Role of Prophylactic Antibiotics in LaparoscopicCholecystectomy: A Meta-Analysis

Abhishek Choudhary & Matthew L. Bechtold &

Srinivas R. Puli & Mohamed O. Othman &

Praveen K. Roy

Received: 22 May 2008 /Accepted: 20 August 2008 /Published online: 9 September 2008# The Society for Surgery of the Alimentary Tract 2008

AbstractBackground The role of prophylactic antibiotics in laparoscopic cholecystectomy in low-risk patients is controversial. Weconducted a meta-analysis to evaluate the efficacy of prophylactic antibiotics in low-risk patients (those withoutcholelithiasis or cholangitis) undergoing laparoscopic cholecystectomy.Methods Multiple databases and abstracts were searched. Randomized controlled trials (RCTs) comparing prophylacticantibiotics to placebo or no antibiotics in low-risk laparoscopic cholecystectomy were included. The effects of prophylacticantibiotics were analyzed by calculating pooled estimates of overall infections, superficial wound infections, majorinfections, distant infections, and length of hospital stay. Separate analyses were performed for each outcome by using oddsratio or weighted mean difference. Both random and fixed effects models were used. Publication bias was assessed byfunnel plot. Heterogeneity among studies was assessed by calculating I2 measure of inconsistency.Results Nine RCTs (N=1,437) met the inclusion criteria. No statistically significant reduction was noted for those receivingprophylactic antibiotics and those who did not for overall infectious complications (p=0.20), superficial wound infections(p=0.36), major infections (p=0.97), distant infections (p=0.28), or length of hospital stay (p=0.77). No statisticallysignificant publication bias or heterogeneity were noted.Conclusions Prophylactic antibiotics do not prevent infections in low-risk patients undergoing laparoscopic cholecystectomy.

Keywords Laparoscopic cholecystectomy .

Prophylactic antibiotics . Superficial infection .

Meta-analysis

Introduction

Laparoscopic cholecystectomy has become the first-linetreatment modality for symptomatic cholelithiasis overopen cholecystectomy. The laparoscopic approach has anextremely low rate of postoperative infection (0.4–1.1%)in comparison to open cholecystectomy, consistingmostly of superficial site infections at the umbilicaltrocar site.1–4 The infection complications of opencholecystectomy are well known and prevalent; therefore,prophylactic antibiotics are routinely indicated. However,the use of prophylactic antibiotics in laparoscopic chole-cystectomy remains unclear despite its popularity. Fewstudies have shown that prophylactic antibiotics inlaparoscopic cholecystectomy decrease the incidence ofpostoperative complications in laparoscopic cholecystec-tomy.5–7 Other randomized controlled trials (RCTs) havedemonstrated no obvious role of prophylactic antibiotics inlaparoscopic cholecystectomy.8–16 However, these RCTs

J Gastrointest Surg (2008) 12:1847–1853DOI 10.1007/s11605-008-0681-x

Scientific Meeting: Data presented at Digestive Disease Week on 19May 2008 at San Diego, CA.

A. Choudhary :M. L. Bechtold : S. R. Puli : P. K. RoyDivision of Gastroenterology,University of Missouri School of Medicine,Columbia, MO, USA

M. O. OthmanUniversity of New Mexico,Albuquerque, NM, USA

P. K. Roy (*)ABQ Health Partners,2nd Floor, Gastroenterology, 5400 Gibson Blvd SE,Albuquerque, NM 87108, USAe-mail: pk2949@yahoo.com

were small or terminated early due to paucity of majorinfections.8–16 Due to the small sample sizes of the RCTs, anadequate power to detect a difference for antibiotic use forthe rare event of infections may not have been achieved. Weconducted a meta-analysis of randomized controlled trials toevaluate the role of prophylactic antibiotics in laparoscopiccholecystectomy.

Materials and Methods

Study Selection Articles and abstracts that evaluated theuse of antibiotic administration for the prevention ofinfection in laparoscopic cholecystectomy were searched.All articles were searched irrespective of language,publication status (articles or abstracts), or results. Asearch was conducted in MEDLINE, EMBASE, CochranCentral Register of Controlled Trials, and Pubmed(1966–October 2007). The search terms used wereprophylactic administration of antibiotics and laparoscop-ic cholecystectomy. Additionally, references lists ofretrieved articles, reviews, and meta-analyses werescanned for potential articles. Lastly, a manual searchof abstracts submitted to the Digestive Disease Week,American College of Gastroenterology, and UnitedEuropean Gastroenterology Week (2000–2007) was per-formed. Inclusion criteria were randomized controlledtrials that used prophylactic antibiotic(s) versus noantibiotics or placebo for laparoscopic cholecystectomywith overall infection as an end point. Exclusion criteriaconsisted of studies that were uncontrolled, not involving

overall infection as an end point, or comparing twodifferent antibiotics rather than placebo or control.

