01What this project has been recording
The stated purpose has not changed in ten years: to detect conduct amounting to advertising violations early, to take necessary action such as administrative guidance, and — by encouraging voluntary efforts by companies and industry associations — to bring promotional activity for ethical pharmaceuticals into proper order. Enforcement is not the sole aim, and that shapes the character of the whole exercise.
The mechanism runs in six steps. MHLW selects monitor institutions from hospitals across Japan. Healthcare professionals there report, through the secretariat (Mitsubishi UFJ Research and Consulting), promotional activity from MRs, MSLs or wholesaler MSs that they judge questionable. An expert case review panel assesses each case. MHLW decides administratively whether it constitutes a legal violation, and where necessary issues administrative guidance in cooperation with prefectural authorities. Industry associations are asked to revise their voluntary codes accordingly.
From FY2019 a general reporting channel was opened to institutions outside the monitor network. Because the base population widened, and because the survey window has ranged from three to nine months depending on the year, year-on-year counts cannot be read directly as "violations went up or down". That caveat is the price of admission for using these numbers.
02Ten years in one table — where the weight moved
| Fiscal year | Survey window | Products with questioned activity | Suspected violation items | Leading category |
|---|---|---|---|---|
| FY2016 | 3 months | 39 | 64 | (categories differ from later years) |
| FY2017 | 5 months | 52 | 67 | Expression liable to cause misunderstanding (41.8%) |
| FY2018 | 8 months | 45 | 74 | Explanation without evidence (14.9%) |
| FY2019 | 8 months | 39 | 57 | Explanation without evidence (24.6%) |
| FY2020 | 8 months | 14 | 17 | Explanation without evidence (29.4%) |
| FY2021 | 9 months | 20 | 26 | Explanation without evidence (38.5%) |
| FY2022 | 9 months | 17 | 23 | Explanation without evidence (39.1%) |
| FY2023 | ─ | 18 | 26 | Explanation without evidence (46.2%) |
| FY2024 | ─ | 18 | 23 | Expression liable to cause misunderstanding (21.7%) |
| FY2025 | ─ | 6 | 11 | Explanation without evidence (36.4%) |
Two things stand out. First, "explanation without evidence" has led almost every year since FY2018, reaching 46.2% of all items in FY2023. Second, counts stepped down from FY2020 and fell to six products and eleven items in FY2025. The FY2020 drop largely reflects the collapse in meeting opportunities during the pandemic; the FY2025 fall is attributed by MHLW to the guideline taking hold.
03FY2016 — the first year
- Hyperphosphataemia agent (MR, oral) ── Using a pamphlet that warned of a side effect specific to the product (excess of a particular nutrient), the MR presented that very side effect as a benefit: "with this drug you can supplement the nutrient in dialysis patients", "you can expect the lab value to rise". A side effect turned into an indication. The institution was in fact running supplementation therapy, and the claim that it would become unnecessary raised concern about prescribing.
- Anti-rheumatic drug (company video on a healthcare professional portal) ── The package insert specifies starting at 300 mg/day, but the video recommended starting at 100 mg/day. A graph favouring the combination arm was emphasised by changing bar colours and making them blink.
- Antipsychotic (in-hospital briefing slides and oral explanation) ── The approved indication is schizophrenia alone, yet the slides added a functional improvement claim and the MR used it to differentiate the product. Other antipsychotics have comparable receptor activity, and the PMDA review report contains no statement that such improvement is expected in humans. The claim did not appear in the product information summary; it existed only in the briefing.
- Bronchodilator (company web seminar figure) ── For the primary endpoint, the vertical axis was partly expanded on the slide to widen the visible gap. The product information summary carried the normal graph; only the slide exaggerated it.
Three of the four first-year cases share one shape: the vetted material is correct, and only the slides, video or spoken words used in the room depart from it. A claim absent from the product information summary appears in a briefing; a normal graph is stretched on a slide. The violation lives not in the material but in the gap between material and delivery. That structure recurs throughout the decade.
The second feature is the reading of a side effect as a benefit. The MR arrives carrying a safety warning and turns its content into a selling point. The existence of the material becomes the entry point for the deviation.
In terms of severity of health harm and egregiousness, no case warranted immediate enforcement. The year's purpose was to make visible that a three-month window already yielded 39 products and 64 suspected items.
