Assessing the Quality of Abstracts in Randomized Controlled Trials Published in High Impact Cardiovascular Journals.

Published

Journal Article

BACKGROUND: In the busy world of cardiovascular medicine, abstracts may be the only part of a publication that clinicians read. Therefore, it is critical for abstracts to accurately reflect article content. The extended CONSORT (Consolidated Standards of Reporting Trials) Statement for Abstracts was developed to ensure high abstract quality. However, it is unknown how often adherence to CONSORT guidelines occurs among cardiovascular journals. METHODS AND RESULTS: We searched MEDLINE for randomized controlled trials published in 3 major cardiovascular journals ( Circulation, Journal of the American College of Cardiology, and European Heart Journal) from 2011 to 2017. Post hoc, interim, and cost-effective analyses of randomized controlled trials were excluded. Two independent investigators extracted the data using a prespecified data collection form and a third investigator adjudicated the data. The primary outcome was frequency of subcategory adherence to CONSORT guidelines. A total of 478 abstracts were included in the analysis. Approximately half of the abstracts (53%; 255/478; 95% CI, 49%-57%) identified the article as randomized in the title. All abstracts detailed the interventions for both study groups (100%) and 81% (95% CI, 78%-85%) reported trial registration. Methodological quality reporting was relatively low: 9% (45/478; 95% CI, 6%-12%) described participant eligibility criteria with settings for data collection, 43% (204/478; 95% CI, 39%-47%) reported details of blinding, and <1% (4/478; 95% CI, 0%-2%) reported allocation concealment. Approximately 60% (301/478; 95% CI, 59%-67%) of the included abstracts provided primary outcome results while 55% (262/478; 95% CI, 51%-60%) reported harms or adverse effects. CONCLUSIONS: There is a high prevalence of nonadherence to CONSORT guidelines among leading cardiovascular journals. Efforts by editors, authors, and reviewers should be made to increase adherence and promote transparent and unbiased presentation of study results.

Full Text

Duke Authors

Cited Authors

  • Khan, MS; Shaikh, A; Ochani, RK; Akhtar, T; Fatima, K; Khan, SU; Mookadam, F; Murad, MH; Figueredo, VM; Doukky, R; Krasuski, RA

Published Date

  • May 2019

Published In

Volume / Issue

  • 12 / 5

Start / End Page

  • e005260 -

PubMed ID

  • 31030545

Pubmed Central ID

  • 31030545

Electronic International Standard Serial Number (EISSN)

  • 1941-7705

Digital Object Identifier (DOI)

  • 10.1161/CIRCOUTCOMES.118.005260

Language

  • eng

Conference Location

  • United States