Article Text

Download PDFPDF

Conservative treatment versus surgery for uncomplicated appendicitis in children: a systematic review and meta-analysis
  1. Ulf Kessler1,2,
  2. Selim Mosbahi1,
  3. Benedict Walker1,
  4. Eva M Hau2,
  5. Michael Cotton3,
  6. Barbara Peiry1,
  7. Steffen Berger2,
  8. Bernhard Egger1
  1. 1 Department of Surgery, HFR Fribourg - Cantonal Hospital, Fribourg, Switzerland
  2. 2 Department of Paediatric Surgery, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland
  3. 3 Department of Accident and Emergency, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland
  1. Correspondence to Dr Ulf Kessler, Department of Paediatric Surgery, Bern University Hospital, University of Bern, Inselspital, Bern 3010, Switzerland; ulf-kessler{at}hotmail.com

Abstract

Objectives To compare conservative treatment with index admission appendicectomy in children with acute uncomplicated appendicitis.

Design Systematic review and meta-analysis.

Data sources Medline, Embase and the Cochrane Library (CENTRAL) from 1950 to 18 February 2017.

Eligibility criteria for selecting studies Studies that assessed both appendicectomy and non-operative management of acute uncomplicated appendicitis in children of less than 18 years of age. Endpoints were postintervention complications, readmission and efficacy (successful outcome of the initial therapy).

Results Five studies met the inclusion criteria (conservative treatment n=189; surgical intervention n=253). Compared with patients undergoing index admission appendicectomy, conservative treatment showed a reduced treatment efficacy (relative risk 0.77, 95% CI 0.71 to 0.84; p<0.001) and an increased readmission rate (relative risk 6.98, 95% CI 2.07 to 23.6; p<0.001), with a comparable rate of complications (relative risk 1.07, 95% CI 0.26 to 4.46). Exclusion of patients with faecoliths improved treatment efficacy in conservatively treated patients. One study was randomised, with the remaining four comprising cohorts assembled by patient or physician choice. Different antibiotic regimens were used between investigations. Follow-up varied from 1 to 4 years.

Conclusions Conservative treatment was less efficacious and was associated with a higher readmission rate. Index admission appendicectomy should in the present still be considered to be the treatment of choice for the management of uncomplicated appendicitis in children.

  • Appendicitis
  • Appendicectomy
  • Conservative Treatmant

Statistics from Altmetric.com

What is already known on this topic?

There is no clear evidence as to the relative advantages or disadvantages of taking a conservative treatment approach for acute appendicitis in children.

What this study adds?

Since this meta-analysis could identify no advantage for conservative treatment, index admission appendectomy should remain the treatment of choice for children with uncomplicated appendicitis.

Introduction

Appendicectomy bears a substantial risk of both perioperative and postoperative complications. Various meta-analyses, including an increasing number of well-designed studies, have provided evidence in support of a role for conservative treatment in the management of uncomplicated appendicitis in adults.1 2 A recent review of practice in Ireland revealed that almost a quarter of consultant surgeons routinely treat uncomplicated appendicitis conservatively.3 In contrast, studies to address this topic in children are rare. To date, no clear evidence has been provided in the literature, either as a systematic review or meta-analysis, as to the relative advantages or disadvantages of taking a conservative treatment approach for acute appendicitis in children. Moreover, in the existing literature, there appears to be a striking breadth in the variety of study design, treatment modalities and results, which further complicates data comparison.4 5

The result of this knowledge gap is uncertainty as to the status of appendicectomy versus conservative treatment for children. Does appendicectomy still represent the ‘gold standard’ treatment for acute appendicitis in children? Alternatively, should its status be questioned, as it has been for adult patients? In clinical paediatric practice, there is the added pressure of increasing demands from both patients and physicians to avoid surgery if not absolutely necessary. Currently, given the paucity of data, there is no clear consensus as to the best approach for a child diagnosed with uncomplicated acute appendicitis.

The aim of our systematic review and meta-analysis was therefore to compare outcome parameters for childhood appendicectomy versus a conservative, non-surgical approach.