Data Extraction Data extraction was independently per-formed by two authors (Choudhary and Bechtold) andreviewed by a third for agreement. Disagreements werediscussed by all three and resolved by consensus. Thetwo authors (AC and MLB) extracted data from eachstudy using a common data extraction form. Details ofstudy design (randomization/blinding), number of sub-jects and dropouts, as well as type, dose, and schedule ofantibiotic administration were recorded. Outcomes ofoverall, superficial, and distant infections as well aslength of hospital stay were recorded. All studies wereassigned a quality score on the based upon the Jadadscale, with 5 representing a high-quality study and 0representing a poor quality.17

Data Analysis The effects of prophylactic antibiotics onlaparoscopic cholecystectomy were analyzed by calculatingpooled estimates of total, superficial, and distant infections.Separate analyses were performed for each outcome usingodds ratio (OR) or weighted mean difference (WMD). Bothfixed and random effects models were used. A statisticallysignificant result was indicated by a p value <0.05 or 95%confidence interval (CI) not including 1. If statisticalsignificance was detected, the number needed-to-treat wascalculated. RevMan 4.2 software was utilized for statisticalanalysis of the data. Publication bias was assessed byfunnel plot. Heterogeneity among studies was assessed bycalculating I2 measure of inconsistency.18,19

Initial search 133 Articles

20 relevant articles selected & reviewed

113 Articles excluded

11 articles excluded

9 RCT’s finally included in study

Comparing antibiotics vs Placebo or control.

Non RCT

Comparing 2 different antibiotics

& modes of administration

Comparing 2 different

modes of prophylaxis.

Involving overall infection as one

of the end point

Figure 1 Article identificationand selection algorithm.

1848 J Gastrointest Surg (2008) 12:1847–1853

Results

The initial search identified 133 articles using the searchterms “laparoscopic cholecystectomy” and “antibiotics”. Ofthese, 20 relevant articles were selected and reviewed bytwo independent authors (AC and MLB). One hundredthirteen studies did not meet the inclusion criteria and wereexcluded, including case reports, case series, reviews, andretrospective studies. Subsequently, 11 additional studiesdid not meet the inclusion criteria and were excluded,including non-randomized prospective studies6,20 and RCTsusing two different antibiotics21 or comparing two modesand doses of antibiotics.22,23 Nine RCTs (N=1,437),published as full-length publications in journals, met theinclusion criteria and were selected for final review andanalysis (Fig. 1). Of the included nine RCTs, three trialswere double-blinded. Table 1 shows the details and Jadadscores for the selected studies (5 = excellent quality, 0 =

poor quality). The studies were of adequate quality (Jadadscores of 2 or more). All RCTs were published from 1997to 2006. Trials were done worldwide, including four trialsperformed in the USA, three trials in Asia, and two trials inEurope. All trials were single-center studies. No significantheterogeneity was present among the studies for any of theoutcomes.

Different antibiotics were evaluated in the selected trials.Three RCTs used cefazolin, two used cefotaxime andcefuroxime, one used cefotetan, and one used cefotetanand cefazolin. Antibiotics were administered preoperativelyin all studies. Three RCTs used multiple doses with the firstdose preoperatively and other doses postoperatively. Pub-lication bias was evaluated by funnel plot with nosignificant publication bias identified (Fig. 2).

Overall Infectious Complications Nine trials providedinformation about overall infectious complications.8–16

Table 1 Description of Studies Included in the Meta-Analysis, Including Jadad Scores

Author Year Location Centers Type of study Jadad score

Chang et al. 2006 Taiwan Single Single-blinded 4Higgins et al. 1999 United States Single Double-blinded 5Illig et al. 1997 United States Single RCT 2Tocchi et al. 2000 United States Single Single-blinded 4Koc et al. 2003 Turkey Single Double-blinded 3Kuthe et al. 2006 India Single Single-blinded 4Mahatharadol et al. 2001 Thailand Single RCT 3Dobay et al. 1999 USA Single Double-blinded 4Harling et al. 2000 UK Single RCT 3

Figure 2 Funnel plot for over-all infections suggesting nopublication bias by showingmultiple studies on both sides ofthe dotted line in an approxi-mately equal distribution.

