative use of therapeutic antibiotic agents. Mohammed et al. [38] concluded in 2014 that adapting antibiotic treatment to bile culture results after PD reduced post-operative wound infections and the severity of post-operative complications. As mentioned before prophylactic antibiotic agents should cover the most frequently cultured organisms. Cefazolin together with metronidazole was appropriate in 55.6% of drainage group patients and in 88.3% of non-drainage patients. We determined that the most appropriate alternative treatment would consist of adding gentamicin to the standard antimicrobial prophylaxis, resulting in complete coverage in 99% of PBD patients and 100% of non-PBD patients.

Strengths and limitations

Our retrospective multicenter analysis has certain strengths and limitations. In this largest series on this topic reported to date, our findings could be based on robust data allowing reliable identification of micro-organisms and their susceptibility patterns in our country. Because intra-operative bile cultures were obtained routinely in participating centers, we were able to sample almost all (94%) operated patients.

This study also has several limitations. Retrospective studies are prone to selection bias. Patients with more comorbidities may be offered PBD rather than primary surgery, although demographic analysis in this study showed no differences between the groups. Actual post-operative infection rates and associated micro-organisms are not addressed in this article. The relation between positive intra-operative bile cultures and post-operative infections has, however, been extensively shown by many reports, so this was not considered relevant extra data. This study focused on establishing the need for alternative broad spectrum prophylaxis regimens especially in patients who underwent PBD. It would be suggested to reconfirm our findings and recommendations in individual pancreatic cancer centers in other regions.

Conclusion

It has been well established that PBD in patients undergoing PD is associated with a higher risk of post-operative infectious complications. Our results from a large Dutch multicenter cohort shows that PBD had a notable influence on intra-operative bile cultures, including a higher rate of polymicrobial contamination and antibiotic resistance. Our findings suggest that in order to reduce the rate of post-operative infectious complications, gentamicin may be added to the widely used standard antimicrobial prophylaxis of cefazolin and metronidazole, which together would provide 99%–100% coverage of isolated micro-organisms from intra-operative bile cultures.

Footnotes

Acknowledgment

This article is not based on a previous communication to a society or meeting.

Author Disclosure Statement

No competing financial interests exist.

References

Yeo

, Cameron

, Sohn

, et al. Six hundred fifty consecutive pancreaticoduodenectomies in the 1990s: Pathology, complications, and outcomes. Ann Surg, 1997; 226:248–257.

Gouma

, van Geenen

RCI

, van Gulik

, et al. Rates of complications and death after pancreaticoduodenectomy: Risk factors and the impact of hospital volume. Ann Surg, 2000; 232:786–795.

Topal

, van der Sande

, Fieuws

, Penninckx

. Effect on centralization of pancreaticoduodenectomy on nationwide hospital mortality and length of stay. BJS, 2007; 94:1377–1381.

Van Heek

, Kuhlmann

KFD

, Scholten

, et al. Hospital volume and mortality after pancreatic resection: A systematic review and an evaluation of intervention in The Netherlands. Ann Surg, 2005; 242:781–790.

Povoski

, Karpeh

Jr , Conlon

, et al. Association of preoperative biliary drainage with postoperative outcome following pancreaticoduodenectomy. Ann Surg, 1999; 230:131–142.

Pisters

, Hudec

, Hess

, et al. Effect of preoperative biliary decompression on pancreaticoduodenectomy associated morbidity in 300 consecutive patients. Ann Surg, 2001; 234:47–55.

DeOliveira

, Winter

, Schafer

, et al. Assessment of complications after pancreatic surgery: A novel grading system applied to 633 patients undergoing pancreaticoduodenectomy. Ann Surg, 2006; 244:931–937.

Morris-Stiff

, Tamijmarane

, Tan

, et al. Pre-operative stenting is associated with a higher prevalence of post-operative complications following pancreatoduodenectomy. Int J Surg, 2011; 9:145–149.

Karsten

, Allema

, Reinders

, et al. Preoperative biliary drainage, colonisation of bile and postoperative complications in patients with tumours of the pancreatic head: A retrospective analysis of 241 consecutive patients. Eu J Surg, 1996; 162:881–888.

10.

Scheufele

, Aichinger

, Jäger

, et al. Effect of preoperative biliary drainage on bacterial flora in bile of patients with periampullary cancer. BJS, 2017; 104:e182–e188.

11.

van Kasteren

, Gyssens

, Kullberg

, et al. Optimizing antibiotics policy in the Netherlands. V. SWAB guidelines for perioperative antibiotic prophylaxis. Foundation Antibiotics Policy Team. Ned Tijdschr Geneeskd, 2000; 144:2049–2055.

12.

Sganga

. New perspectives in antibiotic prophylaxis for intra-abdominal surgery. J Hosp Infect, 2001; 50:S17–S21.

13.

Davies

, Davies

. Origins and evolution of antibiotic resistance. Microbiol. Mol Biol Rev, 2010; 74:417–433.

14.

