This preview shows pages 1–2. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: Reporting of Clinical Trials This lecture will address several issues in reporting: What needs to be included in a formal write-up of a clinical trial. Guidelines for writing a paper vs. writing an abstract, poster or talk. Medical paper format for reporting statistical issues. Professional issues: authorship, responsibility. What needs to be in a clinical trial publication? Important not only to carry out clinical trial carefully, but also to report clearly. Despite peer review process, many reports in medical literature fail to give adequate information to judge the goals, conduct and analysis of clinical trials. In 1994, the first CONSORT (Consolidated Standards of Reporting Trials) was published. An update was published in 2001; it appeared in a number of journals simultaneously. The standards include both a checklist and a flow diagram. At the same time as the update, a study was published comparing reporting before first CONSORT vs. after. Sample size of 211 clinical trial reports from four journals: BMJ, JAMA, Lancet had adopted CONSORT as a standard NEJM did not. Outcomes included number of CONSORT items reported properly, overall quality. Pre-CONSORT mean number of items was 23.4, mean change was increase of 3.7 items. Increase in all 4 journals but significant in the 3 adopters. The improvement was not universal! These are the top-tier medical journals and they were still not perfect. One study of 122 trials of selective serotonin reuptake inhibitors for depression found that only 1 of them (0.8%) described randomization adequately. These guidelines affect statisticians many ways. 1. You need to plan the trial with guidelines in mind, so that you will address each issue. 2. You need to document each item as you go. 3. Interim reporting and tables should follow the structure laid out in the flow chart and help prepare for writing the paper. 4. As a co-author, you should make sure that each item is adequately reported, especially the statistical issues....
View Full Document
- Winter '06