Evaluation of Drain Insertion After Appendicectomy for Complicated Appendicitis: A Systematic Review and Meta-Analysis

This meta-analysis aims to evaluate the comparative outcomes of drain insertion versus no drain after appendicectomy for complicated appendicitis. A systematic search of PubMed, Cochrane Library and Scopus was conducted, and all studies comparing drain versus no drain after appendicectomy for complicated appendicitis were included. Abdominal collection, surgical site infection (SSI), bowel obstruction, faecal fistula, paralytic ileus, length of hospital stay (LOS) and mortality were the evaluated outcome parameters for the meta-analysis. Seventeen studies reporting a total number of 4,255 patients who underwent appendicectomy for complicated appendicitis with (n=1,580) or without (n=2,657) drain were included. There was no significant difference between the two groups regarding abdominal collection (odds ratio (OR)=1.41, P=0.13). No-drain group was superior to the drain group regarding SSI (OR=1.93, P=0.0001), faecal fistula (OR=4.76, P=0.03), intestinal obstruction (OR=2.40, P=0.04) and paralytic ileus (OR=2.07, P=0.01). There was a difference regarding mortality rate between the two groups (3.4% in the drain group vs 0.5% in the no-drain group, risk difference (RD)=0.01, 95% CI (-0.01, 0.04), P=0.36). In conclusion, this meta-analysis has shown that drains have no effect on the development of intra-abdominal collections in complicated appendicitis, but it can significantly increase the risk of postoperative complications such as fistula, surgical site infection (SSI), bowel obstruction, ileus and length of hospital stay.

appendicectomies do not reduce the risk of postoperative complications, and they may increase the rate of other postoperative complications, including enterocutaneous fistula formation (4.2%-7.5%) and wound infection (43%-85%) [19]. Therefore, there is a debate on the overall benefits of using surgical drains routinely in appendicectomy. Our aim was to systematically review the literature and conduct a metaanalysis of the available relevant studies to assess the advantages of using drains versus no drains in complicated appendicitis.

Review
This systematic review was designed, performed and reported as per the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions and Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for meta-analyses reporting [20,21]. Inclusion and exclusion criteria are shown in Table 1.

Inclusion criteria Exclusion criteria
Randomized controlled trials (RCTs)

Search strategy
A literature search was performed by three authors independently in PubMed, Cochrane Library and Scopus up to June 2021. All studies of randomized controlled trials (RCTs) or comparative observational studies comparing the outcomes of drain or non-drain after open or laparoscopic appendicectomy for patients with complicated acute appendicitis were included in this analysis. Single-arm studies, case series and case reports were excluded. The keywords used in the search were "drain" [All Fields] AND ("drainage" [MeSH Terms] OR "drainage" [All Fields] OR "drain" [All Fields]) AND "laparoscopic appendicectomy" OR "laparoscopic appendectomy" [All Fields] OR "open appendicectomy" [All field] AND "acute appendicitis" [All Fields]. The searches were limited to human subjects, and no language or publication date restrictions were applied. The references of relevant reviews and included studies were also checked manually to identify additional potentially eligible studies.

Outcomes
Primary outcomes were the rates of postoperative abdominal collection formation. Postoperative complications including surgical site infection, bowel obstruction, faecal fistula, paralytic ileus, antibiotics duration in days, length of hospital stay (LOS) in days and mortality were considered secondary outcomes.

Study selection and data extraction
Three authors executed the literature, duplicated studies were excluded, the titles and abstracts were evaluated for relevance, and then records were classified as included, excluded or requiring further evaluation. All studies meeting the inclusion criteria of our study were included. Disagreements in the selection of studies were resolved by discussion between the reviewers. However, if the discrepancies remained unresolved, a fourth reviewer was involved.
All data were extracted manually and independently by three authors and revised by another three authors using Excel (2018, Microsoft Corporation, Washington, United States). The information collected from each study was author, year of publication, study design, number of populations in the study, presence of drain or no drain and comparative outcomes.

Assessment for risk of bias
The Cochrane risk of bias tool was used to appraise the risk of bias for the randomized trials [22]. Two investigators independently reviewed all studies and graded the risk as "high," "low" or "unclear" in the following categories: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting and other sources of bias. Newcastle-Ottawa scale (NOS) was used to assess the risk of bias in observational studies [23]. Studies were considered low, medium or high risk of bias if the NOS was 9, 7 or 8 or <6, respectively. Disagreement was resolved with discussion and consultation of a third reviewer (Figures 1, 2  RCT: randomized control trial.

