Discussion: Specific Statistical Considerations Relevant to the Design andAnalysis of Gingivitis Trials

摘要

Dr. Kingman's presentation on statistical issues in gingivitis trials wasreviewed and found to be comprehensive and detailed. The power of different statistical tests was considered and statistical criteria for concluding superiority and equivalency evaluated. It seems that there will be no automatic prescriptions. Specific methodologies for a clinical trial must be selected from among many alternatives based on a well-developed statement of motivation. But the issues are complicated, are still evolving, and will have to be watched closely for some time. Dr. Kingman's paper provides a very useful framework for this task. Implications of his findings on power and the strong recommendation for a placebo control group in equivalency trials are discussed in detail.