ABSTRACT
For a new drug that is being introduced, or if an older drug that is being applied to a new indication, regulatory approval requires one or two successful trials to establish clinical benefit given satisfactory safety. If only one trial is conducted, meta-analysis is not possible or necessary. However, if there are two or more trials, the results should be combined to establish an overall estimate of clinical effect. For a new drug that is approved where there are already established therapies, the benefit versus standard of care and all existing therapies are required for reimbursement. The absence of comparative effectiveness leads to a very high probability of rejection of the submission for reimbursement. For the first drug for a disease, a comparison to an existing therapy is not possible.
