Research Integrity in the Field of Atopic Dermatitis Treatment

Abstract

In this month's issue of Clinical and Experimental Allergy, we highlight the prize-winners from the recent highly successful BSACI conference and present a Cochrane systematic review of topical anti-inflammatory treatments for eczema (atopic dermatitis). Prize-winner abstracts reflected the breadth and depth of research presented at the BSACI conference, which continues to grow and will move to a new, larger venue next year. Please mark 16–18 October at the International Convention Centre in Newport, Wales in your diary! The Cochrane review findings are striking. Not so much for the findings on effectiveness and safety of topical eczema treatments, but for the lack of independence and transparency in the included clinical trials. Eczema is a common, burdensome condition—probably the most burdensome skin condition affecting humans—and topical anti-inflammatory treatments have been used for generations to manage eczema symptoms. In this first comprehensive Cochrane review of the topic, authors identified almost 400 randomised controlled trials in over 50,000 people with eczema. One might hope that such a significant human effort would have yielded some reliable information to guide clinician, carers and people with eczema in making treatment decisions. However, the standout finding in this review was the lack of transparency in the field (Figure 1). Risk of bias was judged to be high in 89% of trials, with the most common issue being concern about selective reporting. While older clinical trials would not be expected to register their plans on a public clinical trial registry or to make their protocol publicly available—it is of some concern that even the modern trials commonly lacked appropriate trial registration. Trial findings were also rarely reported completely, meaning that for many trials and outcomes, it was not possible to include data in meta-analysis. A common issue here was trialists not stating the number of participants evaluated for an outcome.

This Cochrane review is a good example of the ways in which scientific progress, clinical practice and the health of relevant populations are held back when clinical trials are not conducted and reported in a transparent manner. It is relevant that only a tiny minority of these trials were funded independently—where stated, the funder was almost always the manufacturer of one of the topical treatments being tested. It is well-established that clinical trials funded in this way tend to report findings as more favourable, and in commercial trials, there are also obvious incentives for non-transparent reporting, unless this is legally mandated. While it is for the lawyers and regulators to legislate and oversee compliance with transparent reporting of clinical trials, journals also have a place in promoting better clinical trial practice. This is why Clinical & Experimental Allergy places great emphasis on transparency in reporting, especially for potentially influential manuscripts such as clinical trial reports [2]. We publish study protocols and null findings; and we encourage authors to make all relevant materials publicly available—from pre-trial protocol, to statistical analysis plans, participant information sheets and, where ethically feasible, full clinical trial datasets [3-7].

A charity that has consistently advocated for increased transparency of clinical trial reporting is Cochrane. We regularly publish short summaries of Cochrane reviews in Clinical & Experimental Allergy. Please get in touch with Parisut Kimkool [email protected] if you would like to write a Cochrane Corner about a recent Cochrane review that is relevant to the field of Allergy & Immunology. In this issue, authors summarise a recent Cochrane review about pharmacological treatment of gastro-oesophageal reflux in children. These treatments are widely used and prescribed in children—including thickeners, prokinetic agents, H2 antagonists and proton pump inhibitors. Indeed, in infants, there is evidence for increasing use of some of these products in recent years (Figure 2). There is a familiar pattern in the findings of this Cochrane review. For most included trials, there were no relevant data in the public domain that could be extracted. Even for those trials where data could be extracted, the heterogeneity of trial design and outcome reporting meant that meta-analysis could not be conducted. Partly as a consequence of the authors' inability to identify and analyse relevant outcomes from the 36 included trials in 2251 children and adults, findings were inconclusive—with very low certainty evidence for the interventions studied. Authors conclude that further trials are required. Equally important however, is that such future trials should be transparently reported so that all relevant clinical trial data are publicly available to support meta-analyses and evidence-informed decision-making.

One approach that systematic reviewers sometimes use, to address the lack of transparency and reporting in some clinical trials, is to use individual participant data (IPD) from the identified trials. This does seem to improve data quality and certainty of systematic review conclusions [9]. IPD meta-analysis is time-consuming and expensive, requiring data sharing agreements and relationship building with clinical trial teams. However, until such time as clinical trials groups are legally obliged to make both their pre-study plans and their full trial datasets publicly available, IPD will have an important place in the evidence ecosytem.

R.J.B. wrote the first draft. M.H.S. reviewed and approved the manuscript. The corresponding author takes full responsibility for this article.

The authors declare no conflicts of interest.