Assessing the Impact of Changes to the Rome IV Criteria for Clinical Practice in Irritable Bowel Syndrome

Modifications to the Rome IV criteria for clinical practice have been proposed.1Drossman D.A. et al.Gastroenterology. 2022; 162: 675-679Abstract Full Text Full Text PDF PubMed Google Scholar For irritable bowel syndrome (IBS), these include relaxing the minimum symptom frequency threshold, provided the cardinal symptoms are present and impact on quality of life by being bothersome (eg, leading to consultation with a physician), and removing the 6-month minimum symptom duration. We have shown that the Rome IV criteria for IBS select a group of patients with severe symptoms and high levels of psychological comorbidity.2Black C.J. et al.Clin Gastroenterol Hepatol. 2020; 18: 392-398Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar We have also demonstrated that the Rome IV criteria are an improvement over the Rome III criteria in facilitating a diagnosis of IBS, but because of their more restrictive nature, almost 20% of people believed to have IBS according to a physician do not meet these criteria.3Black C.J. et al.Gut. 2020; 70: 1110-1116Crossref PubMed Scopus (25) Google Scholar We assessed how the proposed modifications impact on patient selection and diagnostic performance. We used data from 2 studies reported elsewhere.2Black C.J. et al.Clin Gastroenterol Hepatol. 2020; 18: 392-398Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar,3Black C.J. et al.Gut. 2020; 70: 1110-1116Crossref PubMed Scopus (25) Google Scholar The first was a cross-sectional survey of individuals self-reporting IBS assessing clinical and psychological characteristics of those meeting the Rome IV criteria.2Black C.J. et al.Clin Gastroenterol Hepatol. 2020; 18: 392-398Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar Briefly, we collected complete demographic, symptom, and psychological health data, using validated questionnaires, from 1375 adults registered with 3 UK organizations. Some individuals had consulted a primary care physician, some a gastroenterologist, and some had never consulted a physician, meaning the participants are likely to be generalizable to many individuals living with IBS in the United Kingdom. In this study we showed that applying the Rome IV criteria, instead of Rome III, reduced the proportion of individuals who self-reported IBS meeting diagnostic criteria for IBS from 79% to 59%. Individuals with Rome IV-defined IBS had more severe symptoms and worse psychological health compared with those who only met the Rome III criteria. We used this dataset to assess the effect of the modifications on symptom severity and psychological health in those meeting the new modified Rome IV criteria compared with all other individuals self-reporting a diagnosis of IBS. The second study was a diagnostic accuracy study of the Rome IV criteria for the diagnosis of IBS.3Black C.J. et al.Gut. 2020; 70: 1110-1116Crossref PubMed Scopus (25) Google Scholar In this study we collected data, prospectively, from 577 consecutive patients attending a specialist IBS clinic in the United Kingdom, all of whom underwent a relatively standardized workup, with assessors blinded to symptom status. The reference standard used to confirm IBS was the presence of lower abdominal pain associated with altered stool form or frequency in patients with no evidence of organic gastrointestinal disease after investigation. Sensitivity and specificity of the Rome IV criteria were 82.4% and 82.9%, respectively, with positive and negative likelihood ratios of 4.82 (95% confidence interval [CI], 3.30–7.28) and 0.21 (95% CI, 0.17–0.26), respectively. We used this dataset to assess performance of the modified Rome IV criteria. In both datasets we redefined Rome IV IBS approximating the proposed modifications (Supplementary Table 1).1Drossman D.A. et al.Gastroenterology. 