This study showed that none of the 70 patients referred for a VCE, who had an undelivered initial PC, had gastrointestinal obstruction as proved by follow-up MRE or CTE examinations. Only 20 of the 70 went on to have a VCE.
During the three year study period a total of 714 VCEs were performed. Whilst performing a quality audit we came across a number of patients who had not gone on to VCE because the preceeding PC did not present itself in time. As evident from our results, 60 of 70 consecutive patients with undelivered PC had normal enteroclysis studies. What happened to these patients next? What was the clinician's understanding of our cross-sectional imaging results, and what impact had these on patient management? As no such information was found in the literature, we undertook this retrospective study to get a qualified estimate of the clinical value of our imaging performance.
VCE has proven itself to be the most effective examination in depicting small bowel mucosal lesions, including subtle signs of chronic inflammatory diseases  and obscure bleeding . New capsule enteroscopes designed for oesophageal and colonic exploration are being introduced to the market . In patients with a risk of capsule retention (4%), total enteroscopy is still possible in more than four out of five patients [20, 21]. Small bowel neoplasia remains rare but half of them might be revealed by VCE, i.e. VCE is diagnostically more efficient than any other imaging modality . The overall diagnostic success with VCE has allowed gastoenterologists to proclaim VCE to be a first line procedure in ruling out small bowel disease . However, the detailed depiction of the inner gut surface with VCE sometimes gives rise to interpretation difficulties as the specificity for detecting Crohn's disease is only 53% . Accordingly, erosions of the small bowel do not ultimately lead to a diagnosis of Crohn's disease . In the present study we encountered three such cases with alleged erosions in the terminal ileum, not confirmed at subsequent ileocolonoscopies.
We were able to make two interesting observations. One was that no organic lesion was diagnosed by us in 60 cases by MRE and CTE, i.e. the imaging results did not mirror the clinical suspicion of active small bowel disease. The other was that in most cases the gastroenterologist did not proceed with the primarily intended VCE examination, neither to confirm imaging results nor to verify their preliminary clinical suspicion. It was evident from the medical notes that results of MRE or CTE, together with a clinical re-evaluation of the patient's situation, gave the gastroenterologist sufficient data not to undertake further enteroscopies.
The reason for the low number of patients diagnosed with Crohn's disease with either CTE or MRE in our series is not known, but has probably to do with image resolution. Erosions, typical for Crohn's disease, are superficial and shallow; viz. lesions limited to the mucosa are not possible to unveil by these methods. However, the clinical value of modern enteroclysis examinations seems to be sufficient as few of the primarily intended VCE examinations were done. The true clinical value is still pending, but our follow-up of up to 24 months has yet not revealed any new disease, indicating the absence of false negative imaging studies.
In a patient with a biopsy-proven diagnosis of inflammatory bowel disease, future examinations of the bowel are more focused on complications than on confirmation of diagnosis. In a case like that, cross-sectional imaging with MRE or CTE ought to be performed before deciding on a VCE, saving the costs for at least a PC. Furthermore, a strategy that saves a VCE examination, especially when preceded by a PC, would save health care costs and minimize inconvenience.
In our department, patients under the age of 45 years are allocated to MRE, and older ones to CTE. As Crohn's disease mostly starts in a younger patient group, signs of active Crohn's disease are found more often with MRI than with CT. A diagnosis of Crohn's disease is established when verified endoscopically and proven by histology . Early diagnosis of is important as tailored treatment may reverse active disease and prevent complications. Younger patients with symptoms attributed to the small bowel are likely to harbour early and superficial lesions, that might be difficult to reveal on MRE and CTE examinations, but visible on VCE. A thorough diagnosis is difficult in early disease, mirrored in our series by the observation that referring doctors seemed to be more prone to send younger patients to a final VCE after MRE (15 cases) than after CTE (five cases). Elective imaging of the small bowel is done as an enteroclysis procedure in our institution, i.e. with a naso-jejunal tube, because this allows the radiologist to achieve optimal bowel filling for maximal diagnostic performance . Our examination techniques are on a par with those of leading European experts  which means that our results apply to MRI and CT enteroclyses and not to MRI or CT enterographies in which patients are given contrast by mouth.
All patients were selected and referred by gastroenterologists from one hospital. The fact that the sample population consisted of 2.5 times as many females (50/20) suggests that there might be an imbalance in the inclusion process of patients with a failed PC test. The skewed selection may partly be explained by the limited size of our sample population. Furthermore, it has been shown that women have longer bowel transit times than men, which may indicate that the acceptable PC transit time in women ought to be longer than the 48 h stated .
Given the limited size of our material sample, e.g. only eight cases of established Crohn's disease, solid statistical analyses of the temporal aspects of various diagnostic delays were not possible. However, patients diagnosed with Crohn's disease had examinations performed closer in time than the others. As might be expected, normality of the imaging studies in patients with mild to moderate clinical symptoms, i.e. absence of alarm symptoms, may also be explained by the fact that many of these patients were admitted to examination although their laboratory analyses were close to normal. A limitation of this retrospective study is that not all patients had a VCE. As only 29% of our patients had a final VCE we had to accept the clinical follow-up diagnoses as our "gold standard". One might comment that the absence of a final VCE proved our case, viz. gastroenterologists did not anticipate any further diagnostic information of importance for managing 71% of their patients besides the information obtained by the enteroclysis examinations.