- Research article
- Open Access
- Open Peer Review
Laparoscopic versus open appendectomy in patients with suspected appendicitis: a systematic review of meta-analyses of randomised controlled trials
BMC Gastroenterologyvolume 15, Article number: 48 (2015)
Several systematic reviews (SRs) of randomised controlled trials (RCTs) comparing laparoscopic versus open appendectomy have been published, but there has been no overview of SRs of these two interventions. This overview (review of review) aims to summarise the results of such SRs in order to provide the most up to date evidence, and to highlight discordant results.
Medline, Embase, Cinahl, the Cochrane Database of Systematic Reviews and the Database of Abstracts of Reviews of Effects were searched for SRs published up to August 2014. Study selection and quality assessment using the AMSTAR tool were carried out independently by two reviewers. We used standardised forms to extract data that were analysed descriptively.
Nine SRs met the inclusion criteria. All were of moderate to high quality. The number of randomized controlled trials (RCTs) they included ranged from eight to 67. The duration of surgery pooled by eight reviews was 7.6 to 18.3 minutes shorter using the open approach. Pain scores on the first postoperative day were lower after laparoscopic appendectomy in two out of three reviews. The risk of abdominal abscesses was higher for laparoscopic surgery in half of six meta-analyses. The occurrence of wound infections pooled by all reviews was lower after laparoscopic appendectomy. One review showed no difference in mortality. The laparoscopic approach shortened hospital stay from 0.16 to 1.13 days in seven out of eight meta-analyses, though the strength of the evidence was affected by strong heterogeneity.
Laparoscopic and open appendectomy are both safe and effective procedures for the treatment of acute appendicitis. This overview shows discordant results with respect to the magnitude of the effect but not to the direction of the effect. The evidence from this overview may prove useful for the development of clinical guidelines and protocols.
Appendicitis is the most common reason for acute abdominal pain with a lifetime risk of 8.6% for males and 6.7% for females . The treatment of choice is the surgical removal of the inflamed appendix by using open appendectomy (OA) first described by McBurney in 1894 or by using laparoscopic appendectomy (LA) specified by Semm in 1983 [2,3]. Both surgical methods are safe and well established in clinical practice but there has been a controversy about which surgical procedure is the most appropriate. Therefore, several systematic reviews (SRs) have been conducted summarising, assessing and synthesising the data from primary studies. However, despite similar research questions and methodology, SRs show discordant results for individual endpoints.
The Cochrane Collaboration has introduced a new type of review called an overview of SRs. This method offers a new approach for synthesising the results of the increasing number of SRs. An overview summarises, evaluates and compiles the available evidence from SRs relevant to a single health problem . However, only a few methodological publications on how to conduct an overview are available [4,5]. Therefore, due to a lack of methodological and reporting standards, overviews have varied substantially in performance and in methodological quality, and the benefit of overviews has not been clearly established [6,7]. The purpose of this paper is to conduct an overview of SRs that compares LA versus OA to provide the most up to date evidence and to analyse the reasons for discordant results.
Systematic literature search
Medline, Embase, Cinahl, the Cochrane Database of Systematic Reviews (CDSR) and the Database of Abstracts of Reviews of Effects (DARE) were searched for SRs that compared LA versus OA in patients with suspected appendicitis by using a combination of text words and database specific controlled vocabulary without any restrictions regarding publication date or language (see Additional file 1 available online).The last update search was conducted on August 27, 2014. To identify additional citations missed by electronic searches, references of included studies were checked manually. There is no review protocol or registration available.
Two authors independently screened search results by title and abstract to identify potentially relevant SRs according to inclusion criteria created a priori. We included only the most recent version of a SR when updated versions were available. SRs without any systematic search in at least one database or without critical appraisal of included RCTs were excluded. After the retrieval of potentially relevant studies, full texts were checked against the inclusion criteria once again. Any disagreement was resolved by consensus. In the case of unresolvable discrepancies, a third reviewer was involved in the discussion.
Data extraction and quality assessment
The standardised data extraction form summarised year of publication, inclusion criteria, databases searched, search period and the number of included RCTs. We categorised the relevant outcomes as primary or secondary. Primary outcomes were pain on postoperative day 1, wound infections, intra-abdominal abscesses and mortality. Secondary outcomes were duration of surgery, conversions, length of stay, in-hospital costs (including surgery costs) and time until return to work. We extracted pooled effect sizes and corresponding confidence intervals if reported. To analyse the overlap of included SRs, we used a citation matrix that crosslinks the SRs with their included RCTs to compute the “covered area” (CA) and the “corrected covered area” (CCA) according to Pieper . For the evaluation of the methodological quality of the included SRs, we applied the eleven-item AMSTAR tool due to its reliability, construct validity and feasibility [9-11]. Each assessment question was rated with “yes”, “no”, or “can’t answer”. The data extraction, citation matrix and assessment of methodological quality were conducted by one author and checked by a second. Any disagreements were resolved by discussion or by consultation with a third reviewer.