Statistical analysis
Meta-analysis was performed using Review Manager (RevMan) Version 5.4, The Cochrane Collaboration, 2020. Dichotomous outcomes were pooled with a random-effects model to estimate the odds ratio (OR) or risk difference (RD) (where more than three studies reported zero events in both drain and no-drain groups) with 95% confidence interval. For continuous outcomes, a mean difference (MD) with 95% CI was estimated.
When mean values were not available for continuous outcomes, data on median and interquartile range (IQR) were extracted and subsequently converted to mean and standard deviation (SD) using the wellpractised equation described by Hozo et al. [24].
The results were considered statistically significant if the P-value was <0.05 or if the 95% CI did not include 1.00. Heterogeneity evaluated using the Cochran Q test (χ 2 ) was used to evaluate the heterogeneity and I 2 was reported to quantify it: a value of 0 % indicated no heterogeneity and over 50% indicated significant heterogeneity. Subgroup analysis for studies that included only laparoscopic appendicectomy was considered if the outcome was reported in three or more studies. To check possible causes of heterogeneity and evaluate the robustness of the results, sensitivity analysis was performed by calculating the risk ratio or RD for dichotomous variables. Moreover, the analysis was repeated leaving one study out each time to check the effect of each study on heterogeneity.

Methodological appraisal of included studies
The methodological appraisal of the included observational studies [26,29,30,32,33,[34][35][36][37][38][39][40][41] is illustrated in Table 4. The risk of bias was judged as low in two studies [36,39] and moderate in the rest of the observational studies. An overview of the risk of bias of the included RCTs [25,27,28,31,34] is shown in Figures 1, 2. None of the included RCTs reported random sequence generation and allocation concealment resulting in an unclear or high risk of selection bias. Moreover, there was an unclear risk of performance and detection bias due to a lack of blinding. Attrition bias and reporting bias were judged to be unclear in one study [27]; however, the rest of included randomized control trials (RCTs) showed a low risk of bias in these domains.  Symbol (*) is the number of points given to each study according to the Newcastle-Ottawa scale (* gives one point, and ** gives two points).

Abdominal Collection
The postoperative abdominal collection rate was reported in 16 studies with a total number of 4,007 patients; 10.5% of them developed postoperative abdominal collection (Figure 4). The difference was not statistically significant between the drain and no-drain group (

Subgroup for Open and Laparoscopic Appendicectomy
The

Subgroup for Paediatrics-Only Studies
Analysis of the studies that include only paediatrics patients, seven studies with 2,987 children, revealed that the rate of postoperative collection is statistically lower in the no-drain group compared to the drain group (4.7% vs 10.2%, OR=1.91, 95% CI (1.22, 2.99), P=0.005) ( Figure 6).

Postoperative Complications
The measured postoperative complication outcomes were SSI, faecal fistula, bowel obstruction and paralytic ileus (Figures 7-10

Subgroup for Laparoscopic Appendicectomy
Subgroup analysis for studies that include only laparoscopic appendicectomy cases for SSI was consistent with all studies analysis (OR=2.18, 95% CI (1.05, 4.52), P=0.04).

Length of Hospital Stay
Length of hospital stay was mentioned in 12 studies including 3,468 patients ( Figure 11). The no-drain group showed a statistically significant shorter length of hospital stay (LOS) when compared to the drain group (MD=1.79, 95% CI (1.25, 2.34), P=0.00001). A high level of heterogeneity was observed between included studies (I 2 =88%, P=0.00001).

Mortality
Mortality was reported in five studies with a total number of 656 patients ( Figure 12). There was a difference regarding mortality rate between the two groups (3.4% in the drain group vs 0.5% in the no-drain group, RD=0.01, 95% CI (-0.01, 0.04), P=0.36). The level of heterogeneity was low between the included studies (I 2 =37%, P=0.18).

Sensitivity analysis
There was no change in the direction of the pooled effect size when RD or RR was applied for dichotomous outcomes. Similarly, excluding one study at a time analysis has not shown significant discrepancies with the original analysis.

