Hi no-retreat,

You use the wrong statistical tool. The one you used was a chi squared test - it does not make any sense at all to use that test. The correct test is a paired student's t-test. However, to make such a test, you need to know the results from each patient and we do not have those results. I trust the statistician and the p-numbers provided by Resverlogix - your conclusion, fortunately, is wrong.

Best regards,

BKC