Study selection process
The study selection process is presented in Figure 1. A total of 974 records were identified through the systematic search. After removing the duplicates, the title and abstract of 721 references were screened for meeting the inclusion criteria. After the retrieval of 36 potentially relevant full-text articles (including relevant supplements or appendices), 27 were excluded for the following reasons. One SR did not address patients with suspected appendicitis, one analysed LA without comparison, 18 included study designs of both RCTs and non-RCTs, one did not search systematically in at least one electronic database, and six had no quality assessment of the RCTs. Thus, a total of nine SRs were included in this overview [12-20].
Description of systematic reviews
All the included SRs published in English between 1998 and 2012 performed meta-analyses. Their characteristics are shown in Table 1. The included SRs analysed between eight and 67 RCTs on LA versus OA in patients with suspected appendicitis. Two of the SRs included only RCTs that had recruited only adults. Seven of the SRs applied a language limitation. The number of searched databases ranged from one to six. Medline was the most frequently searched electronic database followed by Cochrane Library and Embase. The number of analysed outcome measures ranged from one to 22. The methodological quality of the included RCTs was assessed by using the Jadad scale [14,15,17], the modified Jadad scale , the 10-point scale proposed by Solomon [13,19], the McMaster University method , the Cochrane risk of bias tool  and in one SR, the authors used their own checklist .
AMSTAR ratings for the reviews
The AMSTAR ratings are summarised in Table 2. All the included SRs were of moderate or high quality. Lack of an assessment of publication bias and the absence of a statement of potential sources of support were the most common flaws. Three SRs published before 2000 failed to conduct a comprehensive literature search by using only one database. One SR did not report the search period . Indeed, key words and MESH terms were stated frequently, but no author provided the complete search strategy. In their analysis of the RCTs, all the SRs presented the study characteristics, performed a critical appraisal, and used adequate methods for combining the results.
Overview of primary and secondary outcomes
The results of the extracted primary and secondary outcomes are given in Table 3. Three SRs showed a reduction of pain on postoperative day 1 in favour of LA compared with OA, but the effect sizes (which varied from -0.8 to -0.7 points on a 10 points VAS) were significant in only two of the SRs. These findings were affected by strong heterogeneity among the primary studies. The incidence of wound infections was significantly less for LA and the odds ratio (OR) ranged from 0.3 to 0.52 with low heterogeneity across the RCTs. Six SRs computed the OR for intra-abdominal abscesses; the values ranged from 1.56 to 2.29. Three meta-analyses showed no significant difference between LA and OA but three others detected significantly higher rates of intra-abdominal abscesses for LA. Only one SR analysed mortality rates; based on seven RCTs, this SR found that the mortality rates were not significantly different between the two surgical approaches.
Based on data pooled from eight SRs, the duration of surgery by LA took 7.6 to 18.3 minutes longer than by OA, though the results were limited by high heterogeneity. Two SRs determined the overall conversion rate to be 9.7% and 11%, with values ranging from 5% to 20% and from 0% to 20%, respectively. LA compared with OA led to a reduction in length of hospital stay of 0.16 to 1.13 days. These findings were significant in seven of eight SRs, though limited by high heterogeneity. The in-hospital costs, including surgery costs, were higher for the laparoscopic approach. For recovery time, the results of three SRs showed a trend in favour of the laparoscopic approach, but only two meta-analyses showed a significant reduction of three days in time until return to work.
Table 4 shows the citation matrix crosslinking nine SRs with 81 primary studies sorted by publication date in ascending order. Using Pieper’s method, the covered area (CA) and the corrected covered area (CCA) was 35.7% and 24.6%, demonstrating a very high degree of overlap.
Despite having the same research question and overlapping search periods, the SRs did not include the same set of RCTs due to their different exclusion criteria. In one case, the authors had no access to the full text , in another the author of an abstract did not answer the request for further information . Differences in the study selection process also resulted from different inclusion criteria. For example, the authors of one SR excluded two studies due to low follow-up . Additionally, there was discordance in excluding studies for the same inclusion criteria. Three SRs analysed a trial in which the assignment of patients to the intervention group had not been random but had instead been based on the schedule of the attending surgeon on call. Meynaud-Kraemer et al.  included one RCT which had been published after their initial literature search because one author providing the needed data was also involved in the primary study . A further comparison of included and excluded studies was not possible since the references of excluded studies were reported only in five SRs [12,18,19,14,16].