Methods

Our study was performed in accordance with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) and the strengthening the reporting of observational studies in epidemiology (STROBE) reporting guidelines.6 7 All published data that met the inclusion criteria (see below), regardless of study design, were considered.

Eligibility criteria

We included those studies that assessed both appendicectomy and the non-operative management of acute uncomplicated appendicitis in children of less than 18 years of age. Studies that included patients with complicated appendicitis (abscess, perforation) at initial presentation for appendicitis (index admission), and that did not report data from uncomplicated disease separately, were excluded. We also excluded studies that included adults or other groups with a potentially worse outcome (ie, patients with faecoliths only, and children immunocompromised owing to chemotherapy or other disease). While we set no language restrictions for our initial search, we did not consider manuscripts published in languages other than English, German or French. We also excluded case reports, duplicate publications and studies that failed to present original data. We subsequently assessed the publications cited by the included studies for any relevant material for possible inclusion.

Information sources and search strategy

We searched the following databases: Medline via PubMed (from 1950 to 2017), the Ovid Embase (from 1950 to 2017) and the Cochrane Library (CENTRAL) (from 1950 to 2017). Using Boolean operators AND and OR, we used all possible combinations of the search terms: ‘child, children, adolescent, infant, toddler, neonate, pediatric, paediatric, appendicitis, appendectomy, appendicectomy, nonoperative, non- operative, conservative, and antibiotic’. We performed the meta-analysis in accordance with the recommendations of the Cochrane Collaboration, using the Grading of Recommendations Assessment, Development, and Evaluation system for methodological quality assessment (http://methods.cochrane.org/gradeing/gradepro-gdt). We performed our literature search and trial selection according to a predefined review protocol. The latest update of this literature search was performed on 18 February 2017.

Study selection

Three authors (SM, BW and UK) independently screened the publications identified in our initial screen for eligibility on the basis of their titles and abstracts; a 90% agreement rate was achieved following their comparison. Where disagreements arose, the full text was assessed by either a paediatric or visceral surgeon (BP, SB or BE), and then discussed prior to reaching a consensus decision.

Data collection process

Data extraction was performed by three authors (EMH, SM and BW). Extracted data were double-checked (UK). Disagreements were resolved with help of three reviewers (BP, SB and BE). We extracted data (in duplicate) regarding general study parameters, diagnostic measurements, treatment regimens, outcome and complications. The following parameters were collected: study design, study group allocation (randomisation, parental choice), power analysis, blinding, age of inclusion, total number of patients, number of conservatively treated children, number of patients undergoing emergency appendicectomy, number of patients having undergone ultrasound or CT scan, detailed inclusion criteria (clinical parameters, symptom duration, laboratory or imaging parameters), inclusion or exclusion of patients with faecoliths on imaging, exclusion of patients with complicated appendicitis, histological results (normal appendix, perforated, other), failure of antibiotic treatment (definition, proceeding and number), recurrence of appendicitis after antibiotic treatment (definition, proceeding and number), complication of antibiotic treatment (type and number), protocol for conservative treatment (type and duration of antibiotic treatment, nutrition or duration of fasting, protocol violations, reasons for opting for oral versus intravenous treatment), intraoperative finding, postoperative complication, reoperation due to parental wish, duration of follow-up and other reported parameters (ie, quality of life, patient satisfaction, treatment costs). Any disagreements were resolved by discussion and subsequent consensus.

Outcomes

The first patient group received conservative (non-operative) treatment with antibiotics. The second group underwent an emergency appendicectomy. The primary outcome was the complication rate. We defined complication as the onset of complicated (perforated or gangrenous) appendicitis or peritonitis occurring after the primary intervention (see online supplementary appendix A). The secondary outcome measures were treatment efficacy and whether the patient required readmission. The compound outcome efficacy was defined to include broader aspects of the two treatment arms than only resolution of acute appendicitis, which will achieve complete success over a certain time. Efficacy of appendicectomy was therefore taken to be the absence of development of post-therapeutic or postoperative complications, including readmission. A successful conservative treatment was taken to indicate the absence of any of the following: failure of antibiotic treatment or recurrence of appendicitis requiring appendicectomy, development of major post-therapeutic or postoperative complications, including readmission. Should a secondary appendicectomy be performed according to parental wish without abdominal symptoms, conservative treatment was deemed successful. However, if secondary appendicectomy was performed in correlation with clinical suspicion of recurrence, conservative treatment was deemed unsuccessful, even in case of normal appendix.

