Subtyping the irritable bowel syndrome by predominant bowel habit:: Rome II versus Rome III

Background The agreement between subtyping irritable bowel syndrome (IBS) patients based on Rome II criteria versus Rome III criteria is unknown. Aim To compare IBS subtyping based on Rome II versus III criteria. Methods The Rome II Modular Questionnaire and the Bristol Stool Form Scale (one-week diary cards) were completed by 249 IBS patients. Based on the Rome II criteria, patients were defined as having diarrhoea- or constipation-predominant IBS, or alternating IBS. Based on the Rome III criteria, patients were divided into IBS with constipation, IBS with diarrhoea, mixed IBS or unsubtyped IBS. Agreement between Rome II and Rome III was assessed with kappa statistics. Results Based on Rome II there were 92 diarrhoea-predominant IBS, 45 constipation-predominant IBS and 112 alternating IBS, and based on Rome III 97 IBS with diarrhoea, 77 IBS with constipation, 16 mixed IBS and 59 unsubtyped IBS. The agreement between Rome II and Rome III subgroups was 46% (kappa = 0.19). Changes from the constipation to the diarrhoea subgroups and vice versa were uncommon (8% of patients). The majority of changes occurred from/to the alternating IBS, mixed IBS and unsubtyped IBS subgroups. Conclusion There is poor agreement between subtyping of IBS patients based on Rome II versus Rome III criteria.