Major depressive disorder treatments remain unsatisfactory, and the development of novel antidepressants is continuing. Therefore, not only the establishment of therapeutic strategies to accumulate evidence on existing therapies, but also the development of novel therapies is required in order to improve the medical standards.
 In principle, parallel, double-blind, randomized, placebo-controlled trials are necessary to assess new compounds for the treatment of major depressive disorders from a scientific perspective. To provide unambiguous evidence of antidepressant activity, well-controlled studies with adequate designs must show efficacy with a statistically significant effect on a clinically meaningful endpoint. For this purpose, it is important to examine all aspects of factors that adversely affect the efficacy and safety assessment in the planning stage of clinical trials and reduce these factors.
 There are several specific characteristics of clinical trials for neuropsychiatric disorders. Some typical features are as follows: 1) a high and variable response, 2) impact on the effect of the baseline severity of disorders, 3) high dropout rates, 4) biases related to subjective measures of clinical symptoms. In this paper, considerations for the planning and performing of clinical trials for major depressive disorders will be discussed based on these features.
 <Author's abstract>