Povoski

, Karpeh

Jr , Conlon

, et al. Preoperative biliary drainage: Impact on intraoperative bile cultures and infectious morbidity and mortality after pancreaticoduodenectomy. J Gastrointest Surg, 1999; 3:496–505.

15.

Jagannath

, Dhir

, Shrikhande

, et al. Effect of preoperative biliary stenting on immediate outcome after pancreaticoduodenectomy. Br J Surg, 2005; 92:356–361.

16.

Cortes

, Sauvanet

, Bert

, et al. Effect of bile contamination on immediate outcomes after pancreaticoduodenectomy for tumor. J Am Coll Surg, 2006; 202:93–99.

17.

Limongelli

, Pai

, Bansi

, et al. Correlation between preoperative biliary drainage, bile duct contamination, and postoperative outcomes for pancreatic surgery. Surgery, 2007; 142:313–318.

18.

Mezhir

, Brennan

, Baser

, et al. A matched casecontrol study of preoperative biliary drainage in patients with pancreatic adenocarcinoma: Routine drainage is not justified. J Gastrointest Surg, 2009; 13:2163–2169.

19.

Sewnath

, Karsten

, Prins

, et al. A meta-analysis on the efficacy of preoperative biliary drainage for tumors causing obstructive jaundice. Ann Surg, 2002; 236:17–27.

20.

Lillemoe

. Preoperative biliary drainage an surgical outcome. Ann Surg, 1999; 230:143–144.

21.

Sivaraj

, Vimalraj

, Saravanaboopathy

, et al. Is bactibilia a predictor of poor outcome of pancreaticoduodenectomy?. Hepatobiliary Pancreat Dis Int, 2010; 9:65–68.

22.

Howard

, Yu

, Greene

, et al. Influence of bactibilia after preoperative biliary stenting on postoperative infectious complications. J Gastrointest Surg, 2006; 10:523–531.

23.

Sohn

, Yeo

, Cameron

, et al. Do preoperative biliary stents increase postpancreaticoduodenectomy complications?. J Gastroinest Surg, 2000; 4:258–268.

24.

Hodul

, Creech

, Pickleman

, Aranha

. The effect of preoperative biliary stenting on postoperative complications after pancreaticoduodenectomy. Am J Surg, 2003; 186:420–425.

25.

Van der Gaag

, Rauws

, van Eijck

CHL

, et al. Preoperative biliary drainage for cancer of the head of the pancreas. N Engl J Med, 2010; 362:129–137.

26.

Lermite

, Pessaux

, Teyssedou

, et al. Effect of preoperative endoscopic biliary drainage on infectious morbidity after pancreatoduodenectomy: A case-control study. Am J Surg, 2008; 195:442–446.

27.

Sahora

, Morales-Oyarvide

, Ferrone

, et al. Preoperative biliary drainage does not increase major complications in pancreaticoduodenectomy: A large single center experience from the Massachusetts General Hospital. J Hepatobiliary Pancreat Sci, 2016; 23:181–187.

28.

Herzog

, Belyaev

, Akkuzu

, et al. The impact of bile duct cultures on surgical site infections in pancreatic surgery. Surg Infect, 2015; 16:443–449.

29.

Augenstein

, Reuter

, Bower

, et al. Bile cultures: A guide to infectious complications after pancreaticoduodenectomy. J Surg Oncol, 2010; 102:478–481.

30.

Martignoni

, Wagner

, Krähenbühl

, et al. Effect of preoperative biliary drainage on surgical outcome after pancreatoduodenectomy. Am J Surg, 2001; 181:52–59.

31.

Sudo

, Murakami

, Uemura

, et al. Specific antibiotic prophylaxis based on bile cultures is required to prevent postoperative infectious complications in pancreatoduodenectomy patients who have undergone preoperative biliary drainage. Wolrd J Surg, 2007; 31:2230–2235.

32.

Rao

, Varghese

, Blake

, Theodossi

. Endoscopic biliary stenting in a district general hospital. Gut, 1995; 37:279–283.

33.

Dixon

, Armstrong

, Duffy

, Davies

. Factors affecting morbidity and mortality after surgery for obstructive jaundice: A review of 373 patients. Gut, 1984; 25:104.

34.

Groen

, Out

, Huibregtse

. Characterisation of the content of occluded biliary endoprothesis. Endoscopy, 1987; 19:57–59.

35.

Grizas

, Stakyte

, Kincius

, et al. Etiology of bile infection and its association with postoperative complications following pancreatoduodenectomy. Medicina, 2005; 41:386–391.

36.

Sugiura

, Ueska

, Ohmagari

, et al. Risk factor of surgical site infection after pancreaticoduodenectomy. World J Surg, 2010; 36:2888–2894.

37.

Rerknimitr

, Fogel

, Kalayci

, et al. Microbiology of bile in patients with cholangitis or cholestatis with and without plastic biliary endoprosthesis. Gastrointest Endosc, 2002; 56:885–889.

38.

Mohammed

, Evans

, van Buren

, et al. Treatment of bacteriobilia decreases wound infection rates after pancreaticoduodenectomy. HPB (Oxford), 2014; 16:592–598.