Surgical and non-surgical management for intussusception in children

  • Protocol
  • Intervention



This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To assess the safety and effectiveness of various contrast media, imaging modalities, pharmacological adjuvants, protocols for delayed repeat enema, and surgical approaches in the management of intussusception in children.


Description of the condition

Intussusception in children is a medical emergency that requires prompt diagnosis and management. It occurs when a segment of bowel (the intussusceptum) invaginates the lumen of another segment of bowel (the intussuscipiens). Both small and large bowel can be involved, but the most common kind of intussusception arises at the junction between the ileum and the caecum,-so-called ileocaecal intussusception (Loukas 2011). Untreated, intussusception may result in bowel perforation, peritonitis, and shock (Ko 2007). Mortality is increasingly rare (Buttery 2011; Davis 2003; Desai 2012; Parashar 2000) but remains a possibility (Iwase 2010; Meier 1996).

Intussusception is one of the most common abdominal emergencies for children less than the age of three (Applegate 2009). Its incidence varies from 0.33 to 2.4 per 1,000 live births (Bines 2002; Eng 2012; Fischer 2004; Huppertz 2006), though evidence suggests this rate is higher in developing countries (Ugwu 2000). Boys are more likely to be affected than girls (Bines 2002), and peak incidence occurs at ages between five and nine months (Daneman 2003; Samad 2012). Recent recipients of the current rotavirus vaccines (RotaTeq, Merck & Co; Rotarix, GlaxoSmithKline Biologicals) may be at slightly elevated risk for intussusception, as suggested by postlicensure monitoring in Mexico (Patel 2011) and Australia (Buttery 2011), but this elevated risk has not been observed in other countries (Ruiz-Palacios 2006; Shui 2012; Soares-Weiser 2012; Vesikari 2006; Yen 2012). A much stronger link between intussusception and an older rotavirus vaccine (RotaShield, Wyeth Laboratories) (Kramarz 2001; Murphy 2001; Peter 2002; Soares-Weiser 2004) led to its worldwide withdrawal 1999.

The cause of intussusception is often idiopathic (Staatz 1998), though any condition that produces pathological lead points (lesions in the bowel) can cause intussusception (Loukas 2011). Of these conditions, lymphoid hypertrophy seems to be the most common (Applegate 2009; Staatz 1998), implicating a viral or bacterial aetiology for most cases (Nylund 2010; Okimoto 2011; Parashar 2000; Staatz 1998). Other potential causes of pathological lead points include Meckel's diverticulum, duplication cyst, polyp, and lymphoma (Daneman 2003; Daneman 2004). Compared with idiopathic intussusception, intussusception caused by lead points is associated with poorer outcomes (Applegate 2009; Loukas 2011).

Diagnosis is challenging, since the symptoms of intussusception are wide-ranging and non-specific (Beasley 1988); the classic triad of symptoms most commonly associated with intussusception consists of vomiting, colicky abdominal pain, and bloody stool, but this triad occurs in less than half of cases (Blanch 2007; Lehnert 2009; Samad 2012). Three studies found that physicians correctly diagnosed intussusception in less than half of initial clinical encounters (Beasley 1988; Blanch 2007; Budwig 1994). Diagnostic delay increases the risk of surgical intervention (Lehnert 2009) and thus emphasizes the importance of prompt and effective management.

Description of the intervention

Non-surgical management of intussusception in children consists primarily of contrast enema (Applegate 2009; Daneman 2004; Ito 2012; Ko 2007). Contrast enema involves instilling contrast medium (usually air, saline, or barium) into the rectum via a rectal tube, with the aim of reducing the intussusceptum by increasing intraluminal pressure (Davis 2003). Fluoroscopy or, in the case of liquid contrast media, ultrasound can guide the procedure and monitor the reduction. Ultrasound avoids the radiation exposure associated with fluoroscopy and has the additional advantage of being an effective diagnostic tool (sensitivity = 98% to 100%, specificity = 88% to 100%) (Applegate 2009).

