Role of prophylactic antibiotics in elective laparoscopic cholecystectomy: A systematic review and meta-analysis.

BACKGROUNDS/AIMS: The role of prophylactic antibiotics for laparoscopic cholecystectomy in low-risk patients is still unclear. This study aimed to verify the conclusion of previous meta-analyses concerning the effectiveness of antibiotic prophylaxis for elective laparoscopic cholecystectomy in low-risk patients.
METHODS: Comprehensive literature searches were performed on electric databases and manual searches. Randomized controlled trials (RCTs), prospective studies, and retrospective studies comparing antibiotic prophylaxis to placebo or no antibiotics in low-risk elective laparoscopic cholecystectomy were included.
RESULTS: This study included 28 RCTs, three prospective studies, and three retrospective studies. In RCTs, prophylactic antibiotics did not prevent deep surgical site infections (SSI) (RR 1.10, 95% confidence interval [CI] [0.45-2.69], p=0.84) but reduced SSI (RR 0.70, 95% CI [0.53-0.94], p=0.02), and superficial SSI (RR 0.58, 95% CI [0.42-0.82], p=0.01). Prospective studies showed prophylactic antibiotics did not reduce superficial SSI (RR 0.35, 95% CI [0.01-8.40], p=0.52) but reduced SSI (RR 0.12, 95% CI [0.04-0.35], p=0.0001). In retrospective studies, antibiotic prophylaxis did not reduce SSI (RR 1.59, 95% CI [0.30-8.32], p=0.58). The pooled data (12121 patients) including RCTs and prospective and retrospective studies showed that prophylactic antibiotics were not effective in preventing deep SSI (RR 1.01 95% CI [0.46-2.21], p=0.98) but effective in reducing SSI (RR 0.67, 95% CI [0.51-0.88], p=0.003) and superficial SSI (RR 0.61, 95% CI [0.45-0.83], p=0.002).
CONCLUSIONS: The use of prophylactic antibiotics is effective for reducing the incidence of SSI and superficial SSI but is not effective for preventing deep SSI in low-risk patients who underwent elective laparoscopic cholecystectomy.

INTRODUCTION

Laparoscopic cholecystectomy is the gold standard method in managing uncomplicated gallbladder stones and other benign gallbladder diseases over open cholecystectomy. Development of minimally invasive laparoscopic cholecystectomy reduced surgical site infection (SSI), length of hospital stay, healthcare costs, and postoperative pain.

Although the use of prophylactic antibiotics before elective surgery has been considered as the best way to prevent postoperative infectious complications, antibiotic prophylaxis for elective laparoscopic cholecystectomy in low-risk group is not recommended in recent guidelines on SSI from the Scottish Intercollegiate Guidelines Network and the American Society of Health-System Pharmacists.12 Nevertheless, low-risk patients who underwent laparoscopic cholecystectomy are still given prophylactic antibiotics in several clinical centers.

Recent meta-analyses345678910111213 investigating the effects of prophylactic antibiotics before elective laparoscopic cholecystectomy for the prevention of SSI have relatively small sample size and low statistical power. Moreover, controversy still exists regarding the effectiveness of antibiotic prophylaxis for elective laparoscopic cholecystectomy. Therefore, we performed an up-to-date meta-analysis to assess the value of prophylactic antibiotics for low-risk elective laparoscopic cholecystectomy in terms of reducing the incidence of SSI, superficial SSI, and deep SSI with randomized controlled trials (RCTs) and prospective and retrospective studies on this topic. This study aimed to verify the conclusion of previous meta-analyses.345678910111213

MATERIALS AND METHODS

The study protocol for this systemic review and meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines.14

Searching and other resources

MEDLINE, Embase, the Cochrane Central Register of Controlled Trials, PubMed, and KMbase were searched using medical subject heading terms and following keyword combinations: “laparoscopic,” “cholecystectomy,” “antibiotic,” “prophylac,” and “meta-analysis.” Table 1 shows the search strategies on electric databases in this review. Additionally, manual searches were performed with reference lists of original articles and systemic review and meta-analyses. The literature search was limited to articles published between 1995 and 2018 without restriction of languages.