J Gastrointest Surg (2008) 12:1847–1853 18491849

Figure 3 Forrest plot demon-strating overall infectious com-plications with prophylacticantibiotic(s) compared to noantibiotic(s) or placebo for lap-aroscopic cholecystectomy.

Figure 4 Forrest plot demon-strating superficial infectionwith prophylactic antibiotic(s)compared to no antibiotic(s) orplacebo for laparoscopiccholecystectomy.

Figure 5 Forrest plot demon-strating major infection withprophylactic antibiotic(s) com-pared to no antibiotic(s) orplacebo for laparoscopiccholecystectomy.

1850 J Gastrointest Surg (2008) 12:1847–1853

The study by Dobay et al.13 demonstrated no infectionsfor either the group, resulting in the inability to analyzethe data. Therefore, the Dobay et al. study is not includedin the Forrest plot. Overall infectious complications weredocumented in 19 of 797 patients (2.4%) treated withprophylactic antibiotics prior to laparoscopic cholecystecto-my versus 23 of 640 patients (3.6%) not treated withprophylactic antibiotics. Pooled analysis revealed no statis-tically significant odds reduction with prophylactic anti-biotics prior to laparoscopic cholecystectomy for overallinfectious complications (OR 0.66; 95% CI 0.35–1.24; p=0.20; Fig. 3). There was no significant heterogeneity amongthe studies (I2=0%, p=0.96). Further subgroup analyseswere performed according to types of infection.

Superficial Wound Infections Eight trials provided infor-mation regarding superficial infections.8–12,14–16 Superficialwound infections were present in 13 of 797 patients (1.6%)who received prophylactic antibiotics prior to laparoscopiccholecystectomy and 15 of 640 patients (2.3%) who did notreceive prophylactic antibiotics. Pooled analysis showed nostatistically significant odds reduction with prophylacticantibiotics prior to laparoscopic cholecystectomy for super-ficial wound infections (OR 0.71; 95% CI 0.34–1.48; p=0.36; Fig. 4). Heterogeneity was not statistically significant(I2=0%, p=0.96).

Major Infections Only four trials offered informationregarding major infections.10,11,15,16 Major infections, inthe form of intraabdominal collections or abscesses, werepresent in two of 630 patients (0.3%) who receivedprophylactic antibiotics prior to laparoscopic cholecystec-tomy versus two of 486 patients (0.4%) who received noprophylactic antibiotics. Pooled analysis demonstrated nostatistically significant odds reduction with prophylacticantibiotics prior to laparoscopic cholecystectomy for majorinfections (OR 1.03; 95% CI 0.25–4.20; p=0.97; Fig. 5).Heterogeneity was not statistically significant (I2=0%, p=0.67).

Distant Infections Only three trials provided informationregarding distant infections.9–11 Distant infections weredefined as any infection away from the wound, includingurinary tract or respiratory tract infections. Distantinfections were present in four of 499 patients (0.8%)who received prophylactic antibiotics prior to laparo-scopic cholecystectomy versus six of 297 patients(2.0%) who received no prophylactic antibiotics. Pooledanalysis showed no statistically significant odds reduc-tion with prophylactic antibiotics prior to laparoscopiccholecystectomy for distant infections (OR 0.49; 95%CI 0.13–1.81; p=0.28; Fig. 6), with no heterogeneityidentified (I2=0%, p=0.77).

Figure 7 Forrest plot demon-strating hospital stay with pro-phylactic antibiotic(s) comparedto no antibiotic(s) or placebo forlaparoscopic cholecystectomy.

Figure 6 Forrest plot demon-strating distant infection withprophylactic antibiotic(s) com-pared to no antibiotic(s) or pla-cebo for laparoscopiccholecystectomy.

J Gastrointest Surg (2008) 12:1847–1853 18511851

Hospital Stay Only three trials offered evaluation regardinghospital stay.12,15,16 Prophylactic antibiotics prior to lapa-roscopic cholecystectomy did not lead to shorter hospitalstays (WMD 0.02; 95% CI −0.10–0.14; p=0.77), with noheterogeneity identified (I2=0%, p=0.52; Fig. 7).

Discussion

Despite controversy surrounding the use of prophylacticantibiotics in laparoscopic cholecystectomy, 79% ofpatients undergoing laparoscopic cholecystectomy havereceived prophylactic antibiotics preoperatively and 63%received antibiotics postoperatively.3 Many studies haveevaluated this issue further with controversial results.

