Abstract
To evaluate the agreement between the Rome III and Rome IV criteria in diagnosing pediatric functional gastrointestinal disorders (FGIDs), we conducted a prospective cohort study in a public school in Cali, Colombia. Children and adolescents between 11 and 18 years of age were given the Spanish version of the Questionnaire on Pediatric Functional Gastrointestinal Disorders Rome III version on day 0 and Rome IV version on day 2 (48 h later). The study protocol was completed by 135 children. Thirty-nine (28.9%) children were excluded because of not following the instructions of the questionnaire. The final analysis included data of 96 children (mean 15.2 years old, SD ± 1.7, 54% girls). Less children fulfilled the criteria for an FGID according to Rome IV compared to Rome III (40.6% vs 29.2%, p=0.063) resulting in a minimal agreement between the two criteria in diagnosing an FGID (kappa 0.34, agreement of 70%). The prevalence of functional constipation according to Rome IV was significantly lower compared to Rome III (13.5% vs 31.3%, p<0.001), whereas functional dyspepsia had a higher prevalence according to Rome IV than Rome III (11.5% vs 0%).Conclusion: We found an overall minimal agreement in diagnosing FGIDs according to Rome III and Rome IV criteria. This may be partly explained by the differences in diagnostic criteria. However, limitations with the use of questionnaires to measure prevalence have to be taken into account. What is Known: • The Rome IV criteria replaced the previous Rome III criteria providing updated criteria to diagnose functional gastrointestinal disorders (FGIDs). • Differences found between Rome IV and historic Rome III FGID prevalence may have been affected by changes in prevalence over time or differences in sample characteristics. What is New: • We found a minimal agreement between Rome III and Rome IV FGID diagnosis, especially in the diagnoses of functional constipation, irritable bowel syndrome, and functional dyspepsia. • The minimal agreement may be partly explained by changes in diagnostic criteria, but limitations with the use of questionnaires to measure prevalence have to be taken into account.