Table 1

Search strategies

Inclusion and exclusion criteria

Inclusion criteria were (i) study design: RCTs and prospective and retrospective studies evaluating the effectiveness of antibiotic prophylaxis for elective laparoscopic cholecystectomy with placebo or untreated controlled group; (ii) population: low-risk patients undergoing elective laparoscopic surgery; (iii) all patients that were given antibiotics before operation and/or postoperative days; (iv) all studies having at least one of the following outcome parameters: SSI, superficial SSI and deep SSI.

Exclusion criteria were (i) studies that are not full-text original articles and (ii) interventions comparing different antibiotic prophylaxis groups without placebo or untreated groups.

SSIs include superficial and deep, and data were collected based on the definition of guideline.1516 High-risk factors for SSI are defined according to the diabetes, obesity, open conversion, emergency operation, preoperative endoscopic or percutaneous biliary intervention, acute cholecystitis, intraoperative gallbladder rupture, obstructive jaundice, immunosuppression, insertion of prosthetic device, and episode of colic within 30 days in guideline.12

Data collection and analysis

This systemic review and meta-analysis was performed following the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions.17

Study selection

Two reviewers (Kim and Yu) independently searched available articles to include eligible RCTs and prospective and retrospective studies. Duplicate articles were removed. The full-text articles of possibly related studies were selected to make a list of trials that fulfilled the inclusion criteria. Disagreements about study selection were resolved through discussion and consensus.

Data extraction

The following data were extracted independently by two reviewers (Kim and Yu) from each study when present: inclusion and exclusion criteria, characteristic of population, study design, type of prophylactic antibiotics and dosage, schedule of administration of antibiotics, randomization method, allocation concealment, number of randomized patients in RCTs and enrolled patients in prospective and retrospective studies, drop-outs, intention-to-treat analysis or per-protocol analysis, SSIs, and superficial and deep infections.

Quality assessment of the studies

Two investigators (Kim and Yu) independently assessed the quality of RCTs included through assessing the following risks of bias check lists provided by The Cochrane Collaboration18: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other biases. Risk of bias was assessed as high, low, and unclear. Jadad et al.19 score was also applied to assess the quality of included randomized studies in the meta-analysis. Of the 28 RCTs,20212223242526272829303132333435363738394041424344454647 25 studies22232425262728293031323334353637394041424344454647 were eligible to be included in the meta-analysis with high-quality Jadad score ≥3, and the other 3 studies202138 were low-quality studies with Jadad score 2. Sensitivity tests were performed to assess whether including three RCTs classified as low-quality is appropriate in this meta-analysis.

Non-RCTs including 3 prospective484950 and 3 retrospective225152 studies were assessed based on the Newcastle-Ottawa Scale,53 which ranges from 0 to 9, with “high quality” defined as a Newcastle-Ottawa Scale total score ≥7. Of the 6 non-RCTs included in this meta-analysis, 2 prospective studies were classified as “high quality.” Sensitivity tests were performed to obtain pooled data rate. Disagreements about assessment were resolved through discussion and consensus.

Statistical analysis

Statistical analyses were conducted using Review Manager version 5.3 (The Cochrane Collaboration). In dichotomous variable analysis, the effect size of the intervention was represented by risk ratio (RR) with 95% confidence interval (CI) and was calculated for continuous variables. A random effect model was applied to calculate the estimated pooled event rate in the presence of significant heterogeneity, determined either when the I2 value was >50% or p-value of Cochrane Q test was <0.1. The publication bias was evaluated to detect “small-study effects” by funnel plot.54 RCTs and prospective and retrospective studies were pooled separately to minimize the bias. Sensitivity analysis was performed to assess whether including non-randomized studies or low-quality studies is appropriate in this meta-analysis.17

RESULTS

A total of 1472 studies were screened through electronic database using our search strategies in July 2018. And 1096 articles were found by hand search. After investigation of title and abstract, 239 articles were screened. Full texts of 63 articles were reviewed and assessed for eligibility, and 34 studies including 28 RCTs and 3 prospective and 3 retrospective trials were included in this meta-analysis. Twenty-nine full-text articles555657585960616263646566676869707172737475767778798081828384 were excluded because of the following reasons: inclusion criteria not met,83 inappropriate intervention,5557707784 inappropriate comparator,565859 insufficient data reported, 606163656667686971727374757678808182 and data duplication.6264 The procedure for the study selection is summarized in the PRISMA flow diagram (Fig. 1).

