Validation of the Red Flags Index for Early Diagnosis of Crohn’s Disease: A Prospective Observational IG-IBD Study Among General Practitioners

Abstract

1. Introduction

Diagnostic delay >12 months is frequent in Crohn’s disease [CD].^1–3 This is associated with lower response rates to therapies and worse outcomes [complicated disease course, increased operation rates].⁴ The overlap of early symptoms and signs between CD and other conditions, such as irritable bowel syndrome [IBS], can be associated with delay in diagnosis.³

Recently, the International Organization for Inflammatory Bowel Disease [IO-IBD] developed a tool to identify early CD and reduce diagnostic delay, using over 201 subjects [85 with CD, 80 with IBS, 36 healthy subjects].³ This index is calculated on a simple questionnaire administered by the physician, and includes eight symptoms and signs suggestive for CD [Supplementary Table 1, available as Supplementary data at ECCO-JCC online]. This index has been estimated to have high sensitivity and specificity [96% and 89% respectively]. Subjects with an index ≥8 are more likely to have suspected CD (odds ratio [OR] 205, 95% confidence interval [CI] 54.3–774, p <0.0001).

We aimed to validate this questionnaire at the community level in a population of patients seen by general practitioners [GP] in two large areas of Lombardy, Italy [Bergamo and Pavia].

2. Report

Consecutive adult patients, with intestinal symptoms and having no medical history of any gastrointestinal disease, referring to the GP, were screened. Patients should have at least one of the following symptoms: chronic abdominal pain, chronic diarrhoea, nocturnal diarrhoea, unexpected weight loss, or perianal lesions.³ The GPs administered the Red Flags [RF] questionnaire to the eligible patients. All patients were referred to the nearest participating centre to confirm or exclude the diagnosis of CD. IBD specialists were blinded to the results of the questionnaire. First-line investigations systematically included blood cell count, serum C-reactive protein, faecal calprotectin [FC], and abdominal ultrasound, according to routine practice. If necessary, second-line investigations were planned [ie colonoscopy and cross-sectional imaging]. Sensitivity, specificity, and positive and negative predictive values [PPV, NPV] of the RF index were calculated. Patients lost to follow-up after the first gastroenterological visit were considered as not having CD.

From November 2016 to November 2019, 64 GPs participated in the study [Bergamo: n = 52, Pavia: n = 12, in charge of a total population of 93 000 subjects] and 112 patients, over a mean number of 53 568 subjects with at least one access to the GP’s visit room for any reason during the study period, were included [median age 35 years, range: 18–69, 37% males]. A total of 66 subjects [59%] completed the study after the first gastroenterological visit [Supplementary Figure 1, available as Supplementary data at ECCO-JCC online].

The prevalence of CD was 3.6% in the study population [4 out of 112 subjects]. Two patients had ileal disease [50%] and two ileo-colonic location [50%], three had B1, and one had B2 phenotype [colonic passable stricture]. Median age was 25.5 years [range 24–27], and three of them [75%] were females. The RF index had 50% sensitivity, 58% specificity, 4% PPV, and 97% NPV.

A combined diagnostic strategy with faecal calprotectin [considered ‘positive’ if having RFI ≥8 and/or FC >250 ng/g] resulted in significantly improved diagnostic accuracy: sensitivity 100% [29–100%], specificity 72% [55–85%], PPV = 21% [5–51%], NPV = 100% [88–100%], Figure 1]; however, only 42 subjects [with three confirmed cases of CD] were available for this analysis.

Figure 1.

Combination of the Red Flags Index [RFI] and faecal calprotectin [FC] improves the chance to diagnose CD.

Open in new tabDownload slide

Over these 42 subjects with available FC, the RF index had 67% sensitivity [9–99%], 74% specificity [5786%], 16% PPV [2–48%], NPV 96% [82–99%]; FC alone had 100% sensitivity [29–100%], 93% specificity [82–99%], 60% PPV [14–94%], 100% NPV [90–100%]. The combination had 67% sensitivity [9–99%], 97% specificity [86–99%], 67% PPV [9–99%], 97% NPV [86–99%].

3. Discussion

The diagnostic accuracy of the RF questionnaire was quite low in terms of PPV when applied alone to a primary care setting, in a region where CD incidence is about 5/100 000 inhabitants. The RF index showed high NPV, and the combination with FC raised the PPV from 4% to 21%.

Since the RF mainly evaluated CD signs and symptoms, the low sensitivity may be explained by the estimated proportion of patients [about 30%] who might have silent CD⁵. Incorporating FC in the diagnostic strategy dramatically increased the accuracy of the RF index for CD diagnosis in the general population.

Of the four patients diagnosed with CD, 75% had early CD according to the Paris Consensus definition,⁶ demonstrating that the RF index + FC can be of value to identify patients requiring early referral to a gastroenterologist and allow early disease control.⁷

In our cohort, 41% of patients did not complete the planned diagnostic work-up although all examinations were accessible by the Italian National Health System reimbursement, and the time to perform all investigations was very short [maximum 3 weeks]. As previously reported,⁸ our findings suggest that the reduction of diagnostic delay requires awareness among GPs, and also education and awareness about inflammatory bowel disease among the general population.

Our study demonstrated that the Red Flag alone does not seem to be a sensitive tool for screening, since the ability to capture CD seems too low, but this can improve with the combination with FC. The study does not validate its use alone, but a diagnostic algorithm which allows the RFI to select patients who need FC testing [based on its good NPV] may be the optimal way to proceed, and may be the basis for future validation studies.

Funding

This study was funded by Takeda Pharmaceuticals, and by the Italian Group for the study of Inflammatory Bowel Disease [IG-IBD].

Conflict of Interest

SD has served as a speaker, consultant, and advisory board member for Schering-Plough, Abbott [AbbVie] Laboratories, Merck and Co., UCB Pharma, Ferring, Cellerix, Millenium Takeda, Nycomed, Pharmacosmos, Actelion, Alfa Wasserman, Genentech, Grunenthal, Pfizer, Astra Zeneca, Novo Nordisk, Cosmo Pharmaceuticals, Vifor, and Johnson and Johnson. GF received consultancy fees from Ferring, MSD, AbbVie, Takeda, Janssen, Amgen, Sandoz, Samsung Bioepis, Celltrion. LPB has received consulting fees from Merck, Abbvie, Janssen, Genentech, Mitsubishi, Ferring, Norgine, Tillots, Vifor, Therakos, Pharmacosmos, Pilège, BMS, UCB-Pharma, Hospira, Celltrion, Takeda, Biogaran, Boerhinger Ingelheim, Lilly, Pfizer, HAC-Pharma, Index Pharmaceuticals, Amgen, and Sandoz; and lecture fees from Merck, Abbvie, Takeda, Janssen, Takeda, Ferring, Norgine, Tillots, Vifor, Therakos, Mitsubishi, and HAC-Pharma.

Author Contributions

GF and SD designed the study; SB wrote the statistical plan; GF wrote the study protocol; GF and ADS enrolled patients; GF and DG coordinated the study; EA performed study diagnostic procedures; GF and SB drafted the manuscript; LPB and SD critically revised the manuscript; all the authors approved the final version of the manuscript.

Conference presentation: hard-copy presentation P244, ECCO Congress, Vienna, February 2020

Acknowledgements

We thank all the GPs who participated in the enrolment of patients.

References