Intraoperative gallbladder perforation and risk of postoperative abscess with or without antibiotics: national cohort study of more than 108 000 cholecystectomies

Several studies have shown that routine antibiotic prophylaxis does not seem to reduce the risk for surgical site infection (SSI) after cholecystectomy1–6. Intraoperative perforation of the gallbladder can lead to spillage of contaminated bile and gallstones into the abdominal cavity. It is questionable whether this increases the risk for SSI. Some authors report that intraoperative gallbladder perforation increases the risk for SSI7,8, whereas a randomized trial did not find an increased risk for postoperative infection after spillage of bile during elective cholecystectomy, and prophylactic antibiotic treatment had no impact9. Similarly, in a recent meta-analysis of 11 observational studies, no increased risk for SSI after surgical gallbladder perforation could be found10. In a retrospective study, spillage of gallstones increased the risk for postoperative infection, but antibiotics did not reduce this risk11. A Turkish study comparing antibiotic prophylaxis with therapeutic treatment in cases of accidental gallbladder perforation during surgery found no significant difference12. The aims of the present study were to investigate if there is an association between intraoperative gallbladder perforation (with and without gallstones left in the abdominal cavity) and postoperative abscess, and to see if there is any evidence that antibiotic treatment in cases of spillage prevents abscess formation. This study was based on data from the Swedish National Quality Register for Gallstone surgery (GallRiks). GallRiks is a validated national database with more than 95 per cent coverage of all cholecystectomies performed in Sweden13. Data recorded in GallRiks include patient-related variables, perioperative data, and 30-day postoperative complications. The study was approved by the Swedish Ethics Review Authority (2019-05405). Data from all cholecystectomies recorded in GallRiks between 2008 and 2019 were used (Fig. 1). Cohort inclusion flowchart Intraoperative gallbladder perforation, sex, age, surgical approach (whether the surgery was completed laparoscopically or not), ASA grade, and acute cholecystitis were classed as predictors. These are well-known clinical predictors of increased risk for postoperative SSI8,14. Prophylactic administration of or treatment with antibiotics were also included as predictors. Administration of antibiotics for less than 24 h was defined as administration with prophylactic intent, and administration during more than 24 h as administration with therapeutic intent. Postoperative abscess within 30 days was chosen as the outcome. Surgical gallbladder perforation is registered in GallRiks as ‘gallbladder perforation with or without suspicion of gallstones left in the abdominal cavity’ by the surgeon performing the procedure. Postoperative infections within 30 days are registered in GallRiks as either abscess or other infections requiring antibiotic treatment. Data from 108 714 cholecystectomies were analysed. Intraoperative perforation with complete extraction of spilled gallstones was recorded in 34 113 cases (31.4 per cent), and in 2624 (2.4 per cent) the surgeon registered suspicion of gallstones left in the abdominal cavity. Antibiotic prophylaxis was given in 19 313 (17.8 per cent) cases, and in 12 726 (11.7 per cent) the patient received antibiotic treatment for more than 1 day (Table 1). Baseline characteristics Values are n (%) unless otherwise indicated. Baseline characteristics Values are n (%) unless otherwise indicated. Of the 20 821 cases with cholecystitis, antibiotics were given in 15 492 (74.4 per cent) cases (prophylaxis in 6200 cases and treatment in 9292 cases). In comparison, 16 547 (18.8 per cent) of the 87 893 cases without cholecystitis received antibiotics. When analysing combinations of cholecystitis and perforation, antibiotic use was uncommon in cases without both perforation and cholecystitis, but in those with cholecystitis or both perforation and cholecystitis, the majority received antibiotics (Fig. 2). Use of antibiotics in perforation and cholecystitis In univariable, as well as multivariable, analyses, gallbladder perforation was found to be associated with postoperative abscess with an odds ratio (OR) of 1.91 (95 per cent c.i. 1.71 to 2.13) in univariate analysis and 1.53 (1.36 to 1.71) when adjusted for other covariates (Table 2). The risk factors with the highest ORs for postoperative abscess were an age more than the median and procedures that could not be completed laparoscopically (2.05 and 2.55, respectively). Acute cholecystitis had an OR of 1.28 (95% c.i. 1.10 to 1.50), antibiotic prophylaxis an OR of 0.99 (95% c.i. 0.86 to 1.16), and antibiotic treatment 1.18 (95% c.i. 0.98 to 1.41); however, the ORs for antibiotic use were not significant. When specifically focusing on the subgroup with gallbladder perforation with suspicion of gallstones left in the abdominal cavity, the OR was slightly higher (1.75, 95% c.i. 1.35 to 2.27; Table 3). Regression analysis of gallbladder perforation as a risk factor for postoperative abscess Regression analysis of gallbladder perforation as a risk factor for postoperative abscess Regression analysis of gallbladder perforation with gallstones left in the abdominal cavity as a risk factor for postoperative abscess Values are n (%) unless otherwise indicated. OR, odds ratio. Regression analysis of gallbladder perforation with gallstones left in the abdominal cavity as a risk factor for postoperative abscess Values are n (%) unless otherwise indicated. OR, odds ratio. This study confirms that intraoperative gallbladder perforation increases the risk for postoperative abscess, as described in other studies4,7,8,11. Incomplete extraction of spilled gallstones increases the risk. If the increased risk for postoperative abscess is explained by contaminated bile, the risk should be substantially increased in cases of intraoperative perforation with cholecystitis compared to perforation without cholecystitis. However, in the interaction analysis in this study, the risk was increased after perforation, regardless of the presence or not of cholecystitis, and for perforation with cholecystitis. Use of antibiotics did not have a clinically relevant impact on outcome. The adjusted OR for antibiotic prophylaxis and treatment was not significant, except for treatment in the group with suspicion of gallstones left in the abdominal cavity, where it was paradoxically associated with an increased risk for abscess. We interpret this as residual confounding. It is not unusual that a dose of antibiotics is given intraoperatively if the gallbladder is perceived to be more inflamed (gangrenous/necrotic) than expected or if there is spillage of bile presumed to be contaminated. The slightly increased OR with antibiotic treatment in the subgroup probably reflects that these cases constitute a selected group of patients with severe cholecystitis. However, if the use of antibiotics were effective in preventing postoperative abscess formation, one would expect the OR for these covariates to be less than 1. The present study was based on data from a large, validated population-based register with nearly 100 per cent completion, and the final analysis included 108 714 patients—making it the largest study to date on this topic. Previous studies in this field have been small and often conducted at a single centre5,7,9–12,15. A meta-analysis10 included 11 observational studies and a total of 5366 patients. However, the present study is limited in that only data registered in GallRiks were analysed and there is therefore a risk of other confounding factors not being included. The timing of prophylaxis or type of antibiotics given is not registered in GallRiks, so we cannot say, from this study, if one type of antibiotics is preferable to another. The biological mechanism behind the increased risk for abscess after surgical perforation is not yet clear. The limited impact of antibiotic administration on the risk for abscess is not enough to recommend prophylactic use or routine administration with therapeutic intent in cases of perforation but may be considered in individual cases. This study was funded with grants from The Swedish Research Council. The authors wish to thank Hans Pettersson, Department of Clinical Science and Education Södersjukhuset, Karolinska Institutet, for help and support with the statistical analysis. The authors declare no conflict of interest. Authors confirm that data is available for viewing if required.