Fig. 1

PRISMA flow diagram for study search.

Study characteristics

The 12121 patients who underwent laparoscopic cholecystectomy from eligible 28 RCTs with 7770 patients, 3 prospective studies with 3123 patients, and 3 retrospective studies with 1228 patients were included in this meta-analysis. Table 2 shows a summary of the characteristics of the included studies published between 1995 and 2018, including years of publication, country, language, study design, intervention, sample size, type of data collection analysis, follow-up days, and outcomes. Most of the patients included in RCTs were classified as low risk for postoperative infection. Exclusion criteria in the majority of RCTs were ASA physical status classification III or higher, antibiotics use within 7 days of the surgery, allergy, complicated gallstone with acute cholecystitis, choledocholithiasis or pancreatitis, previous biliary tract surgery, conversion to open cholecystectomy, diabetes mellitus, immunosuppression, pregnancy, prosthetic heart valve, severe comorbidities such as Child C liver cirrhosis or end-stage renal disease, body mass index >30, and age >70 years or <14 years. However, in the definition of high-risk patients, the criteria of ASA score, BMI, and old age are not exactly the same in included studies.

Table 2

Characteristics of included studies

RCTs, randomized controlled trials; SSI, surgical site infections; NA, not available; PP, Per-protocol analysis; ITT, Intention-to-treat analysis

Risk of bias and Jadad score for RCTs included in this meta-analysis are summarized in Tables 3 and 4 shows the Newcastle-Ottawa Scale for non-RCTs included in this review. In RCTs, only 12 studies212228323638404243444547 showed the data from intention-to-treat analysis, and the other 16 studies20232425262729303133343537394146 showed data from per-protocol analysis.

Table 3

Risk of bias and Jadad score of included randomized controlled trials

+, Low risk of bias; −, High risk of bias; ?, unclear risk of bias

Table 4

Newcastle-Ottawa Scale of included non-randomized studies

Surgical site infections

Subgroup analysis for SSI including both superficial and deep SSI was performed using 28 RCTs,20212223242526272829303132333435363738394041424344454647 3 prospective studies,484950 and 3 retrospective studies.225152 In the 28 RCTs, 83 (2.07%) of the 4018 patients in the prophylactic antibiotic group developed SSI compared with 119 (3.17%) of the 3752 patients in the no prophylaxis group. The prophylactic antibiotic group had less incidence of SSI than the control group (RR 0.70, 95% CI [0.53–0.94], p=0.02) without significant heterogeneity (p=0.96, I2=0%). In 3 prospective studies, 3 (0.16%) of the 1831 patients in the prophylactic antibiotic group developed SSI compared with 22 (1.70%) of the 1292 patients in the no prophylaxis group. The patients in the prophylactic antibiotic group had less incidence of SSI than the control group (RR 0.12 [0.04–0.37], p=0.0002). No significant heterogeneity was found among included prospective studies (p=0.48, I2=0%). In 3 retrospective studies, 15 (2.09%) of the 717 patients in the prophylactic antibiotic group developed SSI compared with 5 (0.98%) of the 511 patients in the no prophylaxis group. No significant difference was observed in the incidence of SSI between the prophylactic antibiotic group and the control group (RR 1.59, 95% CI [0.30–8.32], p=0.58) without significant heterogeneity among included prospective studies (p=0.58, I2=47%). In the overall pooled event rate including 28 RCTs and 3 prospective studies and 3 retrospective studies, 101 (1.54%) of the 6566 patients in the prophylactic antibiotic group developed SSI compared with 146 (2.63%) of the 5555 patients in the no prophylaxis group. The patients in the prophylactic antibiotic group had less incidence of SSI than the control group (RR 0.67, 95% CI [0.51–0.88], p=0.003). No significant heterogeneity among included studies was presented (p=0.47, I2=0%) (Fig. 2).

Fig. 2

Forest plot for surgical site infections in low-risk patients undergoing elective laparoscopic cholecystectomy. A Mantel-Haenszel fixed-effect model was for used for meta-analysis. Risk ratios are shown 95% confidence intervals.