Beyond administrative guidance, the project's stated aim included encouraging voluntary efforts by companies and industry associations. The loop of feeding panel assessments back to associations, and asking them to revise voluntary codes, began here.
04FY2017 — the year data manipulation acquired technique
- Constipation agent (oral) ── "It will replace existing agents", "the indication abroad is broader than in Japan", "it can be used widely for any patient complaining of constipation" — repeated hints at indications outside the approval. The same explanation was reported by other monitor institutions, suggesting activity across a wide area.
- Anticancer drug (slides and oral) ── A reference line found neither in the interview form nor the product information summary was added to a PFS curve at the 12-month point, where the gap was widest. In a randomised trial with over 200 patients per arm, only 11 and 5 patients remained at that point. Asked why, the MR said the line marked the longest observation period — but a graph whose longest observation was 21 months also carried a line at 12 months. The explanation contradicted itself.
- Antibacterial (website and MR-supplied materials) ── Although the study was comparative, the comparator results present in the cited paper were removed. The text said non-inferiority and superiority had been judged, but the results of that judgement were not shown.
- Local anaesthetic (oral) ── Prefacing with "there is nothing I can present as data", the MR nonetheless disparaged competitors by hearsay — "it is said to hurt less than other agents", "it is said to be kinder to the skin". Neither the interview form nor the company website supported any of it.
- Diuretic (product information summary) ── Under the headline "all adverse events and serious adverse events did not differ between under-80s and over-80s" sat a table showing roughly a five-point difference in hypernatraemia incidence in the high starting-dose group, with a caution. The caution was on the same page, but a reader who read only the headline would misjudge safety.
- Anti-rheumatic drug (promotion approval application) ── Asked for the basis of "fewer lipid abnormalities than the class effect of similar agents", the MR replied that he had simply compared the adverse-reaction frequencies printed in each package insert.
- Ulcerative colitis agent (briefing) ── "Adverse reactions occurred in about 10% of cases." The package insert records roughly 55% in clinical trials. The Phase III study cited had excluded from evaluation a principal adverse reaction — a fall in a blood concentration — on blinding grounds. The MR volunteered none of this, and did not mention that the related events sat in the RMP as an important potential risk.
This year's cases show that data alteration runs in two directions: removing and adding. Delete a comparator arm that exists in the source (antibacterial); add a reference line that does not (anticancer). Both preserve the appearance of citation while moving the conclusion.
The second pattern is swapping the basis of comparison. Comparing adverse-reaction frequencies across package inserts; presenting as a whole picture a figure from a study that excluded a major adverse reaction. Every number is real — but they are not numbers that may be compared. That is why "expression liable to cause misunderstanding" accounted for 41.8% of items.
During this fiscal year, on 29 September 2017, MHLW revised the Standards for Fair Advertising of Pharmaceuticals and issued the accompanying commentary notice the same day, adding explicit treatment of digital media and organising comparative advertising, endorsements and patient testimonials. Categories made visible by the monitoring project were folded back into the standard.
Extracting or modifying data when citing, and drawing reference lines without a clear scientific reason, are both impermissible. Cite in a way that preserves the structure of the source.
05FY2018 — the year oral explanation became the main arena
- Parkinson's disease agent (oral) ── Although the package insert prescribes a dosing interval from a contraindicated concomitant drug because of serious adverse reactions, the representative said there was no clear evidence for it and that shortening it would not be cut by insurance review. The same explanation appeared to be circulating widely in the area.
- Dyslipidaemia agent (briefing slides) ── For a three-arm study in the review report, efficacy graphs showed only one or two arms. Only the low-dose and placebo arms were presented, hiding the fact that increasing the dose produced little additional effect.
- Acid secretion inhibitor (oral) ── A 5 mg strength not contemplated for maintenance therapy was recommended on the basis that physicians may reduce doses at their discretion. The representative then extrapolated "10 mg twice daily beats 20 mg once daily" into "5 mg twice daily beats 10 mg once daily". No study supported it.
- Antibacterial (oral and hearing materials) ── The origin of the brand name — a word meaning "outstanding" — was presented, and the product described as "a drug with outstanding efficacy". What the trial had demonstrated was non-inferiority to the comparator.
- Anticancer drug (information provision) ── Unprompted, the representative said "the biosimilar of this product was not approved overseas", adding that no details were available. Several monitor institutions reported the same pattern, along with emphasis on extrapolated indications, on "equivalent/comparable but not identical", and the claim that patients already on the originator cannot be switched (switching is not prohibited).
