[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address Please contact the publisher to request reinstatement.
[Skip to Content Landing]
September 20, 1995

Bayesian Analysis and the GUSTO Trial-Reply

Author Affiliations

The Montreal General Hospital Montreal, Quebec

JAMA. 1995;274(11):874. doi:10.1001/jama.1995.03530110035030

In Reply.  —We agree with Dr Avins that clinical significance is far more relevant to the interpretation of clinical trial results than statistical significance. However, it must be emphasized that our Figure 4 displays the posterior probability density for the difference in mortality rates between t-PA and SK, assuming no prior information is available, so that the area under the curve between two points represents the probability that the rate difference is in that range. This is unrelated to a P value, which only provides the probability of getting data more extreme than that actually observed in the trial, given that the null hypothesis is exactly true.Of course, we also agree with Dr Browne that the misuse and misinterpretation of P values is not the answer to the question of how to best analyze clinical trials. We see little difference between the use of Bayes' theorem in assigning a

First Page Preview View Large
First page PDF preview
First page PDF preview