• Mary Lou V H Greenfield, Jill M Mhyre, George A Mashour, James M Blum, Eugene C Yen, and Andrew L Rosenberg.
• Department of Anesthesiology, University of Michigan Health System, 1500 East Medical Center Dr., Ann Arbor, MI 48109-5861, USA.
• Anesth. Analg. 2009 Jun 1;108(6):1916-21.

BackgroundWe previously assessed all randomized controlled trials (RCTs) from four anesthesiology journals from January 2000 to December 2000. We identified key areas for improvement in the study protocol design and implementation and in data analyses. This study was repeated for the year 2006 to determine if improvements have occurred during the 6-yr interval.MethodsAll RCTs published in 2006 in four anesthesiology journals (Anesthesiology, Anesthesia & Analgesia, Anaesthesia, and Canadian Journal of Anesthesia) were retrieved using a MEDLINE search. Of 2164 articles published in 2006, 200 papers met these search criteria and were considered valid for analysis. We completed a 14-item, validated assessment tool used in our previous study to determine a quality score for each article. Four clinical reviewers each assessed 50 articles, and one reviewer assessed all 200 articles. Points were assigned by consensus. Scores were weighted and compared with the results from the year 2000.ResultsQuality scores improved from the year 2000 to 2006, from a mean overall quality score of 44% (95% CI = 42, 46) to a mean score of 58% (95% CI = 55, 60). Specific areas of study, quality assessment demonstrating improvement, included sample size estimates (52% vs 86%, P < 0.0001), major end-points (44% vs 99%, P < 0.0001), and discussion of side effects (68% vs 82%, P = 0.0019). Low quality scores remained for randomization blinding (4% vs 19% P < 0.0001), observer blinding to continuing studies (1% vs 5% P = 0.116), and post-beta estimates in trials with negative outcomes (16% vs 18%, P < 0.87).ConclusionsThere appears to have been a general improvement in the overall quality of RCT reporting among the major anesthesiology journals from the year 2000 to 2006. However, many articles could be improved with respect to randomization blinding, observer blinding to continuing study results (i.e., no unplanned interim data analysis), and a full discussion of Type II error in negative trials. Responsibility to improve the quality of the anesthesiology literature rests with investigators to design, implement and report high quality RCTs, and with peer reviewers and journal editors to set the standard for manuscript reporting. Periodic reassessments of the literature can serve to improve and maintain the quality of clinical trials reporting.

13 keywords...

hide…