- Glaucoma agent (oral) ── Told that the 14-day prescribing limit on new drugs made adoption difficult, the representative replied that "one bottle lasts a month, so monthly visits work".
- Analgesic (product video on a professional portal) ── The corresponding author of the paper cited for safety was paid as the company's "medical adviser", but no conflict of interest was disclosed in the video.
Channels concentrated on oral explanation at 48.9% — from auditable materials to unauditable speech. "Explanation without evidence" took the top slot for the first time.
Two categories are specific to this year. One is teaching the workaround: it won't be cut by insurance review; the prescribing-day limit can be met with one bottle. Use contrary to the package insert is reframed from a regulatory question into an operational one. The other is coordinated pressure against biosimilars: the same content across multiple institutions cannot be explained as individual deviation.
On 25 September 2018, during this fiscal year, MHLW issued the Guideline on Promotional Activity for Ethical Pharmaceuticals (applicable from 1 April 2019). Rather than drawing a line around "advertising", it takes promotional activity itself as its object, and requires responsibility at management level, prior review of materials by a promotional activity supervisory department, and monitoring and supervision of the field. It was the institutional answer to oral explanation becoming the main arena.
The guideline asks that the appropriateness of materials not rest on individual judgement but be secured by an organisational mechanism: prior review by a supervisory department and periodic monitoring. It is the starting point for treating "the MR said it" as a design problem of the organisation.
06FY2019 — a new name and a wider door
- Analgesic (oral) ── Prefacing with "this is off-label" and "in overseas guidelines", and with no question from the floor, the representative described efficacy in dyspnoea and cough and argued the product could be first line in dyspnoeic patients with renal impairment.
- Antibacterial (briefing slides) ── Results of several trials were pooled. Primary and secondary endpoints sat side by side; studies with different age eligibility and dose-finding studies were mixed in; per-study patient numbers were absent; and non-inferiority studies were shown without comparator arms.
- Renal anaemia agent (oral) ── Rheumatoid arthritis patients had been excluded from the Japanese Phase III study, yet the representative said efficacy could be expected in them.
- Analgesic (pamphlet and oral) ── While noting there had been no head-to-head study, the representative used a difference in dissociation half-life to assert fewer adverse reactions. The review report records comparable incidence of the main adverse events.
- Anticancer drug (information provision) ── With a switch under consideration, the representative reported that "at other institutions the adopted product reverted because of vascular pain and other adverse events", with no detail of any investigation.
- COPD agent (slides and oral) ── "Essentially all COPD patients with comorbid asthma are candidates." In fact the product is not recommended in some patients because of increased pneumonia risk, and no adverse-reaction information was given orally.
- Hypnotic (web lecture) ── A COI slide was shown, but for barely the blink of an eye, too briefly to read.
This was the first year under the guideline. The project was renamed the Promotional Activity Monitoring Project, a general reporting channel was opened to non-monitor institutions, and an awareness video was produced and published.
The common shape is speaking without being asked. Off-label use raised when no one asked; an unfavourable rumour about a competitor delivered unprompted. The guideline leaves room for information provided in response to a healthcare professional's request — and these cases manufacture the request's absence into an opening.
The COI case matters for a different reason. The question moved from "was it disclosed?" to "was it disclosed for long enough to read?"
Reporting channels were widened and outreach to healthcare professionals built into the project — a shift away from relying on what a limited set of monitor institutions happened to observe.
That healthcare professionals receiving questionable information ask, on the spot, for proper information. The reports return to this repeatedly: the interaction itself is a means of correction.
07FY2020 — the year meetings moved online
- Alcohol dependence agent (oral) ── The package insert conditions prescribing on completion of training, yet the representative said "there are cases where prescribing happens as long as the physician has done the training, even if other professions have not" — an explanation readable as permission to ignore the condition.
- Anti-rheumatic drug (oral) ── The trial in the review report switched patients between the product and another agent, but the study design figure in the interview form and product information summary omitted the other agent, implying a switch to monotherapy with the product.
- Anti-rheumatic drug (online group meeting) ── With no trial comparing clinical effect, the representative said the product was "more selective and superior to existing agents in the class".
- Antipsychotic (online meeting) ── Despite a footnote stating that the material did not present a comparison between the product arm and the comparator arm, the representative used each arm's superiority over placebo to claim equivalent efficacy.