This overview aims to summarise SRs comparing LA versus OA for patients with suspected appendicitis to provide the most up to date evidence, and to highlight discordant results. Nine relevant SRs meeting all the inclusion criteria could be identified. Although we imposed no language restriction in order to prevent publication bias, the only relevant SRs we found were published in English. Our overview shows that LA and OA have been extensively analysed by RCTs and SRs, and that both approaches are safe and effective techniques for the treatment of suspected appendicitis and are associated with good clinical outcomes and little harm. The trend for reduced pain on postoperative day 1 after LA was lower in two out of three SRs but limited by high heterogeneity. The risk of abdominal abscesses was higher following LA in three out of six meta-analyses. The most clear and consistent finding with low heterogeneity was the reduction of wound infections after LA. The results of seven pooled RCTs showed no difference in mortality. The laparoscopic approach shortened hospital stay in eight meta-analyses, but again the data was heterogeneous.
The quality of the included SRs was moderate to high and thus met our quality evaluation criterion. Due to poor reporting, we could often not answer the AMSTAR item about ‘a priori design’ using only the publication for the assessment and not making any enquiries to the authors.
Not requesting further information from the authors in cases where data was missing is one weakness of our overview. For instance, there is a loss of information because the data on pain was not extracted from one study because it did not report the moment of pain measurement . Because we extracted only outcomes determined a priori, our presentation of the endpoints is incomplete. To reduce the risk of bias in our work, we included only those SRs for which a search in at least one electronic database had been conducted and which assessed the included RCTs critically by using a checklist.
Despite the different publication dates and number of included RCTs, the direction of effects for the analysed endpoints was the same and did not change over time. The direction of effect size estimates for wound infections and for the duration of surgery was significant in all SRs; however, there was a high variation in these effect size estimates. The discordant results are probably based on a combination of methodological causes and content-related reasons. Although the SRs had the same research question, they included different studies because they used different databases, search strategies and search periods. One SR did not specify the search strategy at all  and eight provided only keywords and general terms [12-15,17-20]. Thus, not a single search is completely comprehensible. Additional sources of discordant results are the different criteria used to select studies for inclusion. Some authors excluded studies due to a low follow-up, the lack of full text, insufficient resources to obtain the relevant paper, or language restrictions. In this overview, there is a low risk of bias concerning the study selection and the data extraction process since almost all the SRs conducted these quality assurance steps. Moreover, for pooling the data, the authors of the SRs applied either the fixed effect model or the random effect model or a combination of both depending on the heterogeneity. The authors of only three SRs contacted the authors of the primary studies to obtain missing data. One author request revealed that a trial classified as a randomised study used an inadequate sequence generation by allocating the patients according to the schedule of the attending surgeon on call.
Our research question focused on only two treatment procedures, but for decision makers, clinicians and patients, an overview including further types of interventions for appendectomy would be more interesting for their daily work and decision-making.
One fundamental disadvantage of overviews is the delayed integration of results from available primary studies. Overviews cannot reflect all the current evidence. In our example, the last published SR conducted its search in February 2012, so that RCTs published after this date have not been considered here. Thus, there is a lack of evidence of more than two years in our work. However, the direction of the effect size estimates is consistent for the analysed outcomes among the SRs and did not change over time. This makes it unlikely that the results of more recent RCTs would change the confidence in the effect estimates. The strength of SRs is their pooling of data on a particular problem from multiple RCTs. In an overview, it is not possible to adopt the methods for pooling data that are used in a SR without special modification, but even if the methods were so modified, success would be only partial because of the poor quality of reporting. Consequently, the results of this overview are presented only descriptively .
The comparison between LA and OA has been intensively analysed in over 70 RCTs; and further studies would unlikely change the results of SRs. Thus researchers and sponsors should rather focus on assessing new surgical approaches comparing single incision LA versus conventional three port LA for which there is currently insufficient evidence [21,22]. Indeed, the surgical appendectomy remains the standard treatment; however, conservative antibiotic therapy of acute appendicitis might be used in selected cases or in conditions where surgical approaches are contraindicated [23,24].
LA and OA are safe and effective procedures for the treatment of acute appendicitis in clinical practice. The direction of the pooled effects was consistent among the SRs. The evidence from this overview could be used for the development and updating of guidelines and protocols .
Corrected covered area
Randomised controlled trial
Addiss DG, Shaffer N, Fowler BS, Tauxe RV. The epidemiology of appendicitis and appendectomy in the United States. Am J Epidemiol. 1990;132(5):910–25.
McBurney IV C. The incision made in the abdominal wall in cases of appendicitis, with a description of a new method of operating. Ann Surg. 1894;20(1):38–43.
Semm K. Endoscopic appendectomy. Endoscopy. 1983;15(2):59–64.
Becker LA, Oxman AD. Chapter 22: Overviews of reviews. In: Higgins JPT, Green S (editors), Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available from www.cochrane-handbook.org.
