Improved Reporting of Randomized Controlled Trials in the Urologic Literature.

Published

Journal Article

BACKGROUND: Well-designed randomized controlled trials (RCTs) have the potential to provide high-quality evidence to inform questions of therapy and prevention, but this potential is contingent on the use of appropriate methods and transparent reporting. OBJECTIVE: To systematically assess the quality of urology RCT reporting and identify trends over time. DESIGN, SETTING, AND PARTICIPANTS: All RCTs published in four leading urology journals in 2013 were identified and compared to a prior analysis of studies from 1996 and 2004. Two reviewers abstracted data based on the Consolidated Standards of Reporting Trials (CONSORT) checklist. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: A summary reporting score (range: 0-22) for each study was determined. Mean overall scores for 1996, 2004, and 2013 were compared using analysis of variance. We used χ2 to compare the reporting frequency of individual criteria. RESULTS AND LIMITATIONS: Mean CONSORT scores for RCTs were 15.6±2.0 in 2013 (n=82), 12.0±0.3 in 2004 (n=87), and 10.2±0.3 in 1996 (n=65); p<0.01. Key deficiencies remain in reporting methods of allocation concealment and group assignment (selection bias), and blinding of participants, personnel, and outcome assessors (performance and detection bias). Study limitations are potential reviewer bias resulting from lack of journal deidentification and the relatively low number of studies reviewed. CONCLUSIONS: There has been a substantial improvement in reporting quality of RCTs in urology since CONSORT. Some methodological criteria remain underreported, and increased efforts are necessary to further this improvement. PATIENT SUMMARY: Treatment decisions are often based on data from randomized controlled trials. We looked at whether these trials in urology are transparent in reporting their design and conduct using a framework known as the CONSORT criteria and found significant improvements over time. Some areas of deficiency remain, and our paper aimed to highlight these drawbacks to promote continued high-quality research.

Full Text

Duke Authors

Cited Authors

  • Narayan, VM; Cone, EB; Smith, D; Scales, CD; Dahm, P

Published Date

  • December 2016

Published In

Volume / Issue

  • 70 / 6

Start / End Page

  • 1044 - 1049

PubMed ID

  • 27503840

Pubmed Central ID

  • 27503840

Electronic International Standard Serial Number (EISSN)

  • 1873-7560

Digital Object Identifier (DOI)

  • 10.1016/j.eururo.2016.07.042

Language

  • eng

Conference Location

  • Switzerland