NIHR Signal Stool test is useful before GPs refer for possible inflammatory bowel disease
Published on 4 December 2018
A stool test by GPs has been shown to support referral decisions for young adults, not suspected of cancer, to investigate possible inflammatory bowel disease (IBD - which includes Crohn’s Disease and Ulcerative Colitis). This study supports current NICE guidelines that the calprotectin stool test can usefully inform patient referral pathways and reduce unnecessary invasive tests such as colonoscopy.
High levels of faecal calprotectin are associated with gut inflammation, as occurs in IBD and other organic intestinal diseases; low levels are associated with functional bowel disorders, requiring different management.
In this cohort study, 789 adults aged 18-46 with gastrointestinal symptoms suggestive of either IBD or functional bowel disorders, and who had calprotectin tests to inform GP referral decisions, were followed for a year to assess diagnostic investigations and outcomes. Of patients with a negative calprotectin test, 99% did not have IBD.
Used as a decision aid, the test result influenced referral decisions for 45% of patients tested: it helped to avoid referral for 40% and led to referral for an additional 5%.
- Diagnostics, Gastrointestinal disorders, Haematological and immunological disorders, Health management, Primary care, Acute and general medicine
Why was this study needed?
Around 10% of presenting problems to GPs involve gastrointestinal symptoms. Most young adults who present with symptoms such as changed bowel habit and pain have a functional bowel disorder, such as irritable bowel syndrome, which does not require invasive investigation.
However, symptoms of IBD and functional bowel disorder can be similar, making it challenging to minimise the burden of unnecessary further investigation without missing the minority who do have IBD and may come to considerable harm without timely investigation.
Currently, where GPs suspect cancer or IBD, patients are referred for urgent investigation. However, there is often considerable diagnostic uncertainty between IBD and functional bowel disorders, and a test to reduce this uncertainty may avoid patients being referred for unnecessary invasive testing. Other research has previously demonstrated the usefulness of calprotectin testing to rule out IBD in patients who do not have concerning or severe symptoms; this study included some patients with concerning symptoms such as bleeding, but otherwise considered at low risk of having cancer.
What did this study do?
In this prospective cohort study, a convenience sample of 789 patients had calprotectin tests to support GP referral decisions (the test was not mandatory for all patients who met eligibility criteria). Inclusion criteria were: aged 18 to 46 with gastrointestinal symptoms but no previous diagnosis of IBD, no clinical suspicion of colorectal cancer, and no use of NSAID painkillers in the previous six weeks. Three hundred and eight (39%) of these patients had concerning symptoms such as bleeding.
At the point of testing, GPs recorded whether they would have referred the patient to secondary care for investigation had the test not been available, and researchers assessed if the test results changed GP referral decisions. Patients were followed for 12 months to allow comparison between their calprotectin test result and their final diagnosis.
This was a well-designed study conducted in the NHS and is likely to be relevant to practice. However, because of the way that patients were selected, the low uptake of testing, the exclusion of patients otherwise suspected of having cancer, and exclusion of all patients who had taken NSAID painkillers within the last six weeks, a cautious interpretation is needed.
What did it find?
- Overall, 17% (132/789) of patients tested had a positive calprotectin test (>100 µg/g), and 6% (50/789) were eventually diagnosed with IBD.
- For the 481 patients without concerning symptoms: of those with negative test results, 99% did not have IBD (negative predictive value); of those with positive test results, 39% did have IBD (positive predictive value).
- For the 308 patients with concerning symptoms such as bleeding: of those with negative test results 98% did not have IBD (negative predictive value); of those with positive test results, 50% did have IBD (positive predictive value).
- Comparing GPs referral intentions for all 789 patients: a negative test result influenced the decision to avoid 317 referrals (40% of all patients), while a positive test result influenced the decision to refer an additional 38 (5%).
What does current guidance say on this issue?
The NICE 2013 guidance for adults recommends faecal calprotectin testing as an option to help GPs distinguish inflammatory bowel disease from functional bowel disorders. It only recommends calprotectin testing in adults with recent onset symptoms, and when cancer is not suspected. NICE also recommends that calprotectin testing requires quality assurance processes and locally agreed care pathways to be in place.
What are the implications?
The study supports NICE guidance that calprotectin testing could be a useful tool to support GPs to rule out IBD in many patients, even if they do have bleeding. Designers of referral pathways may consider how to integrate calprotectin testing in order to reduce referrals for invasive diagnostic investigation.
The study authors recommend that, although not definitive in themselves, calprotectin tests may be useful as part of locally-agreed referral pathways that include: good oversight, close coordination with secondary care, ongoing training, and audit.
The authors also recommend an 8-week safety net review of non-referred patients with negative test results to ensure that patients with false negative result are not lost to follow up.
Citation and Funding
Walker GJ, Moore L, Heerasing N et al. Faecal calprotectin effectively excludes inflammatory bowel disease in 789 symptomatic young adults with/without alarm symptoms: a prospective UK primary care cohort study. Aliment Pharmacol Ther. 2018;47(8):1103-16.
The study was supported by Crohn’s and Colitis UK, Tillotts and South West Academic Health Sciences Network, with additional NIHR support for the research team.
NICE. Faecal calprotectin diagnostic tests for inflammatory diseases of the bowel. DG11. London: National Institute for Health and Care Excellence; 2013.
Delays in diagnosing inflammatory bowel disease can occur as the symptoms overlap with functional bowel disorders and other organic intestinal diseases.
NICE DG11 recently published work by NHS England, the NHSBA Pacific Programme, and the Yorkshire and Humberside AHSN to show that a faecal calprotectin of 100 micrograms/g used as part of a primary care pathway can reduce unnecessary referrals to secondary care and reduce colonoscopy rates.
This study corroborates these findings and shows that this pathway can also be used in patients with alarm symptoms who would otherwise have been referred straight to secondary care.
Kevin Barrett, General Practitioner, New Road Surgery, Hertfordshire
The commentator is chair of the Primary Care Society for Gastroenterology. He is also Clinical Champion for the IBD Spotlight Project for RCGP and Crohn's and Colitis UK.
Diagnostic tests are never perfect and will sometimes suggest that some people who are unwell do not have the disease, and that some people who are healthy do have the disease. This is why test results must always be considered as part of a clinical assessment and alongside results of other investigations. Positive and negative predictive values are useful to help understand how reliable an isolated test result is.
Positive predictive value [PPV]: Suppose a person has a positive test result – how sure can they be that they really do have the disease? If, for example, a test has a positive predictive value of 40%, then for every 100 people whose test results are positive (suggesting they have the disease): 40 really do have the disease, while 60 are in fact healthy.
Negative predictive value [NPV]: Suppose another person has a negative test result – how sure can they be that they do not have the disease? If, for example, a test has a negative predictive value of 90%, then for every 100 people whose test results are negative (suggesting they are healthy): 90 really are healthy, while 10 do in fact have the disease.