Superficial surgical site infections

The incidence of superficial SSI was described in 22 RCTs,21222326282930313233343536373840414243454647 2 prospective studies,4950 and 1 retrospective study.52 In 22 RCTs, 59 (1.68%) of the 3508 patients of the prophylactic antibiotic group developed superficial SSI compared with 96 (2.95%) of the 3258 patients in the no prophylaxis group. The patients in the prophylactic antibiotic group had less incidence of superficial SSI than the patients in the control group (RR 0.59, 95% CI [0.43–0.82], p=0.001) without significant heterogeneity among included studies (p=0.91, I2=0%). In 2 prospective studies, 0 (0%) of the 143 patients in the prophylactic antibiotic group developed superficial SSI compared with 1 (1.03%) of the 97 patients in the no prophylaxis group. No difference was observed in outcomes between the two groups (RR 0.35 95% CI [0.01–8.40], p=0.52). In a retrospective study, 5 (1.79%) of the 279 patients in the prophylactic antibiotic group developed SSI compared with 3 (1.56%) of the 192 patients in the no prophylaxis group. In the overall pooled event rate including 22 RCTs, 2 prospective studies, and one retrospective study, 64 (1.63%) of the 3930 patients in the prophylactic antibiotic group developed superficial SSI compared with 100 (2.81%) of the 3547 patients in the no prophylaxis group. The prophylactic antibiotic group had less incidence of superficial SSI than the control group (RR 0.61, 95% CI [0.45–0.83], p=0.002). No significant heterogeneity was found among included studies (p=0.94, I2=0%) (Fig. 3).

Fig. 3

Forest plot for superficial surgical site infections in low-risk patients undergoing elective laparoscopic cholecystectomy. A Mantel-Haenszel fixed-effect model was for used for meta-analysis. Risk ratios are shown 95% confidence intervals.

Deep surgical site infections

The incidence of deep SSI was described in 19 RCTs,21232526282930313233343536384042434546 2 prospective studies,4950 and one retrospective study.52 In 19 RCTs, 10 (0.35%) of the 2890 patients in the prophylactic antibiotic group developed deep SSI compared with 10 (0.38%) of the 2637 patients in the no prophylaxis group. The prophylactic antibiotics did not reduce the incidence of deep SSI compared with control group (RR 1.01, 95% CI [0.46–2.21], p=0.98). No significant heterogeneity was found among included studies (p=0.77, I2=0%). In 2 prospective studies, 0 (0.00%) of the 143 patients in the prophylactic antibiotic group developed deep SSI compared with 0 (0.00%) of the 97 patients in the no prophylaxis group. In a retrospective study, 0 (0.00%) of the 279 patients in the prophylactic antibiotic group developed SSI compared with 0 (0.00%) of the 192 patients in the no prophylaxis group. In the overall pooled event rate including 19 RCTs, 2 prospective studies, and one retrospective study, 10 (0.30%) of the 3312 patients in the prophylactic antibiotic group developed deep SSI compared with 10 (0.34%) of the 2926 in the no prophylaxis group. No difference was observed in outcomes between the two groups (RR 1.01, 95% CI [0.46–2.21], p=0.98) without significant heterogeneity among included studies (p=0.77, I2=0%) (Fig. 4).

Fig. 4

Forest plot for deep surgical site infections in low-risk patients undergoing elective laparoscopic cholecystectomy. A Mantel-Haenszel fixed-effect model was for used for meta-analysis. Risk ratios are shown 95% confidence intervals.

Sensitivity test

Sensitivity tests were performed by excluding all low-quality studies including 3 RCTs202138 with Jadad score ≤2 and 4 non-randomized studies22505152 with Newcastle-Ottawa Scale score ≤6. In these sensitivity tests of all subgroups, the results were similar to those of the primary meta-analyses in pooled effect size and heterogeneity. Table 5 shows all sensitivity tests of the subgroups. To determine the effect of individual study, sensitivity analysis excluding one study at a time was also performed. Removal of the largest study48 including 2883 patients did not influenced the substantial change of the outcome (RR 0.74, 95% CI [0.56–0.97], p=0.03) in subgroup analysis of SSI.

Table 5

Sensitivity analyses

SSIs, surgical site infections; RCTs, randomized controlled trials; RR, risk ratio; CI, confidence interval.

Publication bias

A funnel plot of the included studies showed asymmetry, suggesting publication bias in the subgroup analysis of SSI (Fig. 5).

