Also, see my earlier post (link below) I wrote after digging deep into the basics of p-value, t-value, null-hypothesis etc. My post only conveys that two-tailed p-test DOES make sense and has nothing to do with decline/improvement/harmful/beneficial aspects of a drug.