2022; 162: 675-679Abstract Full Text Full Text PDF PubMed Google Scholar Patients reporting 2 or more of abdominal pain related to defecation, softer or harder stools at onset of abdominal pain, and more or less frequent stools at onset of abdominal pain, of 3 months duration or more, were classified as meeting modified Rome IV criteria, provided symptoms were bothersome enough to have led to consultation with a physician. In the cross-sectional survey, 1272 participants (92.5%) met modified Rome IV criteria for IBS. Characteristics of these individuals, compared with those with self-reported IBS who did not meet modified Rome IV criteria, are provided in Table 1 and are compared with the original Rome IV data. When data were compared between all 1272 individuals who now met the modified Rome IV criteria and the 102 subjects who did not, higher rates of anxiety (49.0% vs 37.3%, P = .001), somatization (47.5% vs 28.4%, P < .001), and gastrointestinal symptom–specific anxiety (33.8% vs 15.7%, P < .001) persisted, but rates of depression and severe symptoms were no longer higher.Table 1Characteristics of Individuals Meeting Modified Rome IV Criteria Compared With Those not Meeting Modified Rome IV Criteria for IBSOriginal study dataReanalysis of dataMet Rome IV criteria for IBS (n = 811)Did not meet Rome IV criteria for IBS (n = 562)PaP value for Pearson χ2 for comparison of categorical data.Met modified Rome IV criteria for IBS (n = 1272)Did not meet modified Rome IV criteria for IBS (n = 102)PaP value for Pearson χ2 for comparison of categorical data.IBS subtype Constipation142 (17.5)126 (22.4)247 (19.4)23 (22.5) Diarrhea311 (38.4)206 (36.7)480 (37.8)37 (36.3) Mixed stool pattern331 (40.9)185 (32.9)484 (38.1)31 (30.4) Unclassified26 (3.2)41 (7.3)<.00157 (4.5)10 (9.8).054IBS-SSS symptom severity Mild90 (11.2)202 (36.0)262 (21.4)31 (31.4) Moderate333 (41.5)212 (37.8)517 (42.1)27 (30.0) Severe379 (47.3)100 (17.8)<.001448 (36.5)32 (35.6).009IBS limits activities ≥50% of the time573 (70.8)253 (45.2)<.001791 (62.4)37 (36.3)<.001HADS-A categories Normal202 (24.9)226 (40.2)380 (29.9)48 (47.1) Borderline167 (20.6)117 (20.8)269 (21.1)16 (15.7) Abnormal442 (54.5)219 (39.0)<.001623 (49.0)38 (37.3).001HADS-D categories Normal434 (53.5)373 (66.4)749 (58.9)59 (57.8) Borderline191 (23.6)103 (18.3)273 (21.5)21 (20.6) Abnormal186 (22.9)86 (15.3)<.001250 (19.7)22 (21.6).89PHQ-15 categories Mild somatization6 (0.7)23 (4.1)20 (1.6)8 (8.8) Low somatization78 (9.6)138 (24.6)187 (14.7)29 (28.4) Medium somatization270 (33.3)225 (40.0)461 (36.2)35 (34.3) High somatization457 (56.4)176 (31.3)<.001604 (47.5)29 (28.4)<.001VSI tertiles Low196 (24.3)268 (47.8)408 (32.2)57 (55.9) Medium281 (34.8)179 (31.9)432 (34.0)29 (28.4) High331 (41.0)114 (20.3)<.001429 (33.8)16 (15.7)<.001Values are n (%).IBS-SSS, Irritable Bowel Syndrome Severity Scoring System; HADS-A, Hospital Anxiety and Depression Anxiety Score; HADS-D, Hospital Anxiety and Depression Depression Score; PHQ-15, Patient Health Questionnaire-15; VSI, Visceral Sensitivity Index.a P value for Pearson χ2 for comparison of categorical data. Open table in a new tab Values are n (%). IBS-SSS, Irritable Bowel Syndrome Severity Scoring System; HADS-A, Hospital Anxiety and Depression Anxiety Score; HADS-D, Hospital Anxiety and Depression Depression Score; PHQ-15, Patient Health Questionnaire-15; VSI, Visceral Sensitivity Index. In the diagnostic accuracy study, among 573 patients providing complete data, 518 (90.6%) met the modified Rome IV criteria. Of 456 patients with a diagnosis of IBS according to the reference standard, 452 met the modified Rome IV criteria, yielding a sensitivity of 99.1%. Among 117 without IBS, 50 did not meet the modified Rome IV criteria, giving a specificity of 42.7%. Positive and negative likelihood ratios for the modified Rome IV criteria were 1.73 (95% CI, 1.48–2.02) and 0.02 (95% CI, 0.01–0.06), respectively. Among 67 patients meeting the modified Rome IV