Clare Rutterford1, Monica Taljaard2, Stephanie Dixon3, Andrew Copas4, Sandra Eldridge5. 1. Centre for Primary Care and Public Health, Blizard Institute, Barts and The London School of Medicine and Dentistry, Queen Mary University of London, Yvonne Carter Building, 58 Turner Street, London E1 2AB, UK. Electronic address: c.m.rutterford@qmul.ac.uk. 2. Clinical Epidemiology Program, Ottawa Hospital Research Institute, Ottawa Hospital, Civic Campus, 1053 Carling Avenue, Civic Box 693, Ottawa, Ontario K1Y 4E9, Canada; Department of Epidemiology and Community Medicine, University of Ottawa, Ottawa, Ontario, Canada. 3. Department of Epidemiology and Biostatistics, Schulich School of Medicine and Density, University of Western Ontario, 1151 Richmond St, London, Ontario N6A 3K7, Canada. 4. Hub for Trials Methodology Research, MRC Clinical Trials Unit at University College London, Aviation House, 125 Kingsway, London, WC2B 6NH, UK. 5. Centre for Primary Care and Public Health, Blizard Institute, Barts and The London School of Medicine and Dentistry, Queen Mary University of London, Yvonne Carter Building, 58 Turner Street, London E1 2AB, UK.
Abstract
OBJECTIVES: To assess the quality of reporting and accuracy of a priori estimates used in sample size calculations for cluster randomized trials (CRTs). STUDY DESIGN AND SETTING: We reviewed 300 CRTs published between 2000 and 2008. The prevalence of reporting sample size elements from the 2004 CONSORT recommendations was evaluated and a priori estimates compared with those observed in the trial. RESULTS: Of the 300 trials, 166 (55%) reported a sample size calculation. Only 36 of 166 (22%) reported all recommended descriptive elements. Elements specific to CRTs were the worst reported: a measure of within-cluster correlation was specified in only 58 of 166 (35%). Only 18 of 166 articles (11%) reported both a priori and observed within-cluster correlation values. Except in two cases, observed within-cluster correlation values were either close to or less than a priori values. CONCLUSION: Even with the CONSORT extension for cluster randomization, the reporting of sample size elements specific to these trials remains below that necessary for transparent reporting. Journal editors and peer reviewers should implement stricter requirements for authors to follow CONSORT recommendations. Authors should report observed and a priori within-cluster correlation values to enable comparisons between these over a wider range of trials.
OBJECTIVES: To assess the quality of reporting and accuracy of a priori estimates used in sample size calculations for cluster randomized trials (CRTs). STUDY DESIGN AND SETTING: We reviewed 300 CRTs published between 2000 and 2008. The prevalence of reporting sample size elements from the 2004 CONSORT recommendations was evaluated and a priori estimates compared with those observed in the trial. RESULTS: Of the 300 trials, 166 (55%) reported a sample size calculation. Only 36 of 166 (22%) reported all recommended descriptive elements. Elements specific to CRTs were the worst reported: a measure of within-cluster correlation was specified in only 58 of 166 (35%). Only 18 of 166 articles (11%) reported both a priori and observed within-cluster correlation values. Except in two cases, observed within-cluster correlation values were either close to or less than a priori values. CONCLUSION: Even with the CONSORT extension for cluster randomization, the reporting of sample size elements specific to these trials remains below that necessary for transparent reporting. Journal editors and peer reviewers should implement stricter requirements for authors to follow CONSORT recommendations. Authors should report observed and a priori within-cluster correlation values to enable comparisons between these over a wider range of trials.
Authors: David M Murray; Sherri L Pals; Stephanie M George; Andrey Kuzmichev; Gabriel Y Lai; Jocelyn A Lee; Ranell L Myles; Shakira M Nelson Journal: Prev Med Date: 2018-03-16 Impact factor: 4.018