Smith V, Devane D, Begley CM, Clarke M. Methodology in conducting a systematic review of systematic reviews of healthcare interventions. BMC Med Res Methodol. 2011;11(1):15.
Hartling L, Chisholm A, Thomson D, Dryden DM. A descriptive analysis of overviews of reviews published between 2000 and 2011. PLoS One. 2012;7(11):e49667.
Pieper D, Buechter R, Jerinic P, Eikermann M. Overviews of reviews often have limited rigor: a systematic review. J Clin Epidemiol. 2012;65(12):1267–73.
Pieper D, Antoine SL, Mathes T, Neugebauer EA, Eikermann M. Systematic review finds overlapping reviews were not mentioned in every other overview. J Clin Epidemiol. 2014;67(4):368–75.
Shea BJ, Bouter LM, Peterson J, Boers M, Andersson N, Ortiz Z, et al. External validation of a measurement tool to assess systematic reviews (AMSTAR). PLoS One. 2007;2(12):e1350.
Shea BJ, Grimshaw JM, Wells GA, Boers M, Andersson N, Hamel C, et al. Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews. BMC Med Res Methodol. 2007;7:10.
Shea BJ, Hamel C, Wells GA, Bouter LM, Kristjansson E, Grimshaw J, et al. AMSTAR is a reliable and valid measurement tool to assess the methodological quality of systematic reviews. J Clini Epidemiol. 2009;62(10):1013–20.
Bennett J, Boddy A, Rhodes M. Choice of approach for appendicectomy: a meta-analysis of open versus laparoscopic appendicectomy. Surg Laparosc Endosc Percutan Tech. 2007;17(4):245–55.
Golub R, Siddiqui F, Pohl D. Laparoscopic versus open appendectomy: a metaanalysis. J Am Coll Surg. 1998;186(5):545–53.
Li X, Zhang J, Sang L, Zhang W, Chu Z, Li X, et al. Laparoscopic versus conventional appendectomy–a meta-analysis of randomized controlled trials. BMC Gastroenterol. 2010;10:129.
Liu Z, Zhang P, Ma Y, Chen H, Zhou Y, Zhang M, et al. Laparoscopy or not: a meta-analysis of the surgical effects of laparoscopic versus open appendicectomy. Surg Laparosc Endosc Percutan Tech. 2010;20(6):362–70.
Meynaud-Kraemer L, Colin C, Vergnon P, Barth X. Wound infection in open versus laparoscopic appendectomy. Int J Technol Assess Health Care. 1999;15(2):380–91.
Ohtani H, Tamamori Y, Arimoto Y, Nishiguchi Y, Maeda K, Hirakawa K. Meta-analysis of the results of randomized controlled trials that compared laparoscopic and open surgery for acute appendicitis. J Gastrointest Surg. 2012;16(10):1929–39.
Sauerland S, Jaschinski T, Neugebauer EA. Laparoscopic versus open surgery for suspected appendicitis. Cochrane Database Syst Rev. 2010;10:CD001546.
Temple LK, Litwin DE, McLeod RS. A meta-analysis of laparoscopic versus open appendectomy in patients suspected of having acute appendicitis. Can J Surg. 1999;42(5):377–83.
Wei B, Qi CL, Chen TF, Zheng ZH, Huang JL, Hu BG, et al. Laparoscopic versus open appendectomy for acute appendicitis: a metaanalysis. Surg Endosc. 2011;25(4):1199–208.
Rehman H, Rao AM, Ahmed I. Single incision versus conventional multi-incision appendicectomy for suspected appendicitis. Cochrane Database Syst Rev. 2011;7:CD009022.
Xu AM, Huang L, Li TJ. Single-incision versus three-port laparoscopic appendectomy for acute appendicitis: systematic review and meta-analysis of randomized controlled trials. Surg Endosc. 2014;29(4):822–43.
Wilms IM, de Hoog DE, de Visser DC, Janzing HM. Appendectomy versus antibiotic treatment for acute appendicitis. Cochrane Database Syst Rev. 2011;11:CD008359.
Liu ZH, Li C, Zhang XW, Kang L, Wang JP. Meta-analysis of the therapeutic effects of antibiotic versus appendicectomy for the treatment of acute appendicitis. Exp Ther Med. 2014;7(5):1181–6.
Korndorffer Jr JR, Fellinger E, Reed W. SAGES guideline for laparoscopic appendectomy. Surg Endosc. 2010;24(4):757–61.
The authors declare that they have no competing interests.
TJ performed the systematic literature search, extracted, analysed the data and drafted the manuscript. CM selected the literature, extracted the data and performed the quality assessment. ME extracted the data and performed the quality assessment. EN analysed the data and reviewed the manuscript. All authors read and approved the final manuscript.