BMJOG Blog: Depression and IBD, artificial intelligence in colorectal cancer, coeliac disease and managing underperforming endoscopists

In this month’s blog, Dr. Christian Selinger, one of BMJOG’s associate editors and consultant Gastroenterologist in Leeds, selects his top publications from the last month and reflects on the findings and impact of these papers. These articles cover a wide-range of topics, reflecting contemporary luminal GI research in IBD, colorectal cancer, coeliac disease and endoscopy. Following our editorial board meeting this month we are extremely pleased with the progress of BMJOG over the last 12 months- submissions, publications and interest has built hugely. We hope to build on the success of these blog posts and increase the social media impact further to improve the educational value of the journal and content.

In the first paper Dr. Selinger has selected an article from Blackwell and colleagues on the association between antidepressant use and steroid dependency in IBD. The authors report corticosteroid dependency occurring more frequently in continuous antidepressant use (SSRI and TCAs) compared with non-users (19% vs 24% vs 14%, respectively, χ2 p=0.002). The authors included a large number of patients in this population-based study (n=6373) and conclude that antidepressant medication use may be a flag for worse clinical outcomes in patients with ulcerative colitis. Despite these very interesting data it is not clear whether antidepressant medication use is a consequence of poor IBD control resulting in low mood, or if patients with depression are more likely to end up on steroids through not accessing care/poor selfcare, resulting in worse long-term IBD control. Overall, it is vital to assess the psychological state of an individual when reviewing patients in clinic.

The second manuscript centres on a hugely interesting and contemporary area in healthcare, the use of artificial intelligence to aid with diagnosis and treatment of patients. In this paper Ayling et al report their use of the ColonFlag score for prioritisation of endoscopy in colorectal cancer. ColonFlag is an artificial intelligence algorithm based on full blood count parameters, age and sex, and was shown in this analysis to have equal sensitivity and better specificity than faecal-Hb at a cut-off of 10 µg/g for detection of colorectal carcinoma. The authors also compared ColonFlag to faecal immunochemical testing (FIT), which has a much broader evidence base and observe combination of the ColonFlag algorithm with FIT shows slightly better results than either alone. We are very interested to see non-stool based approaches at colorectal cancer risk stratification. Interestingly a cost-effectiveness test would be very helpful to determine it’s utility as a screening tool. AI’s use is increasing and is already being seen as the future in detection of abnormalities in endoscopy. This study points towards the potential utility of AI to aid with screening in a relatively common condition, however a bigger and prospective study is needed to demonstrate the robustness of these findings and transition to clinical use. Overall Dr. Selinger and I feel that this is an exciting idea, with good results, but not ready for clinical application yet.

In the final paper from BMJOG we discuss a slightly controversial concept proposed by Taylor et al, namely do gastroenterologists have medical inertia towards coeliac disease? The authors report on a multicentre study from the UK and compared a number of outcome measures for patients with Coeliac disease and patients with IBD. The authors report that patients with suspected Coeliac disease wait longer for initial endoscopy, guidelines for duodenal biopsy are frequently not followed, and gastroenterologist’s knowledge of Coeliac disease prevalence was poor. The authors conclude that these data suggest ‘medical inertia towards Coeliac disease’. This paper raises several interesting points. Dr Selinger and I agree that Coeliac patients may wait longer than IBD patient for a diagnosis, but the paper does not study the impacts of this delay and it is possible it may have no consequences in the long-term. We also are intrigued by the use of IBD as a ‘control’ group, with the trajectories for patients adhering to treatment differing widely post-diagnosis. It is clear that Coeliac disease is a common and important condition to diagnose and treat. Improving recognition of the illness remains imperative, however, the ongoing involvement of physicians in stable patients and comparability to IBD remain up for debate.

In our external paper, Dr. Selinger has selected an article from one of our sister journals, Frontline Gastroenterology, on the Joint Advisory Group on Gastrointestinal Endoscopy (JAG) framework for managing underperformance in gastrointestinal endoscopy. This position statement from JAG seeks to address underperformance in a supportive and constructive way by understanding the underlying reasons for underperformance and helping endoscopist improve rather than punishment. This paper is especially interesting in the era of COVID-19 compounding the reduced endoscopy numbers during training. We acknowledge that this combination has the potential to result in reduced confidence/performance at ‘independent’ consultant endoscopist level. Whilst this paper highlights the ways in which support can be provided, including coaching, mentoring, training and upskilling, if adequate quality is not met after all these interventions then there must be a system to remove providers in order to safeguard patients and outcomes.


(Visited 246 times, 1 visits today)