How well do reports of clinical trials in the orthodontic literature comply with the CONSORT statement?

Abstract

Objectives: To assess whether clinical trials published in the orthodontic literature comply with the Consolidated Standards of Reporting Trials (CONSORT) statement, and whether compliance has changed over time.

Design: Retrospective study.

Sample: Clinical trials published in four orthodontic journals (American Journal of Orthodontics and Dentofacial Orthopedics, Angle Orthodontist, European Journal of Orthodontics, and Journal of Orthodontics) at three time points [1995/6 (pre-CONSORT), 2000/1 (post-CONSORT) and 2005/6 (post revised-CONSORT)].

Methods: Trials were scored using a 36-point checklist developed from the CONSORT statement. Each item on the checklist was scored as either included, not included, or not applicable if the trial design did not allow its inclusion. The proportion of trials describing randomization, allocation concealment, and blinding was also calculated. The mean scores were compared between time points and between journals.

Results: A total of 151 reports were included. The mean score was 41·5% (SD 10·3). The mean post-CONSORT score was higher than the pre-CONSORT score (P = 0·01; WMD, 5·73; 95% CI, 1·64-9·82), and the mean post revised-CONSORT was also higher than the pre-CONSORT score (P = 0·01; WMD, 4·44; 95% CI, 1·02-7·87). At the post revised-CONSORT time point, the mean score in the journals which had adopted the CONSORT statement was higher than for those which had not (P<0·001; chi square, 17·78; df = 3). Reporting of randomization, allocation concealment and blinding was inadequate at all three time points.

Conclusions: The reporting of clinical trials in the orthodontic literature has improved since the publication of the CONSORT statement, particularly in journals which have adopted the statement. However, the reporting of randomization, allocation concealment and blinding remains inadequate.