Search

It is healthy to periodically question things that we have long taken for granted. Randomized clinical trials are the gold standard for evaluating the benefits and harms of interventions, though often fails to provide the practical evidence to inform medical decision-making. A primary reason is failure to recognize the most important questions for treating patients in clinical practice, and using this as the motivation for the design, monitoring, analysis, and reporting of clinical trials. Our most promising opportunity is placing increased interest on questions of a pragmatic origin to match their clinical importance. Improvements to traditional approaches are needed. Standard approaches synthesizing information obtained from separate marginal analysis of each outcome, fail to incorporate associations between outcomes and recognize the cumulative nature of outcomes in individual patients, suffer from competing risk complexities during interpretation of individual outcomes, and since efficacy and safety analyses are often conducted on different populations, generalizability is unclear. The utility of clinical trials can be enhanced by: (i) using outcomes to analyze patients rather than patients to analyze the outcomes, and (ii) incorporating benefit:risk evaluation into standard trial design and conduct, rather than conducted as a post-hoc exercise. Benefit: risk outcomes such as the desirability of outcome ranking (DOOR), can become a standard, prespecified along with efficacy and safety outcomes. Inclusion and analyses of such endpoints provide important information regarding the effects on patients, unable to be obtained by marginal analyses of efficacy and safety. Prespecified procedures to identify subgroups of patients with a positive benefit:risk profile will further advance clinical trial science. [ABSTRACT FROM AUTHOR]

I thank and commend Dr. Gibson on his thoughtful and thorough discussion regarding the role of I p i -values in evaluating evidence (Gibson [8]). One could replace hypothesis testing and I p i -values with confidence interval estimation. Observed I p i -values should be reported rather than, for example, I p i < 0.05 or I p i > 0.05. Although a I p i -value of 0.06 and 0.86 each result in a failure to reject a null hypothesis when testing at 0.05, they may be interpreted quite differently in light of other evidence such as results for other outcomes. [Extracted from the article]