Faecal calprotectin diagnostic tests for inflammatory diseases of the bowel

Table 1 Instances of symptoms in 269 people on presentation to GP

Initial normal result (<100 micrograms/g)

Three quarters of people (75.8%, 204/269) had an initial normal faecal calprotectin test result. Of these:

Regardless of the final diagnostic pathway being in primary or secondary care:

Initial borderline result (100–250 micrograms/g)

There was an initial borderline calprotectin result for 13.0% of people (35/269). All of these should have had a second calprotectin test 2 weeks later (in accordance with the local guideline). Thirty‑one people were managed according to the protocol with the following results and outcomes:

Initial high result (more than 250 micrograms/g)

Of those tested, 11.2% of people (30/269) had an initial high faecal calprotectin result and all but 1 were referred urgently for specialist investigations according to the local guidelines. Of these 29 people:

The 1 person who wasn't referred, despite a high result, had campylobacter gastroenteritis that resolved spontaneously, so was not sent for colonoscopy.

Table 2 shows the final diagnosis for all people who had a faecal calprotectin test that fulfilled the local guideline criteria. The faecal calprotectin results used are the final test results for each person.

Table 2 Final diagnosis for all people receiving a faecal calprotectin test

Table 2 shows that no one with a normal faecal calprotectin result had IBD. An intermediate result was also a low indicator of IBD (1/17, 5.9%). However, one quarter (25.8%, 8/31) of people with a high faecal calprotectin test result had a final diagnosis of IBD.

The initial management pathway leading to the final diagnosis for all people who had a faecal calprotectin test is shown in table 3.

This table shows that the condition in:

Table 3 Management pathway for all patients receiving a faecal calprotectin test

The audit was not able to determine how the patients who were referred for a faecal calprotectin test would have been managed and diagnosed if the test had not been available.

Comparator data from the Leeds Teaching Hospital NHS Trust, where faecal calprotectin testing is not used in either primary or secondary care, was available for a cohort of 103 patients. They were referred to secondary care with lower GI symptoms where IBS or IBD were likely and cancer was not suspected. Fifty‑five people (53%) had an endoscopy and 2 people were diagnosed with IBD. This gives a diagnostic yield from endoscopic investigations of 3.6% (2/55) for IBD. This is the only 'comparator' data available to assess the diagnostic yield of the locally developed guidelines.

In the York pilot project using locally agreed guidelines for faecal calprotectin testing in 5 primary care practices, 87 people (32.3%) had a specialist investigation (endoscopy, MRI or CT scan) to determine the cause of their symptoms. Of these:

Seventy‑three of these people had an endoscopic investigation (65 colonoscopy and 8 flexible sigmoidoscopy). Of these:

This gives a diagnostic yield from endoscopic investigations of 12.3% (9/73) for IBD for people managed in this care pathway.

Feedback was collected about how valuable GPs found faecal calprotectin testing throughout the project. Informal, face‑to‑face feedback was given by all of the pilot practices. There was positive feedback about the consultant gastroenterologist's visit to each of the practices at the start of the project. The visits provided an opportunity for clinicians to become acquainted with the new care and treatment pathway and with patient selection criteria. It also gave them a chance to discuss any concerns that they had.

A survey of all the GPs in the 5 pilot practices was carried out by the CCG to assess the use of the guidelines for faecal calprotectin testing. Twenty‑one GPs completed this and the results showed a high level of trust in the test results with most (19/21) agreeing that the testing had been useful in making their clinical decision. Eighteen respondents stated that they would continue to use the test in the diagnostic pathway.

A number of common compliance issues were identified during the audit:

Feedback from the GP survey suggested that education is needed to ensure that the test is used appropriately and the results understood, if the test is to be rolled out on a CCG‑wide basis.

Following this pilot, the CCG now needs to reach agreement with the local gastroenterologists and laboratory colleagues on the criteria for faecal calprotectin testing in primary care. The guideline generated 300 referrals for faecal calprotectin testing (269 initial and 31 repeat) across 5 practices in a 26‑week period (10 per month per practice). National general practice profiles, from 2013, published by Public Health England, indicate that the 5 practices involved in the pilot project had 28.4% of the CCG's population of people aged 20–60 years. Therefore, if testing were to be rolled out across the CCG, it is likely that there would be more than 2000 requests for faecal calprotectin tests each year from primary care.

