Quality of reporting and concordance between sources of adverse events in the treatment of moderate-to-severe psoriasis: a cross-sectional study of RCTs from a Cochrane systematic review

Objectives Incomplete reporting of safety outcomes in quality and availability of safety reporting in published articles of randomized clinical trials (RCTs) were described in different medical areas. The number of RCTs assessing systemic treatments for psoriasis has increased considerably. Complete and precise reporting of safety is mandatory for the efficacy/harms balance evaluation. We aimed to assess the quality and availability of safety reporting in published RCTs assessing systemic treatments for psoriasis, as well as the concordance of data between published trials and clinicaltrials.gov (CT). Study Design and Setting We included all RCTs in adults initiated after September 2009, assessing systemic psoriasis treatments compared with placebo or with an active comparator. All trials were selected in duplicate by two independent authors from the latest search of the dedicated Cochrane review. We described quality of safety reporting for all published RCTs, using a modified CONSORT harms scale by using descriptive analysis, and a composite score of 3 key items of safety report. For each RCT, data on adverse events (AEs)/serious AEs (SAEs) were extracted from the publication and CT: total number of AEs/SAEs, patients with AEs/SAEs, SAEs by system organ class classification and deaths. These data were compared between sources for each RCT. Results In total, 128 trials were included in the analysis of reporting quality, and 76 in the analysis of data concordance between sources. The median number of reported CONSORT harms items per article was 9 out of 18 (IQR 7-10), and mean number was 8.39 (sd = 3.02). Items in the methods section were the least frequently reported. The proportion of RCTs reporting the number of SAEs and death were significantly higher on ClinicalTrials.gov than in the published article ((100% (76/76) vs 88.2%, McNemar test, p <0.0016). At least one discrepancy between sources for SAE safety data was found in 30/76 (39.5%) RCTs. Discussion Shortcomings and gaps in the quality of safety reporting in publications of RCTs of systemic psoriasis treatments have been identified. A lack of data in published articles and discrepancies between published articles and CT data complete this finding.