criteria who were not diagnosed with IBS according to the reference standard, 23 had organic disease. Seventeen patients had bile acid diarrhea, 3 exocrine pancreatic insufficiency, 1 small-bowel Crohn’s disease, 1 ulcerative proctitis, and 1 microscopic colitis. The remainder were believed to have another disorder of gut–brain interaction. This reanalysis of data demonstrates that modifications to the Rome IV criteria for clinical practice led to lower absolute differences in symptom severity, compared with people with self-reported IBS who do not meet these criteria. Rates of severe symptoms and anxiety were also lower in patients defined according to these modified criteria compared with when the original Rome IV criteria were used in this study population.2Black C.J. et al.Clin Gastroenterol Hepatol. 2020; 18: 392-398Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar However, our observations could relate to a lack of power because of the smaller number of patients not meeting modified Rome IV criteria. When used to diagnose IBS, although the modified criteria were more sensitive, they were less specific, leading to a much lower positive likelihood ratio than observed with the original Rome IV criteria.3Black C.J. et al.Gut. 2020; 70: 1110-1116Crossref PubMed Scopus (25) Google Scholar The Rome III criteria performed only modestly in diagnosing IBS in previous validation studies,4Ford A.C. et al.Gastroenterology. 2013; 145: 1262-1270Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar,5Sood R. et al.Am J Gastroenterol. 2016; 111: 1446-1554Crossref PubMed Scopus (32) Google Scholar so this improvement in performance of the original Rome IV criteria was welcome. Nevertheless, these modifications may be useful. The high sensitivity seen with the modified criteria means if these are not met at the outset, then a final diagnosis of IBS is highly unlikely, and further investigation should be pursued. Finally, prevalence of organic disease among the 525 patients who met these modified criteria was low at <5%. Our approximation of these modified criteria means the results are preliminary and need to be confirmed by other studies, designed with this specific purpose in mind. It is also important to point out that these modified criteria are intended for use in clinical practice, not in clinical trials or epidemiologic or pathophysiologic studies. Alexander C. Ford, MBChB (Conceptualization: Lead; Formal analysis: Lead; Methodology: Lead; Writing – original draft: Lead). Christopher J. Black, PhD (Conceptualization: Supporting; Formal analysis: Supporting; Methodology: Supporting; Writing – original draft: Supporting). Supplementary Table 1Modifications to the Rome IV Criteria for IBS and our Approximations of ThemModified Rome IV criteria as proposed by Drossman et al1Drossman D.A. et al.Gastroenterology. 2022; 162: 675-679Abstract Full Text Full Text PDF PubMed Scopus (5) Google ScholarModified Rome IV criteria as approximated in this studyRecurrent abdominal pain, on average, at least 1 day per week in the last 3 months and associated with 2 or more or the following:Recurrent abdominal pain, on average, at least 1 day per week in the last 3 months and associated with 2 or more or the following: Related to defecationRelated to defecation Associated with a change in frequency of stoolAssociated with a change in frequency of stool Associated with a change in stool formAssociated with a change in stool form Criteria fulfilled for the last 8 weeksCriteria fulfilled for the last 3 months Patients should have sufficiently bothersome symptoms to seek care or affect daily activityPatients should have sufficiently bothersome symptoms to seek careNote: Key differences between the modified Rome IV criteria and our approximation are shown in italic text. Open table in a new tab Note: Key differences between the modified Rome IV criteria and our approximation are shown in italic text.