Authors Dichotomize Medical Research Findings as Significant Versus Not Significant, Creating a False Sense of Certainty, and Report Outcomes on Patients Whose Results Have Been Previously Reported Without Proper Disclosure.

Abstract

Statistical significance dichotomizes research findings into significant versus not significant, creating a false sense of certainty. It is insufficient to mindlessly report results as significant versus not significant without providing a quantitative estimate of the uncertainty of the data. Authors could provide a confidence interval, draw a P value function graph, or run a Bayesian analysis. Authors could calculate and report a Surprise or S value. Most importantly, authors could thoughtfully consider how the uncertainty within their research data informs the results of their study. And, clinical databases allow researchers to test multiple hypotheses. This could result in reporting outcomes on the same patient or patients in more than 1 study. Such "double-dipping" is not a dilemma in and of itself, but a problem occurs if multiple reporting of outcomes on the same patient or patients is not disclosed in the methods of a study. Absent clarifying disclosure of multiple reporting, a single patient might then be counted twice in future systematic reviews or meta-analyses, resulting in a biased and incorrect review of the literature. Authors using databases to report clinical outcomes must absolutely and explicitly clarify in their methods if the results of 1 or more patients included in their study have been reported in previous publications.