Diagnostic accuracy of a simplified tool for irritable bowel syndrome with predominant constipation in Chinese patients: A multicenter, prospective, observational study

To the Editor: Irritable bowel syndrome (IBS) is one of the most common adult functional gastrointestinal disorders (FGIDs), with an estimated global prevalence of 1.5–11.2%. The prevalence rates for anxiety and depressive symptoms and disorders underscore the psychological impact on IBS patients, emphasizing the need for comprehensive diagnostic tools tailored to specific populations.[1] The main diagnostic reference for IBS is the Rome criteria which is now the standard practice for IBS diagnosis worldwide. The most recent iteration is Rome IV [Supplementary Table 1, https://links.lww.com/CM9/C102]. Despite its broad acceptance, validation studies demonstrated that the Rome criteria only performed modestly in diagnosing IBS and, in clinical practice, Rome criteria are not commonly used. The overlap between the symptoms of IBS and FGIDs must be considered in the diagnosis of IBS and should be taken into consideration in the development of symptom-based diagnostic tools. Presently, there is no specific gold standard for IBS diagnostic method in Asia and only Japanese Society of Gastroenterology (JSGE) published the second edition of evidence-based clinical practice guidelines for IBS in 2020.[2] Hence, we developed the simplified diagnosis tool for IBS with predominant constipation (IBS-C) diagnosis in Chinese patients. The tool was derived based on the relevant parts of the Rome IV and Asian FGID scales. Similar to Rome IV, the simplified diagnosis tool for IBS-C is a patient self-reported questionnaire [Supplementary Table 2, https://links.lww.com/CM9/C102]. It comprises eight questions: one question on the duration of the disease, two questions on the main symptoms, three questions on the change of symptoms, and two questions on stool appearance [Supplementary Table 2, https://links.lww.com/CM9/C102]. The presence of mid–upper abdominal symptoms is assessed in questions 2 and 6. The objective of this study was to assess the diagnostic accuracy of the simplified diagnosis tool for IBS-C in the Chinese population, using Rome IV as the standard reference. This was a multicenter, prospective, observational study (ClinicalTrials.gov Identifier: NCT04968652) that included a face-to-face visit, during which patients underwent the IBS-C diagnostic assessment by an experienced gastrointestinal (GI) clinician. Patients were identified by investigators during routine clinical visits at 10 participating study centers between November 1, 2021, and October 28, 2022. The key eligibility criteria are mentioned in supplementary files, https://links.lww.com/CM9/C102. Diagnostic assessments for patients were first performed using Rome IV criteria [Supplementary Table 1, https://links.lww.com/CM9/C102], followed by the simplified diagnosis tool during the same visit [Supplementary Table 2, https://links.lww.com/CM9/C102]. Patients were assigned to the IBS-C group if their Rome IV results were marked as “IBS-C,” or to the non-IBS-C group if their Rome IV results were marked as “non-IBS-C.” The study protocol and the informed consent form were approved by the Ethics Committee of the First Affiliated Hospital of Sun Yat-sen University. Written informed consent was provided upon enrollment and patients were permitted to discontinue the study at any time. The primary outcome was the sensitivity and specificity of the simplified diagnosis tool related to IBS-C diagnosis in Chinese patients diagnosed with IBS-C using Rome IV criteria as the standard reference. Secondary outcomes were the accuracy of the simplified diagnosis tool for IBS-C, negative and positive predictive values, and kappa coefficient. The definitions for each endpoint are shown in Supplementary Tables 3 and 4, https://links.lww.com/CM9/C102. A total of 301 patients were screened and 300 patients were enrolled and completed the study. FAS included 133 patients assigned to the IBS-C group and 167 patients to the non-IBS-C group based on Rome IV criteria [Supplementary Figure 1, https://links.lww.com/CM9/C102]. Based on the simplified diagnosis tool for IBS-C, 205 (68.3%) patients were diagnosed with IBS-C, and 95 (31.7%) patients without IBS-C [Supplementary Table 5, https://links.lww.com/CM9/C102]. A total of 285 patients were included in the sensitivity analysis. The mean (±SD) age of the total population was 43.9 (±16.4) years; 77.7% of the patients were female and all patients were Asian [Supplementary Table 6, https://links.lww.com/CM9/C102]. Of 285 patients, 133 and 152 patients were assigned to the IBS-C and non-IBS-C groups, respectively, using Rome IV criteria, while 204 patients were diagnosed with IBS-C and 81 without IBS-C, respectively, using the simplified diagnosis tool for IBS-C diagnosis [Supplementary Table 7, https://links.lww.com/CM9/C102]. Using Rome IV as the standard reference, the sensitivity and specificity of the simplified diagnosis tool for IBS-C were 86.5% (95% confidence interval [CI], 79.5–91.8%) and 46.1% (95% CI, 38.4–54.0%), respectively [Table 1]. In the sensitivity analysis, the sensitivity and specificity were 86.5% (95% CI, 79.5–91.8%) and 41.4% (95% CI, 33.5–49.7%), respectively [Supplementary Table 8, https://links.lww.com/CM9/C102]. The accuracy of the simplified diagnosis tool for IBS-C was 64.0% (95% CI, 58.3–69.4%), the positive predictive value was 56.1% (95% CI, 49.0–63.0%), the negative predictive value was 81.1% (95% CI, 71.7–88.4%), and the kappa coefficient was 0.31 (95% CI, 0.22–0.40) [Table 1]. In the sensitivity analysis, the accuracy was 62.5% (95% CI, 56.6–68.1%), the positive predictive value was 56.4% (95% CI, 