Use and misuse of statistical significance in survival analyses.

Abstract

In most preclinical disease models, survival analyses are the gold standard for measuring the efficacy of medical interventions such as therapeutics or vaccines. In these analyses, treatment regimens that promote the survival and/or reduce the morbidity of experimental subjects (e.g., mice) are tested for efficacy. Although these analyses appear to be relatively straightforward, there are associated caveats regarding interpretation of the results that we wish to discuss in this editorial. Of particular concern is overinterpretation of the biological significance of survival data based on statistical significance rather than durability of protection.