Predictive value of the Rome criteria for diagnosing the irritable bowel syndrome

Our aim was to examine the predictive value of the Rome criteria and absence of so-called "red flags" of clinical practice for diagnosing irritable bowel syndrome. Red flags were relevant abnormalities on physical examination, documented weight loss, nocturnal symptoms, blood in stools, history of antibiotic use, and family history of colon cancer.

In retrospective studies, 98 patients who had one or more Rome criteria and lacked red flags were identified by chart review of a 1-yr period. In prospective studies, 95 patients were identified who met the Rome criteria and lacked red flags. Sensitivity, specificity, predictive value of Rome criteria, and absence of red flags were determined. Consultant's final diagnosis was the gold standard. Investigations before and after referral were recorded and reason for referral was determined in prospective studies.

In the retrospective series, the Rome criteria and absence of red flags had a sensitivity of 65%, specificity of 100%, and positive predictive value of 100%. None of these patients required revision of their diagnosis during a 2-yr follow-up. In the prospective study, the positive predictive value was 98%. More than 50% of the patients in this group had been referred because of diagnostic uncertainty and 24% had had an abdominal ultrasound; 66% of those <45 yr old underwent at least partial colonic evaluation.

These findings suggest that the Rome criteria combined with a lack of red flags have a very high predictive value for diagnosing irritable bowel syndrome. Application of these diagnostic criteria has the potential to alter utilization of health care resources.