Irritable bowel syndrome (IBS) is a functional gastrointestinal (GI) disorder, characterized by chronic abdominal pain and altered bowel habit, with a benign natural history. With a worldwide prevalence approximating 10–15%, it is a common disorder and one of the most frequent sources of referral to gastroenterology [
A subset of patients with IBS have diarrhea and abdominal pain as predominant symptoms, presenting similarly to those with inflammatory bowel disease (IBD), thus making the clinical distinction challenging based on symptoms alone [
Clinical indices used to accurately identify inflammation in patients with symptoms exist but are not discerning for IBD and have not been incorporated into routine diagnostic investigation [
The use of point of care ultrasound (POCUS) for noncardiac purposes in medicine by nonradiologists has increased significantly in the last decade [
Consecutive patients presenting for assessment of symptoms including abdominal pain and diarrhea suspicious for IBD to the University of Calgary GI clinic were selected for this pilot study from January 1, 2013, to June 1, 2015. Patients with significant abdominal obesity (BMI > 40), pregnancy, liver disease with ascites, or conditions of previously known luminal inflammation were excluded. Informed consent was obtained prior to the examination and the study was approved by the Conjoint Health Research Ethics Board guidelines.
All patients underwent a physical examination as per standard of care. Laboratory inflammatory markers including C-reactive protein (CRP) were recorded where available, within 2 months of POCUS. Fecal calprotectin was not available at the time of the study at this center. A gastroenterologist (KN), with formal training in US of the GI tract and 3-year experience in the modality, conducted a focused luminal POCUS examination. A standardized approach was used, conducting the examination from the left lower quadrant, examining the colon from the rectum to the cecum, with examination of the rectum, sigmoid, descending, transverse, and ascending colon, and cecum. The terminal ileum was then evaluated, followed by systematic four-quadrant examination to include the remaining small bowel. All exams were completed using a Phillips IU22 US machine utilizing a range of transducers, including high frequency curved (4–9 mHz) and linear probes (12–15 mHz).
Presence of inflammatory activity was documented in binary fashion (“active”/“inactive”). Inflammation was deemed to be present, as per standard sonographic assessment, if there was increased bowel wall thickness (>3 mm for small bowel, >4 mm for large bowel) and the presence of any additional established indicators of inflammation: mesenteric inflammatory fat, lymph nodes, and hyperemia or blood flow as detected on color Doppler imaging [
Patients with normal blood work, negative endoscopy, and histology and who met Rome III criteria for IBS were identified as having IBS (diarrhea predominant) [
All patients underwent ileocolonoscopy after POCUS. The time interval to routine or expedited endoscopy, determined based on clinical suspicion of disease, was recorded. Extent of the endoscopy was recorded including intubation of the terminal ileum or lack of as well as the presence of disease activity, which was recorded in a binary fashion (present/absent), along with documentation of the etiology of the inflammation (confirmation of Crohn’s disease or ulcerative colitis or other suspected etiology) and, if present, the anatomic location. Where clinically indicated, biopsies were taken in both the small and large bowel and results (presence/absence of inflammation) were recorded. If biopsies were taken as part of the standard evaluation, this too was recorded as confirmation of organic disease or its absence.
A subset of patients (14/58) completed a 12-question patient satisfaction questionnaire, with graded responses following the POCUS examination in clinic. The questionnaire evaluated patient satisfaction with the overall experience, patient perception of the value of POCUS, and the willingness to incorporate POCUS as an extension of the physical examination in clinic.
Demographic data and ultrasound quality were evaluated using descriptive statistics. Further analysis was aimed at evaluating the ability of POCUS to detect inflammation in symptomatic patients compared with standard colonoscopy; again, colonoscopy was considered the gold standard. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and the positive and negative likelihood ratios (LR) with 95% confidence intervals (CI) were calculated for the overall POCUS score, compared to endoscopy only. To further evaluate POCUS, the above-mentioned statistics were stratified for segmental analysis, either ileum or colon. Accuracy of the detection of disease location, if present, was also evaluated. In cases where the ileum was not seen in endoscopic examinations, correlations were made between the endoscopic and US segments of the colon alone. Finally, the ultrasound examination quality was measured descriptively. Histology was reported; however, it was not used as the gold standard.
A total of 58 patients presenting to the clinic with symptoms were included. Demographic details are presented in Table
Patient demographic and laboratory investigation data (change title).
|
|
---|---|
Gender | |
Male | 19 (33) |
Female | 39 (67) |
Median age (years) | 32.8 (17.8–72.4) |
Median CRP (mmL, range) ( |
3.4 (0.6–36.7) |
Median time between POCUS and CRP (days) | 39.7 (0–127) |
Median time from POCUS to endoscopy (days, range) | 30 (0–149) |
Adequate POCUS exam quality | 56 (96.5) |
TI intubation on endoscopy | 56 (96.6) |
Inflammatory activity was detected sonographically (positive study) in 9/58 (15.5%) ranging from mild to severe inflammation (Tables
Descriptive ultrasonographic and endoscopic data.
Ultrasound |
|
---|---|
Active inflammation on US | 9/58 (15.5) |
Increased bowel wall thickness | 9/9 (100) |
Lymph nodes | 9 (100) |
Inflammatory fat | 7 (78) |
Hyperemia | 3 (33) |
Preserved wall layers | 10 (100) |
Complications+ | 1 (11%) |
Active site on endoscopy ( |
|
Ileum | 4 (40%) |
Colon | 5 (50) |
Ileocolonic | 1 (10%) |
+Complications identified include any of abscess, stricture, phlegmon, or inflammatory mass.
