Symptom classification in irritable bowel syndrome as a guide to treatment

• Published in: Scandinavian journal of gastroenterology Vol. 44; no. 7; p. 796
• Main Authors: Wiesner, Maureen, Naylor, Sally J, Copping, Ann, Furlong, Anita, Lynch, Andrew G, Parkes, Miles, Hunter, John O
• Format: Journal Article
• Language: English
• Published: England 2009
• Subjects: Adult; Cluster Analysis; Female; Humans; Irritable Bowel Syndrome - classification; Irritable Bowel Syndrome - physiopathology; Irritable Bowel Syndrome - psychology; Irritable Bowel Syndrome - therapy; Male; Middle Aged; ROC Curve; Statistics, Nonparametric; Surveys and Questionnaires; Treatment Outcome
• ISSN: 1502-7708
• DOI: 10.1080/00365520902964705

The treatment of irritable bowel syndrome (IBS) remains unsatisfactory. There are no objective markers for diagnosis, and classification (currently based on symptoms) provides little insight into potential causes or optimal therapy. The aim of this study was to determine whether a Swedish classification of IBS based on cluster analysis of patients' symptoms might provide a guide to successful treatment. Patients in a research clinic for IBS were classified according to criteria published by Ragnarsson & Bodemar (R&B) and also assessed independently by a clinician. Patients fulfilling the R&B criteria for subgroups 1 and 2 received specific treatments, either bulk laxatives or dietary treatment to reduce colonic fermentation, respectively. Patients who did not fit into these categories were given "best treatment" targeted at their predominant symptoms, but not limited in any way. Results before and after follow-up were assessed using a validated symptom-scoring scale. Seventy-one successive patients were recruited, and the numbers falling into R&B subgroups 1 and 2 were 15 (21%), and 28 (39%), respectively, leaving 28 (39%) unclassified. Receiver operating characteristic plots showed that the criteria for separation into subgroups 1 and 2 correlated well with the clinician's assessment. After treatment, symptom scores for the whole group showed a significant improvement (p<0.0001), but results were significantly better in subgroups 1 and 2 than in those unclassified, even when allowance was made for a potential therapeutic placebo effect of 40%. The R&B classification provides a helpful guide to treatment in many cases of IBS.