Biliary Tract Instrumentations Prior to Elective Cholecystectomy: Effect on Biliary Microbiome

Abstract

Background:

Calculus biliary disease is a common condition that requires invasive procedures in complicated cases. The effect of biliary instrumentation on the biliary microbiome and its impact on surgical complications after elective cholecystectomy remains unclear. This study aimed to assess the impact of prior biliary instrumentation on the biliary microbiome, as well as on the clinical outcomes of cholecystectomy.

Patients and Methods:

This retrospective study included all patients who underwent elective cholecystectomy for calculus biliary disease between 2015 and 2020 in a single medical center. Data regarding biliary instrumentation prior to cholecystectomy, biliary cultures obtained during cholecystectomy, and clinical outcomes were collected. A comparison between patients with and without prior instrumentation was performed with regard to biliary cultures and clinical outcomes.

Results:

Of the 508 patients studied, 109 patients underwent biliary instrumentation prior to cholecystectomy. Patients with prior instrumentation were older and more likely to be men (p < 0.0001). Prior instrumentation was also associated with higher rates of conversion to open surgery (p < 0.0001). Positive biliary cultures and polymicrobial growth were both more common among patients with prior instrumentation (p < 0.0001). Prior instrumentation was associated with longer length of hospital stay, as well as higher rates of perioperative complications and surgical site infection (p < 0.0001).

Conclusions:

Prior instrumentation was associated with poorer clinical outcomes and affected the biliary microbiome. The different results of biliary cultures in these patients may suggest that an alternative empiric antibiotic regimen should be considered when treating patients with biliary instrumentation.

Gallstone disease is a common condition and a frequent cause of hospital admissions [1]. Clinical manifestations of symptomatic disease include several different entities, such as biliary colic, acute cholecystitis, choledocholithiasis with or without obstructive jaundice, cholangitis, and biliary pancreatitis [2]. When indicated, antibiotic treatment for gallstone disease is usually empiric, and its choice is guided by the common pathogens that inhabit the biliary tract [3]. The most frequent biliary tract isolates are Escherichia coli, which is found in 25%–50% of cases, and Klebsiella, Pseudomonas, and Enterobacter species, with reports of different prevalence in the current literature ranging from 0.5% to 20% [4,5]. The standard treatment for symptomatic calculous biliary disease is cholecystectomy. However, patients with different presentations of gallstone disease may require additional invasive interventions prior to gallbladder resections, such as cholecystostomy for acute cholecystitis in patients who are medically unfit for surgery, endoscopic retrograde cholangiopancreatography (ERCP) for common bile duct (CBD) clearance, and percutaneous transhepatic biliary drainage (PTBD) for biliary tree decompression in cases of severe cholangitis [5].

An early study conducted in 1992 by Sand et al. [6] consisted of 32 patients and suggested that contamination occurred during ERCP, which may result in long-term bactibilia. Moreover, the effect of different invasive manipulations, either endoscopic or percutaneous, on the biliary microbiome has not been extensively studied. This knowledge may provide additional insights into the guidance of empiric antibiotic therapy in selected cases.

The primary aim of this study was to characterize biliary colonization by micro-organisms from intra-operative biliary cultures from patients who underwent elective cholecystectomy with and without various biliary instrumentations prior to surgery. The secondary aim was to assess the clinical outcomes of patients with and without prior instrumentation with regard to the presence of bactibilia.

Materials and Methods

The study was approved by the Institutional Review Board of the Hillel Yaffe Medical Center (protocol number 0104-20). All patients who underwent elective cholecystectomy for symptomatic calculous gallbladder disease from January 1, 2015 to December 31, 2020 were included in this study. Patients who underwent urgent cholecystectomy during their hospitalization were excluded from the study. Medical records were obtained from the electronic database system of the hospital using ICD-9 codes. The data were individually collected for all patients, and include demography, previous medical history, pre-existing liver diseases, type of biliary events, instrumentation prior to cholecystectomy, surgical procedure, results of biliary cultures obtained during cholecystectomy, and clinical outcomes. The clinical outcome measures included the length of hospital stay after elective cholecystectomy, peri-operative complications, and surgical site infection. Peri-operative complications were recorded according to the Clavien-Dindo score [7].