A prospective non-randomized trial by Frantzides andSykes20 found no beneficial effect of prophylactic cefotetanover chlorhexidine gluconate scrub alone. Chang et al.15

demonstrated that no prophylactic antibiotics (cefotetan) arenecessary after wound closure in an effort to decreaseincidence of superficial wound infections in electivelaparoscopic cholecystectomies. Furthermore, Kuthe etal.16 also demonstrated a similar result with cefuroxime.

Tocchi et al.11 concluded that antibiotics prophylaxisshould be given only in those patients with episodes ofcolic within 30 days of surgery or diabetes. Koc et al.14

concluded no role of prophylactic antibiotics in laparoscop-ic cholecystectomy in 92 patients. Higgins et al.9 alsoconcluded that prophylactic cefotetan and cefazolin have nobeneficial effects in laparoscopic cholecystectomy. Further-more, if no antibiotics were used, savings of ~$30,000 werecalculated at the investigator’s institute (USA).9

In our meta-analysis, prophylactic antibiotics prior tolaparoscopic cholecystectomy resulted in no statisticallysignificant benefit for total infections, superficial infections,major infections, distant infections, and reduction ofhospital stay.

The strengths of this meta-analysis include use of onlyrandomized controlled trials, varying populations (Europe,USA, Asia), and similar outcomes in all studies eventhough various antibiotics were utilized. Also, no hetero-geneity was noted for any of the major outcomes and nopublication bias was noted. Limitations of this meta-analysis include uncertainty about the use of prophylacticantibiotics in high-risk patients undergoing laparoscopiccholecystectomy, which is controversial at this time. High-risk patients have been defined by some investigators asage >60 years or the presence of diabetes mellitus, acutecolic within 30 days before laparoscopic cholecystectomy,jaundice, acute cholecystitis, or cholangitis. Tocchi et al.11

and Koc et al.14 found that the presence of diabetesmellitus, episodes of biliary colic in preceding 30 days ofsurgery, and age >60 years were independent risk factors

for the development of infectious complications; however,Kuthe et al.16 and Chang et al.15 failed to show similarresults. Despite the controversy, none of the RCTs providedseparate data about the effect of prophylactic antibiotics inlaparoscopic cholecystectomy in this particular subgroup ofhigh-risk patients for comparison. In addition, all trialsexcluded those patients with choledocholithiasis and chol-angitis and all trials, except one15, excluded patients withacute cholecystitis. Therefore, since this high-risk popula-tion was not evaluated in the RCTs, this population cannotbe fully evaluated in this meta-analysis.

In conclusion, the current meta-analysis of RCTs on theuse of prophylactic antibiotics in laparoscopic cholecys-tectomy reveals no beneficial effects in low-risk individ-uals. Future multicenter RCTs with adequate statisticalpower and involving a higher number of patients withsubgroups, particularly those at high-risk for infections,are needed to complete the evaluation of prophylacticantibiotics prior to laparoscopic cholecystectomy for high-risk patients.

Acknowledgments No additional acknowledgments. No grant sup-port or external funding were utilized.

References

1. Shea JA, Berlin JA, Bachwich DR, Staroscik RN, Malet PF,McGuckin M, Schwartz JS, Escarce JJ. Indications for andoutcomes of cholecystectomy: a comparison of the pre and postlaparoscopic eras. Ann Surg 1998;227:343–350.

2. Chuang SC, Lee KT, Chang WT, Wang SN, Kuo KK, Chen JS,Sheen PC. Risk factors for wound infection after cholecystectomy.J Formos Med Assoc 2004;103:607–612.

3. McGuckin M, Shea JA, Schwartz JS. Infection and antimicrobialuse in laparoscopic cholecystectomy. Infect Control Hosp Epi-demiol 1999;20:624–626.

4. The Southern Surgeon’s club. A prospective analysis of 1518laparoscopic cholecystectomies. N Engl J Med 1991;324:1073–1078.

5. Shindholimath VV, Seenu V, Parshad R, Chaudhry R, Kumar A.Factors influencing wound infection following laparoscopiccholecystectomy. Trop Gastroenterol 2003;24:90–92.

6. Uchiyama K, Kawai M, Onishi H, Tasni M, Kinoshita H, Ueno M,Yamaue H. Preoperative antimicrobial administration for preventionof postoperative infection in patients with laparoscopic cholecys-tectomy. Dig Dis Sci 2003;48:1955–1959.

7. Al-Abassi AA, Farghaly MM, Ahmed HL, Mobasher LL,Al-Manee MS. Infection after laparoscopic cholecystectomy:effect of infected bile and infected gallbladder wall. Eur J Surg2001;167:268–273.