Pharmacological adjuvants can facilitate non-surgical management, but their efficacy remains controversial. For example, glucagon is an antispasmodic adjuvant used by 10% to 21% of surveyed practitioners (Cachat 2012; Katz 1992; Meyer 1992; Rosenfeld 1999). It provides analgesia (Lappas 1995) and reduces colonic muscle tone (Skucas 1994). However, a recent narrative review suggests that glucagon does not improve the rate of reduction in the non-surgical management of intussusception (Cachat 2012). Other commonly used adjuvants include antibiotics (Ein 2006; Moss 2000; Pepper 2012). Yet, one prospective study concludes that the actual risk of bacteraemia following fluoroscopically-guided air reduction is low (Somekh 1996). Two other studies found an elevated risk for intussusception following antibiotic administration (Hviid 2009; Spiro 2003), raising the concern for recurrence in cases where antibiotics are administered prophylactically.

Surgical management typically entails open laparotomy, though case-series and retrospective studies indicate laparoscopy may be just as effective and result in a shorter length of hospitalisation (Bailey 2007; Bonnard 2008; Kia 2005). Surgical management is generally indicated only if peritonitis, bowel perforation, or shock occurs; if appropriate radiological facilities are unavailable; or if contrast enema fails (American College of Radiology 2007; Daneman 2004). However, because non-surgical management is associated with lower morbidity and shorter hospitalisation (Bruce 1987), delayed repeat attempts at contrast enema may prove more beneficial than conversion to surgical management (Gonzalez-Spinola 1999; Navarro 2004; Sandler 1999; Saxton 1994).

Why it is important to do this review

Intussusception is a common abdominal emergency in children. Despite wide-spread agreement on the use of contrast enema for its non-surgical management, debate persists on the appropriate contrast medium, imaging modality, pharmacological adjuvant, and protocol for delayed repeat enema (i.e. duration of delay and number of repeated attempts) (Beasley 1998; Daneman 2004; Davis 2003; del-Pozo 1999; Littlewood 1998; Liu 1986; Schmit 1999). Debate also persists on the appropriate approach for its surgical management (i.e. open laparotomy versus laparoscopy). Prior reviews of non-surgical management exist (Al-Tokhais 2012; Applegate 2009; Cachat 2012; Daneman 2003; Ko 2007) but are narrative in nature and do not delineate the criteria used to assess studies. In contrast to narrative reviews, systematic reviews use transparent, objective, and reproducible methodology to locate and assess studies (Borenstein 2009). To the best of our knowledge, no systematic reviews of non-surgical or surgical management of intussusception currently exist.


To assess the safety and effectiveness of various contrast media, imaging modalities, pharmacological adjuvants, protocols for delayed repeat enema, and surgical approaches in the management of intussusception in children.


Criteria for considering studies for this review

Types of studies

All randomised control trials (RCTs) comparing contrast media, imaging modalities, pharmacological adjuvants, protocols for delayed repeat enema, and/or surgical approaches in the management of intussusception for children.

Types of participants

Any child who is clinically diagnosed as having intussusception by study authors. Although the Brighton Colloboration has established a validated and standardised case definition (Bines 2004a; Bines 2004b; Kohl 2008; Tapiainen 2006), it has only been used in the context of rotavirus vaccine postlicensure monitoring. As such, we will not require the use of Brighton Colloboration's case definition for inclusion of participants in this review.

Types of interventions

We will include all studies where different contrast media, imaging modalities, pharmacological adjuvants, protocols for delayed repeat enemas, and/or surgical approaches are compared.

Types of outcome measures

Where possible, we will extract the following primary and secondary outcome measures. Outcomes will be assessed at the time points reported by authors, unless otherwise noted. Furthermore, since recurring intussusception is associated with differing outcomes (Applegate 2009), we will conduct our assessment at the patient-level and not the case-level.

Primary outcomes
  1. Number of patients with successfully reduced intussusception. A successful reduction will be characterised by radiologically confirmed reflux of contrast media into the ileum.

  2. Number of patients with radiologically confirmed or clinically suspected bowel perforation or perforations.

  3. Number of patients returning with intussusception.

Secondary outcomes
  1. Number of patients who undergo a bowel resection.

  2. Number of patients diagnosed with sepsis (as confirmed by a positive blood culture).

  3. Radiation exposure (measured in millisieverts) resulting from intervention.

  4. Length of hospitalisation (measured in days) associated with intervention.

  5. Intraluminal pressure (measured in mm Hg) used to achieve reduction.

  6. Number of attempts required to achieve successful reduction.

  7. Length of operation (measured in minutes) in the case of surgical intervention.

  8. Number of any intraoperative complications (as defined by study authors) in the case of surgical intervention.

  9. Number of any postoperative complications (as defined by study authors) in the case of surgical intervention.

  10. Number of intraoperative conversions (i.e. laparotomy required) in the case of laparoscopic intervention.

  11. Time to resumption of full diet (measured in hours).

Search methods for identification of studies

Electronic searches

We will search the following electronic databases: the Cochrane Central Register of Controlled Trials (CENTRAL) (most recent), MEDLINE (via OvidSP) (1950 to present), EMBASE (via Ovid SP) (1980 to present), Science Citation Index Expanded (via Web of Science) (1900 to present), and BIOSIS Previews (1950 to present). We will impose no limitations on language. Duplicate publications will not be included in the review.

Our subject search in MEDLINE will follow the sensitivity-maximising version of the Cochrane Highly Sensitive Search Strategy (CHSSS) for identifying randomised trials in MEDLINE (Lefebvre 2011). We will perform a similar, sensitivity-maximising subject search in EMBASE (Wong 2006).

Search strategies are listed in Appendix 1 - Appendix 5 .

Searching other resources

We will investigate the reference lists of all eligible studies to identify further studies. To locate unpublished studies, we will contact content experts, search the World Health Organization's International Clinical Trials Registry Platform (ICTRP), and explore proceedings from meetings of organisations like the British Association of Paedatric Surgeons (BAPS) or the Radiological Society of North America (RSNA).

Data collection and analysis

Selection of studies

Two review authors will screen titles and abstracts for study eligibility using the inclusion criteria in this review. Where necessary, they will also read the full text of the paper or request additional data from the paper's authors. A third review author will adjudicate disagreements about study eligibility. Review authors will not be blinded to study details during this process. Non-English papers will be translated using Google Translate (Google 2012) or, where possible, a translator.

Data extraction and management

Two review authors will independently extract data and assess risk of bias using a standardised data extraction form. We will resolve disagreements by discussion with consensus agreement, involving a third review author if required. Review authors will not be blinded to study details during this process.

We will extract the following data.

  • General information: study author(s), title, source, contact address, country of study, language of publication, year of publication, any author conflicts of interest, study setting (e.g. hospital emergency department or specialised paediatric hospital).

  • Study characteristics and eligibility for review: study design, randomisation method, allocation concealment, recruitment method, duration of trial, study location, length of follow-up, operator allocation, any obvious concerns of bias.

  • Participants: inclusion and exclusion criteria, age, gender, presence of pathological lead points, anatomical location of intussusception, criteria used to diagnose intussusception, total number of participants, country of origin, number of dropouts or withdrawals and the reasons if recorded.

  • Interventions: number of participants for each intervention, a detailed description of the interventions and comparison interventions including, where relevant, the type, dose, concentration and duration of application.

  • Outcomes: specific outcomes reported and rates of recurrence, perforation, resection, sepsis, and, where applicable, operative complications and intraoperative conversions.

We will enter relevant data into Review Manager (RevMan) 5 software (RevMan 2011).

If any of these data are unclear or missing, we will make an attempt to contact the study authors.