Symptomatic calculous gallbladder disease was defined as one of the following conditions: biliary colic, acute cholecystitis, cholangitis, choledocholithiasis, and biliary pancreatitis. In our institution, patients with an acute biliary event are invited for an elective cholecystectomy four to six weeks after their discharge from the hospital. According to our institution policy, all patients who undergo elective cholecystectomy receive peri-operative antibiotic agents. The standard regimen in our hospital is 2 g of cefazolin within 30 minutes of operation. Patients without evidence of acute cholecystitis during surgery and an uneventful intra-operative procedure do not receive antibiotic treatment post-operatively. Patients with evidence of cholecystitis during surgery, and/or intra-operative bile spillage receive post-operative antibiotic treatment, according to the clinical judgment of the attending physician. Bile cultures are obtained routinely during surgery by the scrub nurse. After resection of the gallbladder, its wall is incised and a swab is taken from the bile inside the gallbladder, under sterile conditions. Cultures are routinely obtained, regardless of the indication for surgery and underlying condition. Cultures are obtained as a quality control measure to treat and prevent postoperative infections by directing antibiotic therapy to specific pathogens. Patients without documented intraoperative bile cultures were excluded from this study.

All eligible patients were divided into two groups: group including patients with no biliary instrumentation prior to surgery, and group B, which included patients with any of the aforementioned types of biliary instrumentation. Biliary instrumentation was defined as percutaneous cholecystostomy, PTBD, and ERCP with and without stent insertion.

Within group B, the patients were further subdivided according to the type of pre-operative instrumentation, namely, cholecystostomy, PTBD, and ERCP. Comparisons between the patients with and without prior biliary instrumentation were performed for baseline and studied variables. For continuous variables, a Shapiro-Wilk test was performed to determine whether the variables were normally distributed. Accordingly, for normally distributed continuous data, an independent t-test was performed. For continuous variables with a non-normal distribution, the Mann-Whitney U test was performed. For categorical variables, Pearson χ² or Fisher exact test was performed as needed. Continuous variables are presented as median (range). Categorical variables were expressed as percentages (numbers). Following the univariate analysis for clinical outcome, a multivariable logistic regression model for peri-operative complications was constructed to determine risk factors. Statistical significance was considered as a two-tailed p value of ≤0.05. Because all data in the study were based on a retrospective chart review, it was assumed that all missing data were missing at random, thus, a pairwise deletion method was chosen whenever data were missing.

Results

Of the 508 patients, 399 patients (78.5%) did not undergo biliary instrumentation before cholecystectomy (group A) and 109 patients (21.5%) underwent prior instrumentation (group B). Among the patients in group B, 25 (22.9%) patients underwent cholecystostomy, 79 patients (72.5%) underwent ERCP with or without stent insertion, and five patients (4.6%) underwent both cholecystostomy and ERCP prior to cholecystectomy. None of the patients underwent PTBD.

The demographic data, baseline medical conditions, and types of biliary events in both groups are summarized in Table 1. Overall, the median age was 43 (range, 14–93) and 33.5% (170) were male. As shown in Table 1, no differences between the groups were observed in medical history, including diabetes mellitus, obesity, fatty liver disease, and history of viral hepatitis. The patients in group B were older and more commonly male (p < 0.0001). In total, 491 (96.6%) patients underwent laparoscopic cholecystectomy, and 17 (3.5%) underwent open surgery. The conversion to open surgery was five-fold higher in group B (9.9%) than in group A (1.8%) (p < 0.0001).

Table 1.

Patient Characteristics

	Group A No prior biliary instrumentation (n = 399)	Group B Prior biliary instrumentation (n = 109)	p
Age, median (range)	43 (19–82)	60 (14–90)	<0.0001
Male gender	29.2% (116)	49.5% (54)	<0.0001
Medical background
Diabetes mellitus	19.6% (78)	23.9% (26)	NS
Obesity	15.6% (62)	18.4% (20)	NS
Fatty liver	4.5% (18)	5.0% (4)	NS
Viral hepatitis	0.5% (2)	1.8% (2)	NS
Biliary event
Gallbladder event	88.9% (354)	67.9% (74)	<0.0001
Bile duct event	11.1% (44)	32.1% (35)	<0.0001
Conversion to open surgery	1.8% (7)	9.9% (10)	<0.0001

Gallbladder event = biliary colic/cholecystitis; bile duct event = choledocholithiasis/cholangitis/pancreatitis.

