Claims Lacking Evidence or Scientific Basis

Category 02 covers 69 cases documented between 2016 and 2024. The underlying structure is consistent: claims about efficacy, safety, or superiority over competing products are made without adequate supporting evidence. Data from fewer than ten patients, unreviewed poster presentations, personal impressions of company representatives, and commercial news articles all circulated as if they constituted evidence. Recurring logical failures include repurposing non-inferiority trial results as proof of superiority, asserting safety differences based solely on mechanism of action, and dismissing risks flagged in a Risk Management Plan by contrasting them with other products. Healthcare professionals are trained to evaluate information critically, yet they operate under time pressure. Unsupported claims distort their decisions and ultimately harm patients. The PMD Act Article 66 prohibition on exaggerated advertising and the MSA Guidelines (Promotional Activity Guidelines, issued by MHLW) Principle (1) requirement for information grounded in scientific and objective evidence can each be read as a direct response to the failures documented here.

Mio"Yui, you've read about twenty cases in this category. What patterns are you noticing?"

Yui"I was surprised by the volume — 69 cases, running without interruption from 2016 through 2024. The same type of failure keeps appearing."

Mio"Right. Can you break down what the failure actually looks like? 'Lacks evidence' covers a lot of ground."

Yui"Reading through them, I see roughly three patterns. First, cases where no supporting basis exists at all — a representative's personal impression, or hearsay that a well-known physician once endorsed the product. Second, cases where a basis exists but it's unreliable — fewer than ten patients, an unreviewed poster, a commercial news article used as a source. Third, and this one is trickiest, cases where data exists but the logic is wrong — using a non-inferiority trial result as proof of superiority, or asserting a safety advantage based solely on mechanism of action."

Mio"Good breakdown. Why is that third pattern — the misuse of logic — the hardest to catch in review?"

Yui"Because a number appears, and at first glance it looks legitimate?"

Mio"Exactly. When something carries the form of 'trial results,' the reader tends to extend trust before scrutinizing the content. A non-inferiority trial only establishes that the difference falls within a prespecified margin — it does not establish superiority. Reframing that as 'our product is better' in promotional material is not data fabrication, but it is a distortion of fact. That can fall under the PMD Act Article 66 prohibition on false or exaggerated claims."

Yui"And the MSA Guidelines Principle (1) requirement for scientific and objective evidence is precisely what's violated there."

Mio"Yes. Now, from a reviewer's side — what would you actually check to assess reliability?"

Yui"For cited literature: whether it's peer-reviewed, the sample size, the study design, whether the result comes from a primary or secondary endpoint. And conflict-of-interest disclosure."

Mio"All of that. There's also a separate blind spot with verbal explanations — at lectures or during sales visits. Those leave less of a record, which makes it easier for representatives to state things without basis. A large share of the cases in this category involve unsupported claims made verbally at web lectures or in-person visits. Reviewing the promotional material alone won't catch them."

Yui"So even material that passes review can generate a different problem in the actual information exchange. Reviewers need to be clear about the scope of what they're reviewing, and then think through what should be monitored downstream."

Mio"Once you can hold that perspective, these cases stop being a catalogue of past failures and become a way to rethink how the review process itself is designed."

69 real cases from the reports