Supplementary file 1

To assess the preintervention rate of perforated and therefore eventually misclassified patients in the conservatively treated group, we calculated the rate of preintervention complicated appendicitis (perforation, peritonitis or gangrenous appendix) found in the patients undergoing index appendicectomy.8

Outcomes were also assessed in a subset of conservatively treated patients where outcomes were distinctly reported for patients without faecoliths.

Data analysis

For statistical analysis, we used Stata V.13. We chose to use random-effect models (DerSimonian and Laird method) to calculate relative risk (RR) with 95% CIs as these models assume that the treatment effect varies between studies (reflecting inconsistency in surgical technique, selection criteria and patient profiles).9 For those studies that contained a 0 in one of the two compared outcome groups, we used Yates’ correction by adding the value of 0.5 to 0 cells, and to all other cells of the respective 2×2 table.10 11 We assessed between-study heterogeneity by calculating the I2 statistic.12 Statistical significance was assumed at the p<0.05 level if the CI did not include the value 1.

On an intention-to-treat basis, those patients initially treated conservatively who later required surgery were considered to be part of the conservatively treated group.

Results

Study selection

Our low stringency initial screen identified 3290 articles (Medline n=1172, Embase n=1918 and Cochrane CENTRAL n=200). Of these, 23 full-text articles were relevant, of which only 5 were ultimately eligible for inclusion (figure 1).13–17

Figure 1

Preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow chart of the study selection process.

Study characteristics and risk of bias

Study characteristics and treatment outcomes for the five studies included for meta-analysis, all of which reported treatment efficacy, complications or readmission rates (in both treatment groups, ie, conservatively and surgically treated patients), are shown together with their inclusion criteria in table 1 and figures 2–4. The report published by Hartwich and colleagues13 was excluded from the meta-analysis of complications given its absence of complications in either treatment group.

Table 1

Characteristics of trials included for meta-analysis, comparing conservative treatment versus appendicectomy in children with uncomplicated appendicitis

Figure 2

Conservative treatment versus appendicectomy: forest plot for complications. RR, relative risk.

Figure 3

Conservative treatment versus appendicectomy: forest plot for treatment efficacy. RR, relative risk.

Figure 4

Conservative treatment versus appendicectomy: forest plot for readmission. RR, relative risk.

Of a total of 442 children diagnosed with acute uncomplicated appendicitis, 189 received conservative treatment and 253 underwent index admission appendicectomies. Despite the evident importance of this subject, the quality of these single-centre studies was low (detailed under online supplementary appendix B). Only one study was randomised,16 with the remainder using treatment cohorts assembled according to patient and physician choice.13–15 17 Children under 5 years were neither included nor reported (table 1). Faecoliths were included in four investigations13 15–17 and excluded in one.14 Furthermore, conservative treatment regimens varied considerably between investigations (table 2), with the duration of follow-up ranging from 1 to 4 years (table 1).

Supplementary file 2

Table 2

Treatment regimens of trials included for meta-analysis, comparing conservative treatment versus appendicectomy in children with uncomplicated appendicitis

In conservatively treated patients, the rate of postintervention complications (2%) did not exceed the preintervention complication rate according to intraoperative findings (5%).

Synthesis of results

Overall, there were few postintervention complications with comparable range in conservative treatment (2%, 3/189) compared with index admission appendicectomy (2%, 4/253) (RR 1.07, 95% CI 0.26 to 4.46; figure 2).

Treatment efficacy rates were higher in surgical patients compared with the conservatively treated patients in all five studies. The overall treatment efficacy was 98% (249/253) in the surgical treatment group and 74% (140/189) in the conservative treatment group; that is, there was a significant reduction in efficacy following conservative treatment (RR 0.77, 95% CI 0.71 to 0.84; p<0.001; figure 3).

