However, EBM requires performing large scale
trials, which require high expenses, prodigious labor, and
sophisticated infrastructures, so they cannot be performed
frequently [1]. Secondly, prediction of the outcome is
required at the time of planning, and optimization of all
the possible factors is not easy. Thirdly, a combinatorial
explosion problem may be encountered when all drugs are
used in a given clinical situation. Fourthly, the different
backgrounds of patients can make results controversial,
even when almost identical large-scale trials are carried
out [1,2]. Lastly and importantly, large-scale trials cannot
produce or even predict the new treatment.