A
Positive endoscopy | Negative endoscopy | |
---|---|---|
Positive US | 8 | 2 |
Negative US | 1 | 47 |
Sensitivity, specificity, PPV, and NPV of POCUS relative to endoscopy.
Overall | Ileum | Colon | |
---|---|---|---|
Sensitivity% |
80.0 |
100.0 |
60.0 |
Specificity% |
97.8 |
98.2 |
100.0 |
PPV% |
88.9 |
83.3 |
100.0 |
NPV% |
95.7 |
100.0 |
96.36 |
Positive LR |
36.8 |
56.0 |
— |
Negative LR |
0.20 |
— |
0.40 |
First suggestion of terminal ileal Crohn’s disease identified on sonography, with thickened distal ileum running over the hypoechoic iliac artery. The white line marks the thickened ileal wall with echogenic or white air in the lumen.
Normal sigmoid colon with normal haustral folds (white arrow).
POCUS exhibited a PPV of 88.9% (95% CI: 51.75–99.72) and a NPV of 95.7% (95% CI: 85.5–99.5). The overall sensitivity and specificity of POCUS were 80% (95% CI: 44.4–97.5) and 97.8% (95% CI: 88.5–99.9), respectively, using a cutoff for bowel wall thickness of 3 mm for the ileum and 4 mm for the colon, with positive and negative LRs of 36.8 and 0.20 (Table
Of the 10 patients with confirmed inflammatory activity on endoscopy, all had CRP (7/10 within 1 month) completed, and 5 were elevated while the other 5 were within normal range. The median CRP value for these 10 patients was 6.9 mg/L. In those with elevated CRP, none of the US examination, confirmed endoscopically, revealed severe disease while all those 3 cases with moderate to severe sonographic disease activity had CRP within normal range, measured within 40 d of the US.
All 14 patients who completed questionnaires described the preparation for US as “easy” and reported an increased understanding of their disease and disease location following the exam. All of these patients believed POCUS had value in guiding disease management and found the increased clinic time acceptable and beneficial to their health.
To our knowledge, this is the first prospective study to evaluate the accuracy of gastroenterologist-performed POCUS in excluding inflammation in symptomatic patients compared to colonoscopy in North America. The accuracy established here is similar to previous published data [
The rate of pathologically proven histologic inflammation in this study, either microscopic colitis or inflammatory bowel disease, was 20/58 (34%). This figure is much higher than that quoted for larger populations of patients with IBS including all subtypes of IBD. Given their symptoms, this population was identified as “high risk” and in need of ileocolonoscopy as part of standard of care investigation and the high pathology rate likely reflects some bias or overrepresentation of pathology, given the convenience sampling from the “high-risk IBD” clinic. However, gold standard evaluation with endoscopy is not always indicated in IBS patients, as it is resource intensive and carries the associated risks of an invasive examination (including patient tolerance) [
The accuracy of detecting disease in the terminal ileum on cross-sectional imaging tends to be greater compared to the colon, as was exhibited in this study on US [
Nine patients with both negative US and endoscopy had microscopic colitis (MC) confirmed histologically. Five of these patients (55%) exhibited extensive small bowel fluid filled loops, a novel finding suggesting possible pathology of the small bowel as well (Figure
Image quality has been suggested to be more variable in US compared to MR or CT, potentially limiting US in some populations, specifically in those with high BMI (>35) with increased abdominal wall adiposity. Only 2 patients in this study had poor quality examinations due to high BMI; however, the US examinations were still deemed sufficient. There are no currently existing training guidelines for gastroenterologists to guide minimal requirement for competency in luminal sonography in Canada, given the evolution of this emerging modality in the field. However, it is a standard part of training in many European countries including Germany and Italy, where at least 100 examinations are recommended [
Patient engagement and satisfaction are important components of establishing rapport and trust and are crucial for long-term success in the management of both chronic functional disorders and inflammatory conditions. US improved patient understanding and was well tolerated and was generally preferred over invasive endoscopy. This tool may prove to be important in facilitating understanding and reassurance for patients at diagnosis and is a radiation-free and relatively inexpensive modality that can be used serially for disease monitoring in established IBD. The sample reflecting patient satisfaction was however small, in this early pilot study.
There are a number of limitations in this study, as it is small and is limited to one academic center with one sonographer performing bedside US examinations. These limitations should of course be kept in mind when considering the generalizability of this study to other centers. This is, however, a novel modality in gastroenterology, and training guidelines along with reporting and quality frameworks must be established. In addition, this preliminary analysis is underpowered; therefore, a single error, false-positive or negative, will significantly adversely impact the sensitivity and specificity in this small sample. Thus, prospectively collected larger studies are necessary to understand accuracy of the modality with and without additional clinical factors such as clinical predictors and inflammatory markers.
The time interval between US and endoscopy was variable with few endoscopic examinations completed more than 3 months after the POCUS. This likely reflects real life practice, however, as these patients were clinically suspected to have IBS and investigation for most of them was triaged as nonurgent. There were no medical interventions initiated for any patients; therefore, no structural changes should result between imaging and endoscopic evaluations. Finally, the sonographer in this study was also performing ileocolonoscopy, not blinded to the US results, which may bias the commentary on endoscopy. In order to minimize this bias, nearly all (94%) had biopsies taken to definitively exclude unrecognized inflammation.
Point of care US is a safe, noninvasive means of accurately detecting inflammation in patients being investigated for symptoms of diarrhea and abdominal pain. It is timely, as it occurs at the bedside during clinical assessment. It is well accepted by patients and can be employed to further direct definitive diagnosis and management and possibly obviate the need for endoscopic evaluation in certain cases. Given the high positive likelihood ratio with a positive exam, it can be effectively used as a triage tool to expedite endoscopy.
The authors declare that they have no competing interests.