Intention to treat and per protocol analyses showed different results in a diagnostic cluster randomized controlled trial evaluating fecal calprotectin in primary care

Sophie Ansems, Marjolein Berger, Patrick Van Rheenen, Karin Vermeulen, Michiel De Boer, Gea Holtman

Background:

General practitioners (GPs) often struggle with the diagnostic uncertainty of distinguishing functional gastrointestinal disorders (FGID) from organic disorders in children with chronic gastrointestinal symptoms. It is essential to limit referrals of children with FGID while not missing organic disorders such as inflammatory bowel disease (IBD). Fecal calprotectin (FCal) testing may be useful as it can safely rule out IBD.

Research questions:

This study aims to evaluate whether FCal testing can reduce referral rates to pediatric specialist care.

Method:

In a pragmatic clustered randomized controlled trial, Dutch GP practices were randomly assigned to an intervention or control group. The intervention group received an online training about the indication, interpretation, follow-up en communication of FCal testing and used it at their discretion. The control group followed Dutch GP guidelines that do not recommend the test. The primary outcome was the referral rate to pediatric care within six months. GPs enrolled children aged 4 to 18 with chronic abdominal pain and/or diarrhoea, aiming to recruit 406 children. Participants were not blinded to the intervention, but researchers were blinded during statistical analysis. We used multilevel logistic regression, on both an intention-to-treat and per-protocol basis.

Results:

Between 01-10-2019 and 01-07-2021, 40 GP practices assigned to the intervention group and 44 to the control group enrolled 203 and 202 children, respectively. Alarm symptoms were more frequently documented in the intervention (26.6%) than the control group (8.9%), while the referral rates were similar (22.8 vs 21.9%, adjusted OR 0.94; 95% CI 0.57-1.54). In the per protocol analysis we found a reduction in referral rates (5.8 vs 20.3%, adjusted OR 0.21; 95% CI 0.09-0.50).

Conclusions:

Our results do not support the routine use of fecal calprotectin in primary care until the reasons why GPs did not adhere to the recommendations given in the online training are known.

Points for discussion:

How can we improve adherence of physicians to complex interventions, like this new diagnostic testing strategy?

What other factors influence the GP's referral behavior in children with chronic gastrointestinal symptoms?

Would you, being a GP, after listening to this presentation, use this test?

#54