The risk of readmission was sevenfold greater for patients having undergone conservative treatment (95% CI 2.07 to 23.6; p<0.001; figure 4).

Outcome in conservatively treated patients with faecoliths (n=27) and without faecoliths (n=101) were detailed in three studies.15–17 Outcomes in patients who had no faecolith were superior to those that had a faecolith (online supplementary appendix C). In patients without faecoliths we found a trend towards less complications (RR 0.33, 95% CI 0.04 to 2.95), a trend towards increased efficacy (RR 1.33, 95% CI 0.87 to 2.02) and a lower readmission rate (RR 0.45, 95% CI 0.28 to 0.73) as compared with those with faecoliths.

Supplementary file 3

However the conservative approach in patients without a faecolith remained inferior to index admission appendicectomy with a lower treatment efficacy (RR 0.80, 95% CI 0.73 to 0.88; p<0.001) and a higher readmission rate (RR 6.28, 95% CI 1.44 to 27.5; p<0.05; online supplementary appendix D).14–17

Supplementary file 4

In addition to the defined outcomes, the study authors reported the positive effects of medical treatment in terms of fewer disability days,14 an improved quality of life13 and lower health-related costs.13 14 Conversely Mudri et al 14 found similar total hospital costs for both groups in correlation with a high failure rate of medical treatment. Tanaka et al 17 also reported on (higher) patient satisfaction following initial surgery.

Discussion

Our systematic review yielded five studies with which to compare the medical versus surgical treatment of children with uncomplicated appendicitis. We had three main findings. First, we identified major methodological limitations in the studies meeting our selection criteria. Second, we could identify no significant advantage for conservative versus surgical treatment on the basis of the assessed outcomes complications, efficacy and readmission. Third, we found that conservative treatment outcomes are compromised in the presence of faecoliths.

Strengths and weaknesses of the study

Despite the importance of the addressed question (namely, whether conservative treatment should be a consideration in children with uncomplicated appendicitis), the impact of the assessed data was surprisingly low.

One limitation is the reduced set of outcome parameters; we did not include other outcomes like length of hospital stay, return to normal activities or costs. The statistical power of our analyses is relatively low with few studies and patients, and even less in the analyses regarding patients with faecoliths.18 Other limitations of our study are partially determined by the quality of the research examined, which was of a low quality overall, with small patient cohorts rendering any subanalyses using meta-regression impossible (eg, for the presence of faecoliths). In addition, follow-up duration ranged from 1 to 4 years. Long-term complications such as adhesion obstruction may occur many years later, giving an obvious bias if follow-up is short. Furthermore, given that many complications of surgery may not warrant readmission, these might not be recorded, further skewing the data. None of the studies included in the meta-analysis treated or reported infants below 5 years of age. Also, results were not reported according to age categories. Considering the known increased perforation rate for infants under 5 years of age,19 conservative treatment in this subgroup may be even less efficacious and more hazardous than in older children. Moreover many ‘children’ over 14 years of age are structurally similar to adults, implying that strict age comparison (rather than, eg, the onset of secondary sexual characteristics) may be misleading.

Strengths and weaknesses in relation to other studies

We found no difference in the risk of postintervention complication between the treatment groups. However, in a recent meta-analysis of adults, there was a significant tendency towards reduced risk of complications in the conservative treatment group (39% for 1430 included patients),1 although a significantly lower treatment efficacy and an increased readmission rate were reported in those conservatively treated patients.1

In contrast to studies in adults, paediatric treatment studies evaluating an unusual therapy are especially difficult. Randomisation, especially in acute disease, is difficult, since parents tend to adhere to the best known medical treatment, instead of evaluation of alternative treatments. In the included studies one major source of study design-related bias was in the allocation of patients by parental or physician choice.20 21

On the one hand, exercising this option seemed to influence patient adhesion positively to the chosen protocol, as evidenced by the higher rate of requested secondary appendectomies (as per parental wish) in the single randomised trial; 6/24 (25%) conservatively treated patients subsequently opted for secondary appendicectomy.16 In the four studies that initially allocated patients to either a surgical or conservative treatment according to patient or physician preference, only 5/165 patients (3%) who initially chose conservative treatment subsequently opted for secondary appendicectomy.13–15 17

On the other hand, allocation to treatment groups according to parental or physician choice might have created a selection bias. Physicians might have also influenced patients’ choice if they felt that the patient might not respond (as in complicated appendicitis) or respond well enough (ie, for a non-pathological appendix) to conservative treatment.

