Novel Symptom Subgroups in Patients With Irritable Bowel Syndrome Are Associated With Healthcare Utilisation in Secondary and Tertiary Care

BACKGROUND & AIMS: Current classification systems for irritable bowel syndrome (IBS) based on bowel habit do not consider psychological impact. We applied a previously validated latent class analysis (LCA) model to a cohort of patients with IBS in secondary and tertiary care to assess whether it predicted levels of healthcare utilisation. METHODS: We applied our LCA model to a referral population with IBS. As described previously, we assigned cluster membership based on gastrointestinal symptom severity and psychological burden. We assessed demographics, symptom severity and quality of life at baseline and levels of healthcare utilisation during 12 months of longitudinal follow-up according to cluster. RESULTS: We recruited 379 patients, of whom 249 (65.7%) met the Rome IV criteria. Those in the four clusters with higher psychological burden had more severe symptoms on the IBS-SSS and lower quality of life scores (p < 0.001 for both). Rates of discharge were generally lower in clusters with higher psychological burden (p = 0.05). Rates of prescribing a drug for IBS (p = 0.001), the mean number of drugs prescribed for IBS (p < 0.001) and the mean number of different drug types prescribed for IBS (p < 0.001 for trend) were highest in the four clusters with higher psychological burden. CONCLUSIONS: In patients with IBS in secondary and tertiary care, the LCA model identified groups of individuals with more severe symptoms and greater impairments in quality of life at baseline and significantly higher rates of healthcare utilisation during longitudinal follow-up.