The quality of randomised controlled trials may be better than assumed
http://www.100md.com
《英国医生杂志》
1 ABC Foundation School of Medicine, Avenida Rebou?as 3387 S?o Paulo, Brazil, 05410-040, 2 ABC Foundation School of Medicine, Division of Medical Oncology, University of Colorado Health Sciences Center, Denver, CO 80262, USA
Correspondence to: A del Giglio sandrabr@netpoint.com.br
As readers of published articles, it is reassuring to know that the quality of published data is probably better than expected from the reporting of the methods. Soares and colleagues have addressed the discrepancies between the proposed methods in original research protocols and those reported in the final article for all 56 randomised controlled trials conducted by the Radiation Therapy Oncology Group since its creation in 1968. Good quality experimental designs were more often adhered to during the conduct of the studies than suggested by the final reports.
As readers of systematic reviews, however, we may have underestimated the quality of experimental methods, especially if reviewers had not checked the original protocols. This may have led to the exclusion of some potentially good quality papers from systematic reviews.1 This would also affect the scoring of quality of papers as a part of a meta-analysis.2 3 Soares and colleagues suggest that this problem can be avoided by accessing the original protocols and contacting trialists to check the accuracy of reported methods. Authors should also adhere to the Consolidated Standards of Reporting of Trials (CONSORT) statement.4
In the near future, papers may be accompanied by internet links to their original protocol and data, allowing researchers to download crude data for meta-analysis. Until that time, it should be borne in mind that although papers fail to report important design and methodological features, they were probably done, especially if the papers are authored by a body such as the Radiation Therapy Oncology Group, the protocols of which undergo rigorous review.
Contributors: AdG was the main author of the commentary and wrote the final manuscript. He will act as guarantor. LJC made modifications to the manuscript and contributed to the final manuscript.
Funding: None.
Competing interests: None declared.
References
Pignon JP, Hill C. Meta-analyses of randomised clinical trials in oncology. Lancet Oncol 2001;2: 475-82.
Detsky AS, Naylor CD, O'Rourke K, McGeer AJ, L'Abbe KA. Incorporating variations in the quality of individual randomized trials into meta-analysis. J Clin Epidemiol 1992;45: 255-65.
Moher D, Pham B, Jones A, Cook DJ, Jadad AR, Moher M, et al. Does quality of reports of randomised trials affect estimates of intervention efficacy reported in meta-analyses? Lancet 1998;352: 609-13.
Moher D, Schulz KF, Altman DG. The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomised trials. Lancet 2001;357: 1191-4.(Auro del Giglio, chairman)
Correspondence to: A del Giglio sandrabr@netpoint.com.br
As readers of published articles, it is reassuring to know that the quality of published data is probably better than expected from the reporting of the methods. Soares and colleagues have addressed the discrepancies between the proposed methods in original research protocols and those reported in the final article for all 56 randomised controlled trials conducted by the Radiation Therapy Oncology Group since its creation in 1968. Good quality experimental designs were more often adhered to during the conduct of the studies than suggested by the final reports.
As readers of systematic reviews, however, we may have underestimated the quality of experimental methods, especially if reviewers had not checked the original protocols. This may have led to the exclusion of some potentially good quality papers from systematic reviews.1 This would also affect the scoring of quality of papers as a part of a meta-analysis.2 3 Soares and colleagues suggest that this problem can be avoided by accessing the original protocols and contacting trialists to check the accuracy of reported methods. Authors should also adhere to the Consolidated Standards of Reporting of Trials (CONSORT) statement.4
In the near future, papers may be accompanied by internet links to their original protocol and data, allowing researchers to download crude data for meta-analysis. Until that time, it should be borne in mind that although papers fail to report important design and methodological features, they were probably done, especially if the papers are authored by a body such as the Radiation Therapy Oncology Group, the protocols of which undergo rigorous review.
Contributors: AdG was the main author of the commentary and wrote the final manuscript. He will act as guarantor. LJC made modifications to the manuscript and contributed to the final manuscript.
Funding: None.
Competing interests: None declared.
References
Pignon JP, Hill C. Meta-analyses of randomised clinical trials in oncology. Lancet Oncol 2001;2: 475-82.
Detsky AS, Naylor CD, O'Rourke K, McGeer AJ, L'Abbe KA. Incorporating variations in the quality of individual randomized trials into meta-analysis. J Clin Epidemiol 1992;45: 255-65.
Moher D, Pham B, Jones A, Cook DJ, Jadad AR, Moher M, et al. Does quality of reports of randomised trials affect estimates of intervention efficacy reported in meta-analyses? Lancet 1998;352: 609-13.
Moher D, Schulz KF, Altman DG. The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomised trials. Lancet 2001;357: 1191-4.(Auro del Giglio, chairman)