NS = not significant.

The biliary culture results are presented in Table 2. Positive cultures and polymicrobial growth were more common in group B (p < 0.0001). Of the 109 patients with prior biliary instrumentations, 77 had positive cultures (70.6%) and 32 had negative cultures (29.4%). Of the 399 patients without prior biliary interventions, 62 had positive cultures (15.5%) and 337 patients had negative cultures (84.5%).

Table 2.

Biliary Cultures

	Group A No prior biliary instrumentation (n = 399)	Group B Prior biliary instrumentation (n = 109)	p
Positive culture	15.5% (62)	70.6% (77)	<0.0001
Negative culture	84.5% (337)	29.4% (32)	<0.0001
Polymicrobial	17.7% (11)	50.7% (39)	<0.0001
Escherichia coli	25.8% (16)	44.2% (34)	NS
Enterococcus	20.1% (13)	39.0% (30)	NS
Clostridium	3.2% (2)	7.8% (6)	NS
Klebsiella	22.6% (14)	28.6% (22)	NS
Enterobacter	4.9% (3)	15.6% (12)	NS

NS = not significant.

The most common micro-organisms in group A were Escherichia coli (25.8%), followed by Klebsiella (22.6%), Enterococcus (20.1%), Enterobacter (4.9%), and Clostridium (3.2%). The most common micro-organisms in group B were Escherichia coli (44.2%), followed by Enterococcus (39.0%), Klebsiella (28.6%), Enterobacter (15.6%), and Clostridium (7.8%).

A comparison of the bacterial cultures in patients with different types of biliary instrumentations who underwent cholecystostomy alone or underwent ERCP prior to cholecystostomy is presented in Table 3. No differences in the rates of positive cultures and polymicrobial growth were found between the groups. The most common micro-organism in the patients who underwent cholecystostomy prior to cholecystectomy was Escherichia coli (96.3%). Enterococcus was the most common micro-organism in the patients who underwent ERCP prior to cholecystectomy. Klebsiella was more common in patients who underwent ERCP (34%) than in those who underwent cholecystostomy alone (5.3%; p = 0.020). A subanalysis was performed, comparing results of biliary cultures in patients with prior ERCP versus patients without prior intervention and patients with prior cholecystostomy versus patients without prior interventions (Tables 4 and 5, respectively). Positive growth was more common among patients with prior ERCP as well as prior cholecystostomy, compared with patients without prior interventions (p = 0.0003, p = 0.03, respectively). Similarly, polymicrobial growth was significantly more common among patients with prior ERCP as well as prior cholecystostomy, compared to patients without prior interventions (p < 0.0001).

Table 3.

Comparison of Culture: Cholecystostomy versus ERCP

	Cholecystostomy (n = 25)	ERCP (n = 79)	p
Positive culture	76.0% (19)	67.1% (53)	NS
Negative culture	24.0% (6)	22.9% (26)	NS
Polymicrobial	52.6% (10)	49.1% (26)	NS
Escherichia coli	96.3% (12)	35.9% (19)	0.04
Enterococcus	36.8% (7)	37.7% (20)	NS
Clostridium	0% (0)	9.4% (5)	NS
Klebsiella	5.3% (1)	34.0% (18)	0.02
Enterobacter	21.1% (4)	15.1% (8)	NS

ERCP = endoscopic retrograde cholangiopancreatography; NS = not significant.

Table 4.

Comparison of Culture: ERCP versus No Prior Biliary Instrumentation

	No prior biliary instrumentation (n = 25)	ERCP (n = 79)	p
Positive growth	15.5% (62)	67.1% (53)	0.0003
Polymicrobial growth	17.7% (11)	49.1% (26)	<0.0001
Escherichia coli	25.8% (16)	35.9% (19)	NS
Enterococcus	20.1% (13)	37.7% (20)	0.04
Clostridium	3.2% (2)	9.4% (5)	NS
Klebsiella	22.6% (14)	34.0% (18)	NS
Enterobacter	4.9% (3)	15.1% (8)	NS

NS = not significant.

Table 5.

