Reporting of harm and safety results in randomized controlled trials published in 5 dermatology journals

Background: Randomized controlled trials (RCTs) are considered the gold standard for assessing efficacy and short-term harm of medicines. However, several studies have come to the conclusion that harm is less well reported than efficacy outcomes.

Objective: To describe harm reporting in publications on dermatological RCTs and assess parameters that could influence the quality of harm reporting.

Methods: Methodologic systematic review of dermatologic RCTs published from 2010 to 2014 in 5 dermatological journals.

Results: Among 110 assessed publications on RCTs, 80 (73%) adequately reported harm and 52% adequately reported its severity. Overall, 40% of the assessed manuscripts perfectly reported and discussed harm. The adequate reporting of harm was significantly associated with the type of trial (odds ratio [OR] 4.41, 95% confidence interval [CI] 1.60-12.35 for multicenter compared with monocentric trials) and having a predefined method for collecting harm data (OR 5.93, 95% CI 2.26-15.56). Reporting of harm severity was better in pharmacologic trials (OR 6.48, 95% CI 2.00-21.0) compared with nonpharmacologic trials and in trials for which a method for collecting harm (OR 5.65, 95% CI 2.00-16.4) and its severity (OR 3.60, 95% CI 1.00-12.8) was defined before the study onset.

Limitations: Assessment was restricted to RCTs and 5 dermatological journals.

Conclusion: Harm is quite well reported in dermatologic journals. Efforts should be made on reporting severity of harm.