RT Journal Article
SR Electronic
T1 Quality of reporting in abstracts of RCTs published in emergency medicine journals: a systematic survey of the literature suggests we can do better
JF Emergency Medicine Journal
JO Emerg Med J
FD BMJ Publishing Group Ltd and the British Association for Accident & Emergency Medicine
SP emermed-2019-208629
DO 10.1136/emermed-2019-208629
A1 Germini, Federico
A1 Marcucci, Maura
A1 Fedele, Marta
A1 Galli, Maria Giulia
A1 Heath, Tevin
A1 Mbuagbaw, Lawrence
A1 Salvatori, Valentina
A1 Veronese, Giacomo
A1 Worster, Andrew
A1 Thabane, Lehana
YR 2019
UL http://emj.bmj.com/content/early/2019/11/06/emermed-2019-208629.abstract
AB Objective We investigated the association between the publication of the Consolidated Standards of Reporting Trials extension for abstracts (CONSORT-EA) and other variables of interest on the quality of reporting of abstracts of randomised controlled trials (RCTs) published in emergency medicine (EM) journals.Methods We performed a survey of the literature, comparing the quality of reporting before (2005–2007) with after (2014–2015) the publication of the dedicated CONSORT-EA in 2008. The quality of reporting was measured as the sum of items of the CONSORT-EA checklist reported in each abstract, ranging from 0 to 15. The main explanatory variable was the period of publication: pre-CONSORT-EA versus post-CONSORT-EA public. Other explanatory variables were journal’s endorsement of the CONSORT statement, number of centres participating in the study, study’s sample size, type of intervention, significance of results, source of funding and study setting. We analysed the data using generalised estimation equations, performing a univariate and a multivariable analysis.Results We retrieved 844 articles, and randomly selected 60 per period for review, after stratifying for journal. The mean (SD) number of items reported was 6.4 (1.9) in the period before and 6.9 (1.8) in the period after the publication of the CONSORT-EA, with an adjusted mean difference (aMD) of 0.47 (95% CI −0.13 to 1.06). Abstracts of trials of pharmacological interventions had a significantly larger mean number of reported items than those of trials of non-pharmacological interventions (aMD 1.59; 95% CI 0.94 to 2.24).Conclusions The quality of reporting in abstracts of RCTs published in EM journals is low and was not significantly impacted by the publication of a dedicated CONSORT-EA.