8. Harling R, Moorjani N, Perry C, MacGowan AP, Thompson MH.A prospective, randomized trial of prophylactic antibiotics versusbag extraction in the prophylaxis of wound infection in laparo-scopic cholecystectomy. Ann R Coll Surg Engl 2000;82:408–410.

9. Higgins A, London J, Charland S, Ratzer E, Clark J, Haun W,Maher DP. Prophylactic antibiotics for elective laparoscopiccholecystectomy: are they necessary? Arch Surg 1999;134:611–613.

1852 J Gastrointest Surg (2008) 12:1847–1853

10. Illig KA, Schmidt E, Cavanaugh J, Krusch D, Sax HC. Areprophylactic antibiotics required for elective laparoscopic chole-cystectomy? J Am Coll Surg 1997;184:353–356.

11. Tocchi A, Lepre L, Costa G, Liotta G, Mazzoni G, Maggiolini F.The need for antibiotic prophylaxis in elective laparoscopiccholecystectomy: a prospective randomized study. Arch Surg2000;135:67–70.

12. Mahatharadol V. A reevaluation of antibiotic prophylaxis inlaparoscopic cholecystectomy: a randomized controlled trial. JMed Assoc Thai 2001;84:105–108.

13. Dobay KJ, Freier DT, Albear P. The absent role of prophylacticantibiotics in low-risk patients undergoing laparoscopic cholecys-tectomy. Am Surg 1999;65:226–228.

14. Koc M, Zulfikaroglu B, Kece C, Ozalp N. A prospectiverandomized study of prophylactic antibiotics in elective laparo-scopic cholecystectomy. Surg Endosc 2003;17:1716–1718.

15. Chang WT, Lee KT, Chuang SC, Wang SN, Kuo KK, Chen JS,Sheen PC. The impact of prophylactic antibiotics on postoperativeinfection complication in elective laparoscopic cholecystectomy: aprospective randomized study. Am J Surg 2006;191:721–725.

16. Kuthe SA, Kaman L, Verma GR, Singh R. Evaluation of the role ofprophylactic antibiotics in elective laparoscopic cholecystectomy: aprospective randomized trial. Trop Gastroenterol 2006;27:54–57.

17. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ,Gavaghan DJ, McQuay HJ. Assessing the quality of reports ofrandomized clinical trials: is blinding necessary? Control ClinTrials 1996;17:1–12.

18. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–1558.

19. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuringinconsistency in meta-analyses. Br Med J 2003;327:557–560.

20. Frantzides CT, Sykes A. A reevaluation of antibiotic prophylaxisin laparoscopic cholecystectomy. J Laparoendosc Surg1994;4:375–378.

21. Orozco H, Sifuentes-Osorio J, Chan C, Medina-Franco H,Vargas- Vorackova F, Prado E, Arch J. Comparison of ceftibutenvs. amoxicillin/clavulanic acid as antibiotic prophylaxis incholecystectomy and/or biliary tract surgery. J Gastrointest Surg2000;4:606–610.

22. Pourriat JL, The French Multicentric Group. Antibiotic prophy-laxis of laparoscopic cholecystectomies with 1 versus 2 grams ofcefotetan. Rec Adv Chemother 1995;19:554–556.

23. Zurbuchen U, Ritz JP, Lehmann KS, Groene J, Heidari M, BuhrHJ, Germer CT. Oral vs intravenous antibiotic prophylaxis inelective laparoscopic cholecystectomy-an exploratory trial. Lan-genbecks Arch Surg 2008;393(4):479–85. (Jul).

Discussion

John B. Marshall, M.D. (Columbia, MO): This is apractical paper that has the potential to change practicehabits. A majority of surgeons presently give prophylacticantibiotics before laparoscopic cholecystectomy. Whilerandomized controlled trials have not shown a benefit, anumber of the trials have been underpowered and notincluded enough subjects to exclude a benefit. Meta-analysis is a statistical technique that permits the results ofdifferent studies to be combined. The results of this well-conducted meta-analysis found no benefit from prophylac-tic antibiotics given before laparoscopic cholecystectomy.This is an important finding given the cost implications andvarious other potential deleterious effects of prescribingunwarranted antibiotics. Most of the trials in this studyexcluded so-called high-risk patients, though the variousstudies tended to define high risk in various ways.Additional investigation is needed in the high-risk subset.However, the verdict seems clear in most patients under-going laparoscopic cholecystectomy, prophylactic antibiot-ics are not needed.

J Gastrointest Surg (2008) 12:1847–1853 18531853