The pilot indicates that, with guidance, the test is useful for identifying people with a low risk of IBD, but who have functional disorders that can be managed appropriately in primary care. The project team consider that GPs need greater awareness of the limitations of the test, how to manage the results, when to refer a patient to secondary care and with what urgency to make the referral.

Further learning from York Teaching Hospital NHS Foundation Trust and NHS Vale of York CCG can be found in this NICE shared learning case study published in 2018.

Table 4 Faecal calprotectin testing in secondary care January 2013 to January 2014

It should be noted that, as can be seen from table 4, 49 people with abnormal faecal calprotectin results did not have a procedure. The reasons for this are likely to include resolution of symptoms, endoscopy carried out elsewhere, and in some cases an alternative diagnosis being made before endoscopy.

The gastroenterology team considered that before the availability of faecal calprotectin testing, these patients would probably have had an invasive endoscopy procedure.

Local data indicated that the referral and endoscopy numbers were in line with national data, but that these were increasing year‑on‑year. This prompted the gastroenterology team to propose the introduction of faecal calprotectin testing in primary care as a way of enabling appropriate and fast‑tracked referral of patients with a suspicion of IBD to secondary care. It also prevented unnecessary IBS referrals and instead allowed primary care centres to manage the treatment of people in this group. The team was hoping to see a reduction in endoscopies in line with the results from the introduction of the test into secondary care.

The gastroenterology research group had used the AXON database for data specific to Wandsworth CCG to carry out a retrospective analysis of Hospital Episode Statistics data, and assessed referrals to outpatient lower‑gastrointestinal specialities (for example, gastroenterology and colorectal surgery) for people aged 16–40. In Wandsworth CCG, for this age group, there has been a consistent annual referral pattern of about 3000 patients to these outpatient specialities.

Through the CCG prioritisation process, the trust's business case to implement faecal calprotectin testing into all 43 GP practices in the CCG with support from 2 laboratories had already been approved. A joint project team was formed, which included a consultant gastroenterologist (project lead), the CCG GP lead, the laboratory leads, and a pathology project manager.

It was decided that calprotectin testing should be offered to people aged 16–40 presenting to their GP with lower gastrointestinal symptoms following a clinical assessment using Rome III criteria for IBS where a diagnosis is uncertain. A pathway algorithm was developed and agreed. People aged 16–18 were included because local experience suggested these young people would generally be referred directly to adult services or soon after their initial appointment with paediatric services. People aged over 40 were excluded because the team decided that this group, presenting to their GP with new changes in their bowel habit or other symptoms, would always warrant a referral to gastroenterology for specialist assessment. This algorithm was disseminated across the CCG to all practices with a supporting letter and additional information on the test.

GPs were advised that if the faecal calprotectin result was:

The project team reported that GP engagement was a major challenge to the adoption project, particularly in relation to data collection with GPs expressing concerns about data sharing and information governance. Data was collected from the central pathology database in secondary care. Data on follow‑up and outcomes of patients who had negative faecal calprotectin results (and therefore not referred to secondary care) were collected from the patient database at each GP practice.

During a 5‑month pilot project (March–July 2014) all requests from primary care for faecal calprotectin testing were identified from the local pathology database. Of the 43 practices, 29 made requests for 269 patients to have a faecal calprotectin test during this time. Of these patients, 61 (22.7%) had an abnormal result, compared with 35.3% of the patients tested in secondary care between January 2013 and January 2014. The project team have been unable to assess whether the results of the tests in primary care had an effect on GP decision‑making about endoscopy referral.

The St George's and Wandsworth team reported that if other CCGs and secondary care sites wish to introduce faecal calprotectin testing into primary care it is important to work together to engage with practices before implementation. This is particularly important for data collection, because it is helpful to define and agree the data that is to be collected.

At the end of the project the team analysed the St George's endoscopy database for the 8‑month period before and the 8‑month period after Wandsworth CCG GP practices were invited to ask for faecal calprotectin testing for people in the target age range (table 5).

Table 5 Lower gastro‑intestinal endoscopies before and after introduction of faecal calprotectin testing in primary care for people aged 16–40

The project team has been unable to show whether faecal calprotectin testing can reduce the CCG's lower gastrointestinal endoscopy workload for a number of reasons:

The project team now plan to continue to educate primary care practitioners about faecal calprotectin and to collect long‑term follow‑up data on the use of faecal calprotectin testing and its effect on referral patterns.