Comparison of Culture: Cholecystostomy versus No Prior Biliary Instrumentation

	No prior biliary instrumentation (n = 25)	Cholecystostomy (n = 25)	p
Positive growth	15.5% (62)	76.0% (19)	0.03
Polymicrobial growth	17.7% (11)	52.6% (10)	<0.0001
Escherichi. coli	25.8% (16)	96.3% (12)	0.007
Enterococcus	20.1% (13)	36.8% (7)	NS
Clostridium	3.2% (2)	0% (0)	NS
Klebsiella	22.6% (14)	5.3% (1)	NS
Enterobacter	4.9% (3)	21.1% (4)	NS

NS = not significant.

A comparison of the clinical outcome measures between the patients in groups A and B is presented in Table 6. The total median length of stay was three days, ranging from 1 to 29 days. The length of stay was longer in group B (p < 0.0001). The total rate of peri-operative complications was 6.8% (34 patients). There were no mortality or degree 4 complications in either group. All the complications were classified as Clavien-Dindo grade 1–3 and did not require interventions under general anesthesia. The total rate of surgical site infection (SSI) was 30.1% (25 patients). The rates of peri-operative complications, as well as SSI, were more common among patients in group B (p < 0.0001). The subgroup analysis of the clinical outcomes of the patients in group B, who underwent cholecystostomy prior to surgery and those who underwent ERCP prior to surgery are presented in Table 5. Although the patients who underwent cholecystostomy prior to cholecystectomy exhibited a longer post-operative stay than those who ERCP prior to cholecystostomy (6.5 days vs. 4 days), they did not experience statistically higher rates of complications.

Table 6.

Clinical Outcomes

	Group A No prior biliary instrumentation (n = 399)	Group B Prior biliary instrumentation (n = 109)	p
LOS, median (range)	3 (1–29)	4 (2–24)	<0.0001
Peri-operative complications	3.5%	18.4%	<0.0001
SSI (n/total)	14.8% (8/54)	58.6% (17/29)	<0.0001

LOS = length of hospital stay; SSI = surgical site infection.

Because group B included higher rates of male gender, a comparison of clinical outcomes and rates of conversion to open cholecystectomy between male and female patients was performed (Table 7). Peri-operative complications and SSI were more common among male patients (p = 0.004 and p = 0.02, respectively). Length of stay was also longer among male patients (p = 0.003). No difference in rates of conversion from laparoscopic to open cholecystectomy was found. A multivariable analysis to determine risk factors for peri-operative complications was performed. As presented in Table 8, male gender was not associated with peri-operative complications, and biliary instrumentation prior to cholecystectomy remained associated with peri-operative complications (p < 0.0001).

Table 7.

Comparison of Clinical Outcome and Conversion to Open Cholecystectomy: Male versus Female Patients

	Males (n = 170)	Females (n = 337)	p
LOS, median (range)	3 (2–29)	3 (1–20)	0.003
Peri-operative complications	11.2% (19)	4.5% (15)	0.004
SSI	44.1% (15/34)	20.4% (10/49)	0.02
Conversion to open surgery	5.3% (9)	2.4% (8)	0.08

LOS = length of hospital stay; SSI = surgical site infection.

Table 8.

Multivariate Analysis of Risk Factors for Peri-Operative Complications

		95% CI
	OR	Lower	Upper	p
Male gender	2.0306	0.9716	4.2438	0.06
Instrumentation	5.8031	2.7736	12.1415	<0.0001

OR = odds ratio; CI = confidence interval.

Discussion

This study aimed to assess the effect of biliary instrumentation, including cholecystostomy and ERCP, on the biliary microbiome by comparing the patients who underwent elective cholecystectomy with and without prior instrumentation.

As expected, the patients who underwent instrumentation were admitted because of cholecystitis, choledocholithiasis, cholangitis, and/or biliary pancreatitis.

In our study, conversion to open cholecystectomy was more common among patients who underwent biliary instrumentation prior to surgery. Such association between prior instrumentation and higher rates of conversion has been previously reported by Domingez et al. [8], Ercan et al. [9], and Ischizaki et al. [10]. Nair et al. [11] investigated the impact of CBD stents on the outcome of elective cholecystectomy and found that the incidence of conversion was higher in patients with stents. This result may be explained by a foreign body reaction, including local inflammation and fibrosis in extrahepatic bile ducts, as has been shown in several studies [12,13]. Nevertheless, previous studies have also reported that an age >60 years and male gender are predictors of conversion to open cholecystectomy [14 –16].