An additional issue was that the histopathological data were inconsistently documented, making it difficult to differentiate between secondary surgeries due to recurrence or parental wish, versus a diagnosis such as non-specific pain.14 Additionally, patients readmitted with appendicitis-like symptoms may not be as rigorously evaluated as on their initial hospitalisation. For example, Hartwich et al 13 reported that imaging was not performed in all of the patients readmitted for appendicitis-like symptoms.

The selection of patients with uncomplicated appendicitis by means of refining inclusion and exclusion criteria is a difficult topic in adults, and even more so in children.21 Selection of patients for such studies creates a dilemma between inclusion of very few patients, where perforation is highly improbable, and inclusion of a larger study population that might be diluted with perforated cases. The rate of preoperative complications according to intraoperative findings in the included studies of 5% was relatively low as compared with studies in adults.21

Interestingly, a potential predictor of failure for conservative treatment that received little attention in study designs was the presence of faecoliths; four of the five assessed studies included patients with faecoliths.13 15–17 In adult patients Vons et al 21 recently reported that faecoliths were associated with complicated (perforated) appendicitis as well as with the failure of conservative management. Furthermore, a recent prospective trial on conservative management in children with appendicoliths was halted owing to an unacceptably high failure rate of 60%.22 Our results are in line with these findings. However, since index admission appendicectomy remained superior to conservative treatment, faecoliths are not the only factor that contributes to failure of conservative treatment.

Another important source of bias was the diversity of protocols for conservative treatment. Fasting as well as antibiotic regimens varied substantially between studies. In addition, none of the included studies addressed the question as to whether local resistance profiles influenced the antibiotic regimen, and no culture results were reported.

Implications for clinical practice

The complication and efficacy rates described in our meta-analyses of children differed from data collated for adults, where antibiotic treatment for uncomplicated appendicitis seemed to be a valuable option. Since evidence from the current literature is imprecise, there is no indication, as yet, on the basis of the published data, to abandon index admission appendicectomy as the treatment of choice for children with uncomplicated appendicitis.

Unanswered questions and future research

The pathogenesis of appendicitis is unresolved. It is thus not surprising that it remains unclear as to whether the conservative treatment including antibiotics and bowel rest in select paediatric patients might provide a reasonable alternative to appendicectomy. In order to resolve this issue, further analyses of pathophysiological aspects of the diseases as well as more stringent treatment trials will be necessary to overcome the shortcomings of the studies evaluated in our meta-analysis. Several trials are running on this topic, as the CONTRACT trial (CONservative TReatment of Appendicitis in Children a randomised controlled Trial) (http://www.nets.nihr.ac.uk/projects/hta/1419290) and the HAPPIEST trial (Hasselt APPendicitis Immunology and Environmental Cohort STudy) (https://clinicaltrials.gov/ct2/show/NCT02391675?term=appendicitis&rank=29).

Conclusion

As yet, there is no good evidence to support abandoning appendicectomy for uncomplicated appendicitis as the treatment of choice in children.

References

Footnotes

  • UK, SM, BW and EMH contributed equally.

  • Contributors UK conceptualised and designed the study, drafted the initial manuscript, performed and interpreted the statistical analysis, was part of the writing committee, critically reviewed and revised the manuscript, and performed the revisions as proposed by the reviewers. EMH, SM and BW performed the literature search, extracted the data, performed and interpreted the statistical analysis, were part of the writing committee, and critically reviewed the manuscript. MC interpreted the data, was part of the writing committee, and critically reviewed and revised the manuscript. BP conceptualised the study, was part of the writing committee and critically reviewed the manuscript. BE and SB performed the duplicate data extraction, analysed and interpreted the data, were part of the writing committee, and critically reviewed the manuscript for important intellectual content. UK, SM, BW, and EMH contributed equally to this work.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement No additional data arise from this study.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles