• 
• 

Three Part Question

In [children with clinically suspected appendicitis] is [Magnetic Resonance Imaging accurate and reliable] in [confirming or excluding a diagnosis of acute appendicitis]?

Clinical Scenario

A 7 year-old boy presents to the emergency department with acute onset abdominal pain. His symptoms are not entirely classical for acute appendicitis but you are concerned this could be the start of its presentation and refer him to your surgical colleagues. The surgical registrar on call asks you to arrange an abdominal ultrasound for him which is carried out whilst he is waiting for a hospital bed. Unfortunately the report is returned as “unable to visualise the appendix”. The surgical team say they will adopt a “wait and watch” approach but his parents are unhappy with this as their son is in pain and ask you if there are any other investigations that can be performed. You are reluctant to request an abdominal CT due to radiation exposure and wonder if there is any evidence for the use of MR scanning in diagnosing appendicitis in children as you have seen it used to diagnose appendicitis in pregnant women.

Search Strategy

Ovid MEDLINE 1946 to January Week 4 2014

exp appendix / OR appendi$.mp. / OR exp appendicitis / OR exp appendectomy / OR appendectomy.mp. / OR appendicectomy.mp. / OR exp abdominal pain / OR abdominal pain.mp. / OR right iliac fossa pain.mp. / OR RIF pain.mp. / OR right lower quadrant pain.mp. / OR RLQ pain.mp. /

AND
exp magnetic resonance imaging / OR magnetic resonance imaging.mp. / OR MRI.mp. / OR exp magnetic resonance spectroscopy / OR magnetic resonance spectroscopy.mp. / OR MR scan.mp / OR MR.mp. / magnetic resonance tomography.mp. /
Limit to (English and humans and “all child (0 to 18 years)” and diagnosis (best balance of sensitivity and specificity)”).

Search Outcome

59 papers of which 3 were relevant to the search question

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Herliczek et al 2013 USA	Case review of 60 consecutive paediatric cases where MR scan of the appendix was performed within 24 hours of inconclusive US scan for suspected appendicitis. Age 7-17 (mean 13.4) M 28 : F 32 MR scan reports by 2 independent radiologists of different training levels blinded to the clinical outcome and initial MR report	1b cohort study with good reference standards applied to all patients	Comparison of reviewed MR scan result with operative and histopathology findings if child underwent surgery. Comparison of reviewed MR scan result with clinical outcome if child did not undergo surgery.	Sensitivity of MR = 100% (95% CI 0.72-1.0) Specificity of MR = 96% (95% CI 0.87-0.98) Positive predictive value = 83% (95% CI 0.55-0.95) Negative predictive value = 100% (95% CI 0.93-1) Overall test accuracy = 97% Inter-observer agreement κ = 1.00	Single centre, observational study Small sample of convenience. Unpowered study. Excluded patients who had MR scan >24 hours after US scan, potential selection bias Did not look at outcome for patients who had inconclusive US scan but no MR scan. No patient younger than 7 in study Did not use the same MR scanner or protocol for all patients Indications for performing investigations were at the discretion of the clinician. No standardised criteria used. However this is comparable to everyday practice.
Johnson et al 2012 USA	Prospective analysis of 42 children with suspected appendicitis and stable clinical conditions underwent MR scan and either US scan, CT scan or both within 24 hours of presentation Age 4-17 Images interpreted by 4 independent experienced radiologists who were blinded to outcome and the results of the other investigations	1b cohort study with good reference standards applied to all patients	Comparison of MR result with clinical outcome determined by result of CT / US, histopathology of surgical specimens and clinical follow up.	Sensitivity of MR = 100% (48/48, 95% CI 93-100) Specificity of MR = 99% (119/120, 95% CI 95-100) Positive predictive value of MR for diagnosis of acute appendicitis = 98% (48/49, 95% CI 89-100) Negative predictive value of MR = 100% (119/119, 95% CI 97-100) 0% failure rate of examination Diagnostic quality imaging in 100% of cases	Small, single centre observational study Convenience sample, study unpowered. No standardisation of diagnostic criteria for recruitment other than clinician’s opinion. However this is comparable to everyday clinical situations where the responsible physician decides whether to operate, investigate or observe. No standardisation of comparative imaging study. Despite study sensitivity of 100% there is a relatively large confidence interval for this value suggesting the sample size may be too small. The study did not directly look at the use of MR if first imaging was equivocal. The study did not discuss any comparison between the accuracy of MR scan and the other imaging modalities.
Moore et al 2012 USA	208 MR scans performed on patients aged 3-17 with clinically suspected appendicitis requiring imaging. MR scan used as first line investigation. Fast MR protocol with no contrast. MR available 24/7 Not preselected with USS	1b cohort study with good reference standards applied to all patients.	Correlation of MR finding with surgical or clinical outcome. All patients followed up.	Sensitivity of MR for diagnosing appendicitis = 97.6% (95% CI 87.1-99.9) Specificity of MR for diagnosing appendicitis = 97.0% (95% CI 93.2-99) PPV = 88.9% (95% CI 76-96.3) NPV = 99.4% (95% CI 96.6-99.9)	Single centre trial. Convenience sample, not powered. Confidence intervals relatively wide, especially for sensitivity and PPV indicating sample size may not be sufficiently large No set criteria or standardisation for patient selection. Decision to perform MR at discretion of surgeon. No inclusion or exclusion criteria given. Unknown if any children in the study period presented with symptoms and were operated on without prior imaging. 1 of 6 paediatric radiologists interpreted MR scan. May have been more reliable to dual report images and then assess for inter-relater reliability.
			Time from request of investigation to scan	Average time from request to examination = 78.7 minutes, median 65 min
			Time for scan	Average time for investigation = 14.2 minutes, median 12minutes
			Time from end of scan to report	Average time from end of examination to available report = 57min, median 46min
			Other diagnoses	MR also able to diagnose 2 other pathologies in the group who were negative for appendicitis

