The P-value is the significance probability of obtaining a value of the test statistic that is as extreme, in relation to the null hypothesis in a particular data set more informative statistical measures such as the likelihood ratio and the Bayesian posterior probability have been suggested for drawing inferences from clinical trials and epidemiologic studies. Causal inference is not statistical in nature; rather it strives to provide scientific explanations of criticisms of proposed explanations that would describe the observed data pattern. In this context, it is important to remember that a finding may not be medically important, or a causal hypothesis may even not be true even if a study shows a significant P-value.