- Antibacterial (online group meeting) ── The slides named only active ingredients, but the competitor's brand name was spoken aloud, repeatedly, in comparison.
- Renal anaemia agent (online meeting) ── With no comparative data, the representative said "no adverse reactions were seen in the Japanese trials, so it is safer than other agents".
- Web seminar ── None of the speakers disclosed a conflict of interest. According to the speakers, the sponsoring company had not asked them to.
The channel mix inverted: online individual meetings at 35.7%, face-to-face at 21.4%. With it came a clear split — put nothing in the document, say it aloud (the competitor's name; the footnote neutralised). The document passes review; the audio leaves no trace.
The COI failure at the web seminar was caused not by speakers' omission but by the sponsor not asking. Responsibility sat with the company.
Reports fell to 14 products and 17 items, but this largely reflects the collapse in meeting opportunities and cannot be read as evidence of improvement.
The guideline applies to promotional activity as a whole, regardless of medium. Being online is not an exemption.
For company-sponsored seminars, that the company take responsibility for the appropriateness of the occasion — including asking speakers to disclose conflicts of interest.
08FY2021 — the year disparagement surged
- SGLT2 inhibitor (explanation) ── Unprompted: "there is no indication for HFpEF, but efficacy is reported in the literature. Since chronic kidney disease was added as an indication, physicians say it has become easier to prescribe for HFpEF patients by attaching the CKD diagnosis."
- Dermatitis agent (material shown in an online meeting) ── The study design did not compare the product arm with agent A directly, yet a between-group difference estimate absent from the original paper was added between them. The same appeared in the product information summary and on the company website, implying a head-to-head comparison.
- Renal anaemia agent (explanation) ── With no such finding in the trials, the representative said "the gradual rise in Hb means a lower risk of thromboembolism", implying a lower risk than competing agents.
- Diabetes agent (explanation and slides) ── Where a Japanese subgroup analysis showed no significant difference, the representative said "the difference clearly comes through in the Japanese population too". Challenged, he switched to authority: "Professor X says sufficient efficacy can be expected."
- Migraine preventive (online meeting, supervisor present) ── The shared comparison table itself was unobjectionable, but without being asked the representative disparaged competitors ("agent A requires a double loading dose", "agent B causes constipation"). The supervisor was present, suggesting the practice was organisational.
- Heart failure agent (online briefing) ── Only the overseas Phase III result, which reached significance, was presented; the domestic Phase III result, which did not, was not mentioned at all. The same was confirmed at multiple institutions.
Disparagement of competitors jumped to 19.2%. What deserves attention is the signature of organisation: conduct in front of a supervisor; identical explanations at several institutions. Individual deviation does not account for it.
The second theme is exaggeration by omission. In the heart failure case not one statement is false. The trial that reached significance is described; the one that did not is passed over. From this year on, "efficacy emphasised alone" sits alongside active exaggeration as a principal category.
The SGLT2 case shows an added indication being used as a door into off-label use. Once the conversation turns to attaching a different diagnosis, the boundary of the approval is effectively void.
Adding data absent from the original paper when preparing a graph is "going too far", and presenting a between-group estimate for arms that were not compared directly invites factual misunderstanding.
Where cases suggest promotional activity that is organisationally improper, companies and industry associations were asked to enforce compliance themselves.
09FY2022 — patient-facing materials and disparagement by proxy
- Antipyretic analgesic (patient-facing material) ── A figure made the systemic patch look stronger and broader than oral or local agents: the product coloured deeply out to the extremities with "systemic" set in larger type, while the oral agent was labelled "acts systemically" but drawn in pale colour over a narrow area with no emphasis.
- Mineral preparation (online briefing) ── The electronic package insert permits dilution only in physiological saline, yet, unasked, the representative said "at another site they dissolve it in 50 mL of saline" and "somewhere else they used 5% glucose and had no problem" — administration methods conveyed as hearsay.
- Monoclonal antibody (explanation) ── The indication is limited to pruritus, but the representative said efficacy had been shown for the rash that follows the itch. The basis was a secondary endpoint of the Japanese Phase III study. Providing secondary-endpoint information is not itself a problem; using it to describe an unapproved indication promotes off-label use.
- Diabetes agent (online briefing) ── Showing a graph of the product arm alone, the representative supplied the comparator arm orally: "not shown on the slide, but the other arm showed a concentration-dependent rise in lactate, so a lower risk of lactic acidosis can be expected". Lactic acidosis is set in the RMP as an important potential risk.