Fig. 5

Funnel plot for determination of publication bias in the subgroup analysis of SSI from all included studies. RR, Risk ratio; SE, Standard error.

DISCUSSION

This systematic review and meta-analysis included a total of 12121 patients from the 7770 patients in 28 RCTs, 3123 patients in 3 prospective studies, and 1228 patients in 3 retrospective studies. This study concluded that the use of prophylactic antibiotics in low-risk patients undergoing elective laparoscopic cholecystectomy prevents SSI and superficial SSI other than deep SSI. Previous meta-analyses345678910111213 were based only on data from RCTs and concluded that prophylactic antibiotics were not effective in preventing postoperative SSI in low-risk elective laparoscopic cholecystectomy, except for two studies.1012

In previous several meta-analyses on this topic, only RCTs were included to determine the overall effect rate. Therefore, these meta-analyses were performed with relatively small sample sizes and were statistically under-powered. To overcome the limitation of the study including only RCTs with small sample sizes, this meta-analysis comprised a total of 12121 patients from both RCTs and non-RCTs to obtain appropriate statistical power in the subgroup analysis of SSI, superficial SSI, and deep SSI. Moreover, all available RCTs and non-RCTs that have not been published in English were used in this study to reduce language bias. Although there was a possibility of inducing significant heterogeneity by combining RCTs with non-RCTs, there were no significant differences in results between RCTs alone and RCTs and non-RCTs in all subgroup analyses.

This systematic review and meta-analysis has several limitations. First, of the 28 RCTs, 3 trials have low quality with Jadad score. In addition, of the 6 non-RCTs, 4 studies were of “low quality” assessed by the Newcastle-Ottawa Scale. Therefore, sensitivity test was performed to obtain pooled data rate. Second, there was significant heterogeneity among non-RCTs in subgroup analysis of SSI. Third, few RCTs provide the data regarding high-risk patients associated with diabetes mellitus, steroid or immunosuppressive therapy, biliary obstruction, jaundice, antibiotic intake 7 days prior to surgery, emergency cholecystectomy, acute or chronic cholecystitis 6 weeks before surgery, and open conversion surgery. Therefore, whether prophylactic antibiotics play a role in high-risk laparoscopic cholecystectomy still remains unclear. Fourth, the number of patients in the included studies is insufficient to avoid type II error because most of the trials, including RCTs and non-RCTs, in this meta-analysis showed a relatively low incidence rate of SSI and very low difference in incidence rate between the prophylactic antibiotic group and control group. Fifth, there was publication bias evaluated by funnel plot that showed asymmetry, which suggests that small sample size studies reporting negative results have not been published. Sixth, most of the included RCTs in this review were performed per-protocol analysis instead of intension-to-treat analysis, inducing misleading results.85

In addition, an important consideration of this meta-analysis is the inconsistency of the inclusion criteria in the included studies. The Scottish Intercollegiate Guidelines Network1 and the American Society of Health-System Pharmacists guideline2 suggest intraoperative gallbladder rupture, open conversion, acute cholecystitis, jaundice, immunosuppression, pregnancy and implantation of prosthetic devices as high-risk factors for SSI. The Scottish Intercollegiate Guidelines Network1 also include diabetes, emergency surgery, long duration of procedures, ASA score of 3 or higher, recent episode of colic within 30 days before surgery and age >70 years as high-risk group for SSI. However, several studies232526272846 included patients with an ASA score of 3 or the ASA score was not applied to inclusion criteria in several RCTs.2232384043 Therefore, future studies need to provide a consistent set of inclusion criteria based on guidelines for defining low-risk or high-risk groups for SSI.

In conclusion, the overall pooled data of this meta-analysis from the present data including RCTs, prospective studies, and retrospective studies support the use of prophylactic antibiotics prior to elective laparoscopic cholecystectomy in low-risk patients to prevent SSIs and superficial SSIs. To assess the exact beneficial effects of antibiotic prophylaxis for laparoscopic cholecystectomy, more well-designed multicenter RCTs with large sample size, different population groups, and adequate statistical power for high-risk patients are necessary because most of the trials on this topic have focused on particularly the low-to-moderate risk patients. Furthermore, subgroup analyses including overall infection, extra-abdominal infection, and duration of postoperative hospital stay are also required with recent RCTs and non-RCTs.