Assessment of risk of bias in included studies

We will assess risk of bias using The Cochrane Collaboration's 'Risk of bias' tool as detailed in the section 8.5 in Cochrane Handbook for Systematic Reviews of Interventions (Cochrane 2011). In particular, we will assess the following domains: selection bias (due to inadequate random sequence generation or allocation concealment), performance bias (due to disparate operator experience with surgical and non-surgical reduction techniques), detection bias (due to inadequate blinding of outcome assessment and data analysts), attrition bias (due to incomplete outcome data), reporting bias (due to selective reporting), and other potential biases. Where we assess studies as having 'unclear risk' in any domain, we will contact authors to clarify the risk of bias. We will initially include all studies irrespective of the risk of bias. We will also perform a sensitivity analysis. If the sensitivity analysis shows substantial difference, we will exclude studies with high risk of bias from the review.

Measures of treatment effect

We will conduct our analysis following the guidelines set out in the Cochrane Handobok for Systematic Review of Interventions (Higgins 2011).

For dichotomous data, results will be presented as summary risk ratio (RR) with 95% confidence intervals (CI) and number needed to treat to benefit (NNTB) or number needed to treat to harm (NNTH) as appropriate. For continuous data, results will be presented as mean difference (MD) if outcomes are measured in the same way between trials. The standardised mean difference (SMD) will be used to combine studies that measure the same outcome but use different methods. For rate data, results will be presented as rate ratio with 95% CI. For survival data, results will be presented as hazard ratio (HR) with 95% CI.

Unit of analysis issues

Studies may feature multiple reduction attempts, repeated observations of their participants, or cluster randomisation. Where possible, we will correct for their effects following the suggestions in Chapter 16 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). In general, our assessments will take place at the patient-level and not the case-level. Where a unit of analysis issue may exist, we will perform a sensitivity analysis comparing them to studies without the same unit of analysis issue(s).

Dealing with missing data

In the case of missing summary or outcome data, we will seek these data from the study authors. Where data are not forthcoming, we will attempt to impute these data using the methodology outlined in Chapter 16 of Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

In the case of allocated participants missing in the reported results, we will, where possible, attempt to restore them to their appropriate group using an intention-to-treat (ITT) analysis. We will also seek clarification from the study authors.

No studies will be excluded for missing data, but we may perform a sensitivity analysis to determine their implications for the meta-analysis.

Assessment of heterogeneity

Where appropriate, we will use the I² statistic and Cochran Q statistics to assess heterogeneity (Borenstein 2009). We will deem heterogeneity to be substantial if the I² is more than 50% or Chi² P value is less than 0.10. In the case of substantial heterogeneity, we will perform prespecified subgroup and sensitivity analyses.

Assessment of reporting biases

If we suspect reporting bias, we will seek unreported data from study authors. If these data are not forthcoming and pose a high risk of bias, we will explore their implications for the meta-analysis by sensitivity analysis. In addition, we will use a funnel plot to check for publication bias if more than 10 studies are included in the meta-analysis (Egger 1998).

Data synthesis

We will analyse data using RevMan. For studies judged to have similar interventions, populations, and outcomes, we will use random-effects models and consider a P value of 0.05 or less to be statistically significant.

Subgroup analysis and investigation of heterogeneity

We expect the following areas to significantly contribute to study heterogeneity, and we will accordingly conduct subgroup analyses where necessary:

  1. Care setting. Different care settings, such as tertiary care centres, are associated with differing outcomes (Bratton 2001; Calder 2001; Rosenfeld 1999).

  2. Participants with confirmed presence of pathological lead point. The presence of these lead points is associated with differing outcomes (Loukas 2011).

  3. Participants with previous intussusceptions. Recurrance is associated with different patient characteristics and outcomes (Applegate 2009).

  4. Colonic structures involved in the intussusception. Intussusception involving different colonic structures (e.g. ileocaecal versus ileoileal) is associated with different outcomes (Loukas 2011).

  5. Studies with high risk of bias. These studies will be identified as having one or more domain judged 'high risk' by the risk of bias tool, as suggested in Chapter 8 of Cochrane Handbook for Systematic Reviews of Interventions (Cochrane 2011).

  6. Age. Children less than the age of one and greater than the age of three are more likely to possess pathological lead points (Applegate 2009).