Comment(s)

Acute appendicitis is the most common emergency surgical presentation in children. In up to a third of cases diagnosis can be difficult and for these patients either a ‘watch and wait’ policy is adopted or further investigations are performed. In most developed countries guidelines recommend appendicectomy if clinical suspicion of appendicitis is high. Atypical presentations will undergo imaging, usually by way of Ultrasound. This approach can present several problems. Without definitive pre-operative diagnosis, despite high clinical suspicion, there are still a significant number of ‘negative appendicectomies’. The ‘watch and wait’ approach is a valuable tool, as children with appendicitis do not improve clinically. This does however have the drawback of potentially worsening a patients’ clinical condition, prolonging suffering and making surgery technically more challenging and risky with increased risk of post-operative complications. The waiting period can be frustrating for patient, parent and clinician, in addition to prolonging the duration of stay and added financial cost. Ultrasound is a useful investigation if the operator is able to demonstrate an acutely inflamed appendix. However if the appendix cannot be visualised, it cannot be said that the appendix is normal. It is a good rule-in test for appendicitis but not a very sensitive rule-out test. It is also good at identifying other pelvic pathologies, especially in girls. The reliability of ultrasound is dependent upon the skill and experience of the operator, as is the time taken to perform the examination. The position of the appendix, presence of ileus, patients’ body habitus and gender are also important. The examination can be quite painful and may be distressing for a child already in pain. In some institutions, Computed Tomography (CT) may be performed after equivocal ultrasound investigation. This is very sensitive for diagnosing appendicitis and is also able to identify other pathologies. However, it involves large does of radiation and often IV contrast, and therefore there is obvious caution regarding its use in all patients, but especially in children and pregnant women. Magnetic Resonance (MR) imaging is not a new modality but its availability, use and application is expanding. With significant recent technological advances, fast, contrast free images can be produced that give excellent delineation of anatomy, especially in acutely inflamed tissues. It also has the obvious advantage of being radiation-free and its use in the imaging of pregnant women with suspected appendicitis is well established. The three research studies appraised are all similar in that they are cohort studies that look at the ability of non- contrast MR scanning to accurately diagnose or rule out appendicitis in comparison to the reference standards of histopathological diagnosis after surgery or resolution of symptoms. They are all single centre studies that look at relatively small numbers of patients with inclusion criteria representative of clinical practice. The sensitivity and specificity of MR scanning for the diagnosis of acute appendicitis is seen to be high throughout, with sensitivity (rule-out) ranging from 97.6-100% and specificity (rule-in) ranging from 96-99%. It is interesting that all of the studies identified were carried out in The USA. No studies have been identified that look into the economic implications of MR imaging of suspected appendicitis in children. If MR is accepted as part of the pathway of investigating appendicitis in children, the financial implications, particularly for organisations such as the NHS will have to be considered.

Editor Comment

RB

Clinical Bottom Line

These studies provide evidence that MR imaging is sensitive and specific in the diagnosis of acute appendicitis in children. As the financial implications have yet to be studied and access to MR scanners is not widely available and accessible in the UK, it is recommended that MR is reserved for cases of equivocal diagnosis after US scan in children in preference to CT if available.

References

1. Herliczek TW, Swenson DW, Mayo-Smith WW Utility of MRI After Inconclusive Ultrasound in Pediatric Patients With Suspected Appendicitis: Retrospective Review of 60 Consecutive Patients American Journal of Roentgenology 2013;200: 969-973
2. Johnson AK, Filippi CG, Andrews T, Higgins T, Tam J, Keating D, Takamaru Ashikaga T, Braff SP, Gallant J Ultrafast 3-T MRI in the Evaluation of Children With Acute Lower Abdominal Pain for the Detection of Appendicitis AJR. American Journal of Roentgenology 2012;198: 1424-1430
3. Moore MM, Gustas CN, Choudhary AK, Methratta ST, Hulse MA, Geeting G, Eggli KD, Boal DKB MRI for clinically suspected pediatric appendicitis: an implemented program Pediatric radiology 2012: 1056-63
4. Moore MM, Gustas CN, Choudhary AK, Methratta ST, Hulse MA, Geeting G, Eggli KD, Boal DKB MRI for clinically suspected pediatric appendicitis: an implemented program Pediatric radiology 2012: 1056-63