HOME
Return to Table of Contents
SYSTEMATIC REVIEW OF UTILITY OF MANNING AND ROME CRITERIA IN DIAGNOSING IRRITABLE BOWEL SYNDROME
AC Ford1, NJ Talley2, SV van Zanten3, N Vakil4, P Moayyedi1
1McMaster University, Hamilton; 2Mayo Clinic College of Medicine, Rochester, Minnesota, USA; 3University of Alberta, Edmonton; 4University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin, USA
INTRODUCTION: Lower gastrointestinal (GI) symptoms are common in the general population. Despite worries about underlying malignancy, the majority of these will be due to functional disorders, such as irritable bowel syndrome (IBS). There are a number of criteria for diagnosing IBS but the accuracy of these are unknown in clinical practice.
METHODS: The authors conducted a systematic review of the literature to identify studies that examined the accuracy of accepted diagnostic criteria (Manning, Rome I, II, and III) and computer models in diagnosing IBS. Searches of MEDLINE and EMBASE (until October 2007) were conducted. Studies were required to report prospectively on unselected cohorts of adult patients with lower GI symptoms attending for investigation (colonoscopy, barium enema, or CT colography). Those that compared the accuracy of specific diagnostic criteria or a computer model’s opinion with the results of lower GI investigation were eligible for inclusion. Two authors independently assessed studies for eligibility and extracted data in order to estimate sensitivity and specificity. Data were combined using a random effects model.
RESULTS: 10 papers were eligible, evaluating 2355 patients. Diagnosis reached by a computer model alone (6 studies, 1526 patients) had a sensitivity of 84% (95% confidence interval (CI) 74 to 92%) and a specificity of 88% (95% CI 77 to 96%) for detecting functional lower GI disease. Diagnosis reached by the Manning criteria (4 studies, 574 patients) had a sensitivity of 78% (95% CI 62 to 90%) and a specificity of 72% (95% CI 55 to 87%). Finally, diagnosis using the Rome I criteria (1 study, 602 patients) had a sensitivity of 71% (95% CI 66 to 76%) and a specificity of 85% (95% CI 80 to 89%).
CONCLUSIONS: Computer models, Manning or Rome I criteria are only moderately accurate in diagnosing IBS. There are no data on Rome II or Rome III despite the former being described eight years ago.