Q（Question）

Regarding principle ②, which states that 'negative information about the efficacy, safety, and quality of prescription drugs, such as failure to demonstrate superiority, must also be provided,' what types of information are expected to be provided?

A（MHLW answer）

For example, when citing data from original research papers, information must be provided accurately without distorting the content of the original paper — which means not excerpting only the portions favorable to the company's product, but also including information such as instances where the company's product failed to demonstrate superiority and information on risks such as adverse reactions.

So what (meaning): When citing from original research, favorable excerpts alone are not sufficient. Data showing the product failed to demonstrate superiority, and information on adverse reactions and other risks, must be included so that the overall message of the original paper is accurately conveyed.

So why (rationale): Selective citation creates a distorted impression that can lead healthcare professionals to make suboptimal prescribing decisions; presenting the full context of original research is a core principle for ensuring appropriate treatment choices for patients.

Commentary — background, application, practical notes

The obligation to provide negative information is explained with more concrete specificity in this Q&A than anywhere else. What the Ministry of Health, Labour and Welfare focused on is the principle of 'not distorting the content of the original research paper.' This prohibits not only the active misconduct of fabricating data but equally the passive misconduct of manipulating the recipient's impression by selecting and emphasizing only convenient data.

A typical scenario is a randomized controlled trial comparing the company's own product against a comparator in which the primary endpoints — such as mortality and hospitalization rates — showed no statistically significant difference between groups, yet the materials present only secondary endpoint data (e.g., symptom improvement scores or changes in laboratory values) where the company's product appeared superior, omitting the primary endpoint results entirely. This selective presentation creates the false impression that the trial as a whole supports the company's product's superiority.

The point most easily misunderstood is the assumption that 'using only data that has been publicly published means there is no problem.' Even when only published data is used, the way it is selected and presented may still distort the context of the original research, and this is equally prohibited. In concrete terms, knowing that a trial exists in which the company's own product 'failed to demonstrate superiority' and then excluding that trial from the information selected for materials is precisely what constitutes 'arbitrary selection of information.'