According to regulation (21 CFR 300.50) drugs may be combined when “each component makes a contribution to the claimed effects” for a given dosage and patient population. Within this framework, drugs may be combined for a variety of reasons. A relatively straightforward case is a combination of drugs with different mechanisms that treat different symptoms. Secondly, another component may be added to the principal active drug in order to discourage abuse. Another potential rationale might be a claim of synergistic efficacy, i.e., more than additive effects. Such synergy allows for a smaller dose of each drug, but that by itself does not improve the patient’s wellbeing. It only has value if there is some clinical advantage such as improved safety or tolerability, i.e., toxicity is not synergistic as well. From a statistical standpoint, rigorously addressing such a claim requires a numerical tradeoff between benefits and harms.