Discussion
Acute appendicitis is a very common condition with a lifetime risk of 7% [2]. Although appendicectomy is recognized as one of the commonest surgical procedures, there is around a 9% chance of having a complication related to appendicectomy and mortality ranges between 0.24% and 4% [7,[16][17]. SSIs including wound infection and postoperative collection are the commonest to happen after complicated appendicitis compared to simple appendicitis [8,14,15]. However, there is another wide range of postoperative complications related to complicated appendicitis such as ileus, obstruction, stumpitis, dehiscence, fistula formation and rarely death [42].
Acute uncomplicated appendicitis is a term usually used to describe a localized inflammation of the appendix with no localized or generalized effect or extension to adjacent structures [10][11][12][13][14][15][16]. As a rule of thumb, complicated appendicitis includes perforated appendicitis which can include localized contamination, abscess formation and single or more quadrant contamination with pus or faeces; it also includes appendicular mass, fistulation and intestinal obstruction [43]. In 2012, a proposed intraoperative classification system was proposed by Gomes et al. [44]. This classification ranged from grade 0 being a normal looking appendix to grade 5 being diffuse peritonitis with a range of the severity of the inflammation in-between. Despite that, this classification has not been adopted widely.
Our meta-analysis has found no significant difference in the postoperative intra-abdominal collection between the two groups which is about the same in most of the included studies [35][36][37]39]. However, Song and Jung and Pakula et al. have found a statistical difference in high-rate intra-abdominal collection and abscess formation in their drain group patients [32,33]. This variation in the rate of intra-abdominal/pelvic collection postoperatively may attribute to many reasons such as too small a drain diameter to drain thick fluids, infection introduced by drain materials or blockage of the drain by a clot or debris [38].
The rate of the intra-peritoneal collection after complicated appendicitis was reported to be 8.6% in the literature; hence, the use of a surgical drain following surgery has shown to be of no significant benefit for complicated appendicitis, which is consistent with our meta-analysis finding (10.1%) [45]. Otherwise, in other surgeries where a high risk of the intra-abdominal collection can result in a high morbidity and mortality rate, a surgical drain should be used.
The commonly reported risk factors for collection are perforated appendicitis, high total leucocyte count and duration of symptoms [46,47]. Irrigation of the peritoneal cavity was found to be an increasing risk of development of the intra-abdominal collection, and the pus can spread during washout from the localized area of infection in the right iliac fossa to the other compartments in the abdomen resulting in generalized peritonitis [47]. Moreover, a recent systematic review concluded that suction only is not inferior to suction and irrigation in terms of the development of the postoperative collection and the adoption of a suctiononly approach may reduce the operative time [45,48]. Assessment of the effect of these known risk factors for the collection was not possible in this review due to the lack of data in the included studies.
Interestingly, the incidence of the post-appendicectomy collection was found to be higher in laparoscopic appendicectomy when compared to the open approach in the Cochrane meta-analysis that was performed in 2002; however, in the subsequent update of the same review in 2004 and 2010, the likelihood of intraabdominal collection decreased from OR=2.77 in 2002 to OR=1.87 in 2010 [48,49].
Several reports suggested that drain can be one of the causes of intestinal obstruction and paralytic ileus during the postoperative period, which eventually can result in a delayed postoperative recovery [50]. This can be either mechanical obstruction due to the presence of the drain as a foreign body inside the abdominal cavity or due to infection and subsequent scarring that can be introduced through the drain from outside the abdomen [51]. This review showed statistically significant differences in the rate of postoperative bowel obstruction, ileus and fistula formation in the drain group compared to the no-drain group. This is significantly related to the high morbidity rate and delayed recovery noted in the drain group. The development of postoperative enteric fistula after appendicectomy represents only 0.5% of cases; this can be due to either a leak from appendicular stump, iatrogenic injury to the caecum or presence of distal bowel obstruction [52,53]. As per the results that previously mentioned and analysed in this review, the absence of a drain was of significant benefit for the length of hospital stay (LOS) than for the use of a drain. The hospitalization is significantly shorter in the no-drain group which has a positive effect on patient quality of care and the cost of the health system in general. In addition to that, early discharge prevents patients from hospital-acquired infections and promotes psychological well-being [54]. We can argue that patients with complicated appendicitis and drain insertion generally would have a higher risk of developing postoperative complication and prolonged hospital stay.
Prophylactic drain after abdominal operations has been studied extensively in the literature. A Cochrane review concluded that it is unclear whether routine abdominal drainage had any effect on the reduction of mortality at 30 days or postoperative complications after pancreatic surgery [55]. Regarding colorectal surgery, the recommendation of a recent meta-analysis of randomized controlled trials was against the usage of the drain as there was no difference in anastomotic leak rate, morbidity and mortality between the drain and no-drain groups [56]. Moreover, the recommendation remains against the drain after laparoscopic cholecystectomy and liver resection [57,58].
The present meta-analysis is the most comprehensive meta-analysis of the literature to date. In 2015, a Cochrane review of five studies compared drain versus no drain in complicated appendicitis [59]. The authors reported that there is no difference between the two groups regarding the intra-abdominal collection, SSI and mortality, which is in line with our results. However, the authors have not compared the outcomes regarding bowel obstruction, fistula, ileus and antibiotics duration. This should be taken into account that our pooled population size was much larger than the aforementioned meta-analysis, indicating that our findings may be less subject to type 2 error.

Limitations
There are obvious limitations in included studies as most of them are observational studies and not welldesigned RCTs which are inevitably subject to selection bias. Due to the lack of available data, we were not able to conduct subgroup analysis for the types of complicated appendicitis (perforated appendicitis, appendicular mass, etc) or the immunocompromised and age extremities patients (paediatrics vs elderly) as these factors might have biased our findings in favour of an intervention. Moreover, the reported outcomes by the included studies, except for bowel obstruction, fistula and ileus, have been heterogeneous. Authors recommend to shed more light on the lack of a uniform definition or proper scoring system for complicated appendicitis in the literature; hence, an agreement about setting a clear definition and scoring system for complicated appendicitis should be adopted for the future research and trials in order to meet the demand of the fast-growing medical care.

Conclusions
This meta-analysis has shown that drains do not affect the development of intra-abdominal collections in complicated appendicitis, but they can significantly increase the risk of postoperative complications such as fistula, surgical site infection (SSI), bowel obstruction, ileus and length of hospital stay (LOS).

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.