- Oncology agent (online meeting) ── Without showing any competitor data, the representative named the competitor and asked, "their trial did not produce data this good — what do you think, doctor?" An attempt to have the healthcare professional supply the disparagement.
- Osteoporosis agent (explanation) ── A conceptual diagram from a review article was presented as evidence of a better anabolic window. The diagram illustrated a way of thinking, not measured clinical effect.
Three features. First, patient-facing material enters the record. Colour depth, coloured area and type size create an impression of superiority without a single incorrect word.
Second, disparagement by proxy: rather than criticising the competitor, invite the healthcare professional to do it. Moving the speaker avoids the formal breach.
Third, a deliberate division of labour between slide and speech (the diabetes agent). Do not write it; say it. The split that appeared in FY2020 has become planned.
Providing secondary-endpoint information is acceptable; using that result to describe an unapproved indication promotes off-label use and is improper. The line was stated explicitly.
Use appropriate materials and data. Do not conflate a conceptual diagram with evidence of clinical effect. Do not let an efficacy narrative cancel a risk recorded in the RMP.
10FY2023 — the year "explanation without evidence" approached half
- Hyperlipidaemia agent (material in an online briefing) ── A survey of preferences for formulation characteristics was presented, but the respondents were severe asthma patients, not candidates for the product, and the survey came from a country with a different health insurance system. The graph shown had also been cut down to the parts favourable to the product.
- Antiviral (mass email direct mail) ── From a drug-interaction comparison graph, the "no interaction" group present in the original paper had been deleted.
- ENT agent (online briefing) ── "Several physicians in this prefecture served on the efficacy and safety evaluation committee for the Phase III study, and they told us it works well and they have high hopes for it." The impressions of trial committee members offered as the basis.
- Antineoplastic (face-to-face) ── "The reason adverse events are more frequent in the Japanese-only trial data than in foreigners is that Japanese physicians are conscientious, so more reports come in." The review report states that caution is needed for those events, suggesting the MR was unaware of it.
- Dermatitis agent (face-to-face) ── "It is used for pruritus in dialysis patients." Asked for the data, the representative answered that there was none.
- Renal anaemia agent (product pamphlet) ── The pamphlet's arc ran: reaching the target haemoglobin early yields better renal outcomes → this product improves anaemia early → please consider it. The cited work was observational, drew no conclusion requiring the target to be met by week 12, and stated that a randomised trial would be needed. With no evidence of faster improvement than same-mechanism agents, "patients in whom you want to improve anaemia quickly" was listed as a candidate group.
- Diabetes agent (face-to-face) ── Asked why so many strengths existed, the representative answered "we anticipate use for obesity", and volunteered that an application was pending. Safety related to weight loss is set as an important potential risk.
"Explanation without evidence" reached 46.2%, the highest share in ten years. What this year's cases reveal is the quality of what gets offered as a basis: the impressions of a committee member, the conscientiousness of Japanese physicians, a foreign survey of the wrong population. Each keeps the form of resting on something. The space is not empty; a substitute has been put in it.
Mass email direct mail appears as a new medium — less amenable to external control than a meeting, and carrying the same old deletion from a source.
The renal anaemia pamphlet shows that misdirection can arise from the arc of a document even when no individual sentence is false. It argues for review at the level of composition, not sentence.
Extracting or modifying data when citing figures from source literature is a guideline violation. Citing the impressions of trial evaluators or authorities as a basis, without scientific or objective grounds, is inappropriate.
In February 2024 MHLW issued Q&A No. 4 on the promotional activity guideline, clarifying what kind of comparative information about competitor products may properly be provided.
The reports also warn about the opposite failure: companies becoming so cautious that they provide nothing beyond what is already in the product information summary. What is asked for is not silence but honest information provision with a basis.
11FY2024 — the company-sponsored lecture as a way around
- Diabetes agent (online group meeting and pamphlet) ── "This is the only treatment shown to improve survival." The basis was time to all-cause death — a secondary endpoint. The pamphlet carried the same claim.
- CNS agent (online group meeting) ── "If treatment continues beyond 18 months, this gap is expected to widen further." No interim results from the long-term study had been published.
- Antineoplastic (face-to-face) ── A medical affairs staff member said the target protein is expressed in almost all solid tumours and that the product would come to be usable in every tumour type, and that both this product and a competitor were slated for expansion so either could be used.