In our study, the patients in the instrumentation group were older and more likely to be male, which may be responsible, in part, for the higher rates of conversion to an open procedure. Thus, a comparison of clinical outcomes and conversion to open cholecystectomy between males and females was conducted. This analysis showed male patients had higher rates of peri-operative complications and SSI (0.004 and 0.02, respectively), as well as longer hospital stay (p = 0.003). Following the multivariable analysis, pre-operative biliary instrumentation remained associated with higher rates of peri-operative complications and male gender did not. This result suggests that higher rates of peri-operative complications were, indeed, related to prior biliary instrumentation. When analyzing the results of the biliary cultures obtained during cholecystectomy, positive cultures were found to be more common in patients with prior instrumentation. Similarly, studies that have investigated the impact of stents and other foreign bodies in other anatomic locations, including the pulmonary and urinary systems, have also found that the presence of a foreign body changes the microbiome and may create a biofilm that must be addressed [17,18].

As previously reported in several studies, Escherichia coli was found to be the most common bacteria in both study groups and the second most common bacteria was Klebsiella [4,5]. Contrary to previous reports, the second most common bacteria found in this study were Enterococcus species in both groups. Klebsiella growth, which is considered a frequent finding in biliary cultures [5], was the third most common bacterium found in this study. Cultures were positive for Klebsiella in only 22.6% and 28.6% of the patients without and with prior instrumentation, respectively. Notably, polymicrobial growth was also more common among the patients who underwent biliary instrumentation. Comparing the cultures obtained from patients with prior ERCP, no differences in rates of polymicrobial growth or growth of specific pathogens were found. However, in other studies, a controversy exists regarding monomicrobial growth, which is predominantly Escherichia coli [19,20].

Evaluation of clinical outcomes in this study demonstrated that the length of hospital stay was significantly longer among the patients in group B. In addition, the rates of peri-operative complications and SSIs were higher in the patients with biliary instrumentation. Indeed, a longer hospital stay has been previously reported to be associated with ERCP and stent insertion preceding cholecystectomy [11]. Notably, when subanalyzing the results of clinical outcomes in group B, no differences in the rates of peri-operative complications or SSI were found between the patients who underwent cholecystostomy or ERCP. Hence, the higher rates of SSI and peri-operative complications that were found among the patients who underwent biliary instrumentation prior to cholecystectomy did not result specifically from cholecystostomy or ERCP; however, both were similarly associated with these measures of poorer clinical outcomes. Patients who underwent cholecystostomy had a longer hospital stay, compared to patients who underwent ERCP (p = 0.014).

Following the multivariate analysis, pre-operative biliary instrumentation remained associated with higher rates of peri-operative complications and male gender did not. This result suggests that higher rates of peri-operative complications were, indeed, related to prior biliary instrumentation. Higher post-operative infection rates caused by biliary tract contamination by pre-operative instrumentation in complicated cases, as well as the difference in biliary cultures, may suggest a need for culturing bile during cholecystectomies in order to be able to adjust specific antibiotic therapy.

This study has several limitations that must be acknowledged, including the known limitations of a retrospective design as well as a single-center study. In addition, patients needing cholecystostomy may be sicker patients to begin with. Thus, these patients are maybe subject to poorer clinical outcomes even though the cholecystectomy was performed electively, at least six weeks after the hospitalization for the acute biliary event. Another study limitation is that group B included all the patients who underwent either cholecystostomy, ERCP (with or without stent insertion), or both. These procedures are inherently different, and their effect on the biliary microbiome must be considered when interpreting the results of this study.

Conclusions

Biliary instrumentation prior to cholecystectomy was associated with higher rates of positive gallbladder cultures and polymicrobial growth compared with patients without prior interventions. This result may suggest that an alternative empiric antibiotic regimen should be considered in these patients. In addition, biliary instrumentation was found to be associated with longer hospital stays, and higher rates of perioperative complications and surgical site infections. These differences were more significant with cholecystostomy prior to cholecystectomy than with ERCP. Furthermore, larger prospective studies are needed to validate our results and investigate the effects of biliary instrumentation on the biliary microbiome in order to adjust the optimal antibiotic regimen. Collection of data regarding common biliary microbiome in different institutions may be helpful for development of center-specific protocol for peri-operative prophylactic antibiotic regimen, specifically in patients with prior biliary intervention. In addition, future studies may further explore the differences in the clinical outcomes and biliary microbiomes in patients who underwent ERCP with and without stent insertion, as well as in patients who underwent cholecystostomy

Footnotes

Funding Information

This research did not receive any grant or funding from funding agencies in the public, commercial, or not-for-profit sectors.

