observational data(One of the key lessons we should learn historically is that making treatment decisions based on observational data is not the way forward. You need a "gold standard" randomised controlled trial, designed to find an answer to a question by recruiting patients in similar circumstances either to take the drug or to take a placebo. Their doctors and the researchers do not know which ones are taking the genuine trial drug.Alex_Odeychuk)
anecdotes(соответствуют уровню доказательности С (по шкале уровней доказательности A, B, C, D) и могут быть полезны для выбора терапии в ситуациях, которые не освещены в рандомизированных исследованиях Alex_Odeychuk)