49.3–63.3%), the negative predictive value was 77.8% (95% CI, 67.2–86.3%), and the kappa coefficient was 0.27 (95% CI, 0.17–0.37) [Supplementary Table 8, https://links.lww.com/CM9/C102]. Table 1 - Diagnostic accuracy of the simplified diagnosis tool for IBS-C (FAS) (N = 300). Variables Values Rome IV, n (%) IBS-C 133 (44.3) Non-IBS-C 167 (55.7) Simplified diagnosis tool, n (%) IBS-C 205 (68.3) Non-IBS-C 95 (31.7) Sensitivity (%) (95% CI)* 86.5 (79.5–91.8) Specificity (%) (95% CI)* 46.1 (38.4–54.0) Accuracy (%) (95% CI)* 64.0 (58.3–69.4) Positive predictive value (%) (95% CI)* 56.1 (49.0–63.0) Negative predictive value (%) (95% CI)* 81.1 (71.7–88.4) Kappa coefficient (95% CI)† 0.31 (0.22–0.40) CI: Confidence interval; FAS: Full analysis set; IBS-C: Irritable bowel syndrome with predominant constipation. *95% CI was determined using the Clopper–Pearson exact method. †95% CI was determined using the Cohen (1960) method. Overall, the simplified diagnosis tool displayed high sensitivity for diagnosing IBS-C. A high proportion of patients diagnosed with IBS-C using Rome IV were also diagnosed with IBS-C using the simplified diagnosis tool for IBS-C (i.e., a low false negative rate). The use of this tool in routine clinical practice may minimize unnecessary investigations during the diagnostic workup for patients with IBS-C, thus reducing the economic and psychological burden on these patients.[3,4] The modest specificity of this tool for IBS-C highlights several important points. First, it may incorrectly diagnose with IBS-C in patients who do not have the disease (i.e., false positives). The potential false positives must be considered when interpreting the results of the simplified diagnosis tool for IBS-C in clinical practice. Further research is also required to confirm the sensitivity and specificity of this tool for the diagnosis of IBS-C in Asian populations. Second, there is a lack of a gold standard diagnostic tool or standardized biomarker to diagnose IBS with acceptable precision in studies that assessed the accuracy of diagnostic tests for IBS. A major limitation of using Rome IV in Asian populations is that it was validated in Western populations, who have a lower prevalence of dyspeptic upper abdominal symptoms than patients from Asian countries. In the present study, 34.6% of patients who were diagnosed with IBS-C using Rome IV had upper abdominal symptoms when assessed using the simplified diagnosis tool for IBS-C, and half of these patients reported that the upper abdominal symptoms were related to meal intake. By incorporating questions about upper abdominal symptoms, the simplified diagnosis tool for IBS-C may better distinguish between patients with IBS-C and functional disease and therefore be a more accurate diagnostic tool for use in Asian patients than Rome IV. Abdominal discomfort or bloating is often the most important feature of IBS among Asian patients, which may lead to underdiagnosis of IBS using Rome IV. These differences in symptoms between patients in Western and Asian countries may contribute to the modest sensitivity of the Rome criteria in diagnosing IBS in Asian patients, as well as the modest specificity of the simplified diagnosis tool for IBS-C using Rome IV as a reference. Consistent with these observations, the kappa coefficient in this study indicated poor agreement between Rome IV and the simplified diagnosis tool for IBS-C. Overall, 13.5% of patients diagnosed with IBS-C using Rome IV criteria were judged as not having IBS-C by the simplified diagnosis tool, while 53.9% of patients diagnosed with non-IBS-C using Rome IV were judged as having IBS-C using the simplified diagnosis tool. In addition to the factors described earlier, this disagreement could be due to the stricter diagnostic criteria used by Rome IV, which exclude patients with mild-to-moderate symptoms. As patients were required to recall their symptoms from 6 months before completing the questionnaire, this could have resulted in a recall bias. Although any recall bias is unavoidable in questionnaire-based study, shorter timeframe of the simplified diagnosis tool has minimized the recall bias, but it is susceptible to the occurrence of infrequent events. To minimize selection bias, eligible patients included in the study were diagnosed during routine clinic visits by experienced GI clinicians who were trained with both Rome IV and the simplified diagnosis tool for IBS-C. There was no loss to follow-up as all assessments were completed within the initial visit. In conclusion, the simplified diagnosis tool demonstrated good sensitivity and may help identify more Chinese patients with IBS-C; thus, it might be an alternative to Rome IV in China and even Asia, although further validation is required by conducting more studies consisting of various sample sizes and other Asian or international cohorts to further validate the results of this study. This would help in greater regional or global applicability of the simplified diagnosis tool questionnaire. Moreover, the simplified diagnosis tool for IBS-C is easy to administer and accounts for the differences in symptoms between Asian and Western populations. The simplified diagnosis tool for IBS-C may provide a valuable input for developing Rome V. Acknowledgments Medical writing assistance was provided by Zhi Yang Loh, BSc; Liting Hang, PhD; and Alice Carruthers, PhD, of Nucleus Global Asia-Pacific. We would like to thank all researchers and patients who participated in the study, and AstraZeneca for sponsoring this study. Conflicts of interest The study was funded by AstraZeneca, China. Availability of data and materials The data that support the findings of this study are available from the corresponding author upon reasonable request.