  7. Geographical region. Regional differences in epidemiology, equipment availability, and operator experience exist (Beasley 1998; Liu 1986; Schmit 1999; Ugwu 2000).

We will assess differences among subgroups using analysis of variance (Altman 1996).

Sensitivity analysis

We will conduct sensitivity analysis where unforeseen or arbitrary decisions are made as per the guidance provided in section 9.7 of Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). Regardless of the outcome of analysis, we will report all results in our review, even if they were not included in the meta-analysis.


We would like to thank Jooly Joseph and Madan Mohan Palliyil for initiating this review. We would also like to thank the Cochrane Colorectal Cancer Group--especially Marija Barbateskovic--for their support and guidance.


Appendix 1. MEDLINE Search Strategy (via OvidSP)

1. exp Intussusception/
2. ((intestin* and invagination*) or intususcep* or intussuscep*).mp.
3. 1 or 2
4. exp Infant/
5. exp Child/
6. (infant* or child* or newborn*).mp.
7. 4 or 5 or 6
8. 3 and 7
9. randomized controlled
10. controlled clinical
11. randomized.ab.
12. placebo.ab.
13. clinical
14. randomly.ab.
15. trial.ti.
16. 9 or 10 or 11 or 12 or 13 or 14 or 15
18. 16 and 17
19. 8 and 18

Appendix 2. EMBASE Search Strategy (via OvidSp)

1. exp intussusception/
2. ((intestin* and invagination*) or intususcep* or intussuscep*).mp.
3. 1 or 2
4. exp child/
5. (infant* or child* or newborn*).mp.
6. 4 or 5
7. 3 and 6
8. randomized controlled trial/
9. randomization/
10. controlled study/
11. multicenter study/
12. phase 3 clinical trial/
13. phase 4 clinical trial/
14. double blind procedure/
15. single blind procedure/
16. ((singl* or doubl* or trebl* or tripl*) adj (blind* or mask*)).ti,ab.
17. (random* or cross* over* or factorial* or placebo* or volunteer*).ti,ab.
18. 13 or 10 or 14 or 16 or 9 or 15 or 11 or 8 or 17 or 12
19. "human*".ti,ab.
20. (animal* or nonhuman*).ti,ab.
21. 20 and 19
22. 20 not 21
23. 18 not 22
24. 7 and 23

Appendix 3. Cochrane Library Search Strategy

#1 MeSH descriptor Intussusception explode all trees
#2 ((intestin* and invagination*) or intususcep* or intussuscep*):ti,ab,kw
#3 (#1 OR #2)
#4 MeSH descriptor Infant explode all trees
#5 MeSH descriptor Child explode all trees
#6 (infant* or child* or newborn*):ti,ab,kw
#7 (#4 OR #5 OR #6)
#8 (#3 AND #7)

Appendix 4. Science Citation Index Search Strategy (via Web of Science)

#1 Topic=(((intestin* and invagination*) or intususcep* or intussuscep*))
#2 Topic=((infant* or child* or newborn*))
#3 Topic=((controlled trial or controlled clinical trial or placebo or clinical trial or random* or trial or cct or rct))
#4 (#3 AND #2 AND #1)

Appendix 5. Biosis Previews Search Strategy

#1 Topic=(((intestin* and invagination*) or intususcep* or intussuscep*))
#2 Topic=((infant* or child* or newborn*))
#3 Topic=((controlled trial or controlled clinical trial or placebo or clinical trial or random* or trial or cct or rct))
#4 (#3 AND #2 AND #1)

What's new

28 October 2013New citation required and major changesNew author team, major changes and new citation
5 October 2013AmendedCopy edited and ready for publication

Contributions of authors

All authors helped draft the protocol. In addition, RMG and TK designed the review, JK provided content area advice, and ACW provided methodological advice.

Declarations of interest

Richard McGee is a recipient of a postgraduate research scholarship from the National Health and Medical Research Council (NHMRC), Australia.

No other interests to declare.