Systematic reviews in dermatology: opportunities for improvement

Systematic reviews and meta-analyses stand atop the level-of-evidence pyramid.¹ These articles are often highly cited and frequently influence guidelines and clinical decision-making at the point of care. Given their significant impact, it is paramount that these studies are conducted with the utmost quality, and the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines have been developed to promote best practices for these studies.²

In this issue of BJD, Croitoru et al. use the PRISMA guidelines to evaluate the quality of systematic reviews published in dermatology journals between 2013 and 2017.³ They find that the vast majority of systematic reviews had at least one non-reported PRISMA item.

Protocol and registration was one of the most often underreported items, with only 15% of systematic reviews stating that they had registered a protocol. Increasing protocol registration is important to improve the quality of systematic reviews published in the dermatology literature. Similar to the role of clinical trial registration in registries such as ClinicalTrials.gov, registration of systematic reviews in registries such as The International Prospective Register of Systematic Reviews (PROSPERO) can promote transparency and methodological rigor for these important studies.⁴ The finding by Croitoru et al. that stated use of a registered protocol was associated with better PRISMA reporting supports the value of protocol registration.

It is important to keep in mind that a primary purpose of protocol registration is to increase transparency and still allow for changes to methodology after registration. Interestingly, Croitoru et al. did not register this systematic review (although they clearly stated this and the rationale behind the decision in their methods section), but they did include a predefined protocol. They had to deviate from their original proposed analysis plan owing to low event rates, which precluded their planned use of logistic regression for some outcomes. When protocols are registered and deviations are documented, it gives readers the opportunity to judge for themselves whether a modification was justified while still allowing the authors flexibility to adjust to unexpected challenges encountered while conducting the review.

Another important finding reported by Croitoru et al. is that funding source and risk-of-bias assessments were frequently underreported. While many studies were unfunded, given the potential impact of systematic reviews on guideline development and treatment patterns among clinicians, it is important to clearly state funding or lack thereof to ensure transparency regarding potential conflicts of interest.

In 2019, BJD began requiring registration of systematic reviews and meta-analyses prior to publication.⁵ However, not all dermatology journals require protocol registration. There is an opportunity to improve the quality of systematic reviews and meta-analyses published in our field by consistently requiring registration across journals. Broadly instituting such a practice will educate authors about the importance of study registration and will ensure that published systematic reviews are conducted to the highest quality standards. We strongly encourage all authors and editors to register their systematic reviews and use PRISMA guidelines when performing systematic reviews.