Author Disclosure Statement

We confirm that there are no known conflicts of interest associated with this publication, and there has been no financial support for this work.

References

Shabanzadeh

. Incidence of gallstone disease and complications. Curr Opin Gastroenterol, 2018; 34:81–89.

Attasaranya

, Fogel

, Lehman

. Choledocholithiasis, ascending cholangitis, and gallstone pancreatitis. Med Clin North Am, 2008; 92:925–960.

Portincasa

, Molina-Molina

, Garruti

, et al. Critical care aspects of gallstone disease. J Crit Care Med (Targu Mures), 2019; 4;5:6–18.

Hanau

, Steigbigel

. Acute (ascending) cholangitis. Infect Dis Clin North Am, 2000; 14:521–546.

Gomi

, Solomkin

, Schlossberg

, et al. Tokyo Guidelines 2018: Antimicrobial therapy for acute cholangitis and cholecystitis. J Hepatobiliary Pancreat Sci, 2018; 25:3–16.

Sand

, Airo

, Hiltunen

, et al. Changes in biliary bacteria after endoscopic cholangiography and sphincterotomy. Am Surg, 1992; 58:324–328.

Dindo

, Demartines

, Clavien

. Classification of surgical complications: A new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg, 2004; 240:205–213.

Domínguez

, Rivera

, Bermúdez

, et al. Analysis of factors for conversion of laparoscopic to open cholecystectomy: A prospective study of 703 patients with acute cholecystitis. Cirugia Espanola, 2011; 89:300–306.

Ercan

, Bostanci E

, Teke

, et al. Predictive factors for conversion to open surgery in patients undergoing elective laparoscopic cholecystectomy. J Laparoendosc Adv Surg Tech, 2010; 20:427–434.

10.

Ishizaki

, Miwa

, Yoshimoto

, et al. Conversion of elective laparoscopic to open cholecystectomy between 1993 and 2004. Br J Surg, 2006; 93:987–991.

11.

Nair

, Uzzaman

, Fafemi

, et al. Elective laparoscopic cholecystectomy in the presence of common bile duct stent. Surg Endosc, 2011; 25:429–436.

12.

Karsten

, Davids

, van Gulik

, et al. Effects of biliary endoprostheses on the extrahepatic bile ducts in relation to subsequent operation of the biliary tract. J Am Coll Surg, 1994; 178:343–352.

13.

Karsten

, Coene

, van Gulik

, et al. Morphologic changes of extrahepatic bile ducts during obstruction and subsequent decompression by endoprosthesis. Surgery, 1992; 111:562–568.

14.

Licciardello

, Arena

, Nicosia

, et al. Preoperative risk factors for conversion from laparoscopic to open cholecystectomy. Eur Rev Med Pharmacol Sci, 2014; 18:60–68.

15.

Sippey

, Grzybowski

, Manwaring M

, et al. Acute cholecystitis: Risk factors for conversion to an open procedure. J Surg Res, 2015; 199:357–361.

16.

Sultan

A M

, El Nakeeb

, Elshehawy

, et al. Risk factors for conversion during laparoscopic cholecystectomy: Retrospective analysis of ten years' experience at a single tertiary referral centre. Dig Surg, 2013; 30:51–55.

17.

Khoddami

, Chew

, Lange

. Problems and solutions of stent biofilm and encrustations: A review of literature. Turk J Urol, 2020; 46(Suppl 1):S11–S18.

18.

McGinniss

, Imai

, Simon-Soro

, et al. Molecular analysis of the endobronchial stent microbial biofilm reveals bacterial communities that associate with stent material and frequent fungal constituents. PLoS One, 2019; 29;14:e0217306.

19.

Rua

, Liao

, Peng

, et al. Microbial profiles and risk factors of preexisting biliary infection in patients with therapeutic endoscopy. Gastroenterol Res Pract, 2019; 2019:1527328.

20.

Chandra

, Klair

, Soota

, et al. Endoscopic retrograde cholangio-pancreatography-obtained bile culture can guide antibiotic therapy in acute cholangitis. Dig Dis, 2019; 37:155–160.