- Peripheral nervous system agent (face-to-face) ── Sleep-disorder incidence was given as "over ten percent" for competitor B and 2.4% for the company's A. In fact B's figure was all adverse reactions and A's was somnolence alone; for somnolence, B was under 1%.
- Metabolic agent (face-to-face) ── For a relaxed paediatric dose limit, the MR conveyed only "it can now be given up to 30 mg", with no background and no safety information. Asked for the basis, he could not answer; two weeks later he returned with nothing but an interview form with a sticky note.
- Vaccine (online group meeting) ── Unprompted, the MR said "B requires reconstitution before use, A does not", extracting only the competitor's disadvantages and none of the company's own, such as a shorter shelf life.
- Metabolic agent (face-to-face) ── Relaying word of mouth that two patients in the prefecture had needed surgery after adverse events on competitor B, the MR said "this is an abnormal situation, so please switch your adopted product to our A".
- Kampo preparation (product briefing material) ── After the transition period for the revised package insert format ended on 31 March 2024, materials carrying the old format were still being distributed.
- Allergy agent (company-sponsored lecture slides) ── The speaking physician repeatedly described short-duration and low-dose long-term regimens outside the approved dosage, said of competitor B that "A and B are exactly the same, so switching is safe", and closed on a slide reading "A is safe, simple, and a big patient benefit". The materials were distributed without prior review by the promotional activity supervisory department, and after discovery no correction was issued to all attendees.
The decisive development is the company-sponsored lecture as a new channel. The speaker is an outside physician, not an employee. But vetting the materials is the company's responsibility, and here they were distributed unvetted, with inadequate correction afterwards. The report names the causes: failure to confirm the speaker's understanding of the guideline, and the absence of layered internal review. Going beyond case description into root cause is what distinguishes this year.
A new category also appeared: expressions that play on healthcare professionals' anxiety. A single phrase — "abnormal situation" — used to request a switch. Not persuasion by data, not criticism of a competitor's weakness, but an appeal to feeling.
The peripheral nervous system case is the finished form of shifting the denominator. Both figures are real and both come from package inserts. One is a total; the other is one event. Misleading without lying.
On the lecture case the report goes further than usual: the sequence of conduct "strongly suggests an absence of corporate compliance awareness". On RMP information, it notes that Section 3 of the guideline requires, as a precondition for promotional activity, the proper provision of information necessary for appropriate use — including contraindications and the RMP — and that cases were nonetheless seen where RMP information was inadequate. MHLW states that it takes seriously the fact that the guideline's principles are not being observed and "strongly requests" that they be enforced afresh.
It also records that, for products in strong competition, cases were seen suggesting organisationally improper promotional activity at some companies.
- Prevention of recurrence: internal systems, assurance of material appropriateness, training and monitoring, written procedures, and rapid response to improper activity.
- MR education. Beyond guideline awareness, basic statistical literacy, so that efficacy, safety and statistical results can be explained as a set. The report notes cases ending with "I was wrong, I'm sorry" — some of which suggest deliberately improper information provision on the assumption that an apology suffices. Those are treated as highly egregious and watched closely.
- For company-sponsored seminars, that speaking physicians and pharmacists also understand the guideline, and that companies check slide content.
- That companies not become inhibited. Information genuinely sought by healthcare professionals should be answered in good faith.
12FY2025 — year ten: fewer reports, and the lecture hall in focus
- Vaccine (wholesaler MS, oral) ── Delivering a quotation, a wholesaler's MS promoted the vaccine as one that "confers lifelong immunity, so a single dose is enough". No long-term follow-up data support it.
- Ophthalmic agent (oral) ── After introducing an overseas Phase II study with few patients and no statistical testing, the representative added: "the professor who was involved in the trial says the product's action may have suppressed the night-time change" — an eminent academic's impression as the basis for efficacy.
- Respiratory agent (company-sponsored lecture) ── The speaking physician, departing from the stepwise treatment selection recommended in the professional society's guideline, presented content amounting to applying an option intended for severe patients to untreated patients, with memorable phrasing and insufficient description of patient criteria. The company had vetted the slides; they were not corrected, and no correction came from the chair. Several physicians gave similar lectures more than once.
- Psychotropic agent (in-hospital briefing) ── "It is the only drug for the type-○○ condition and works well even where other companies' drugs fail." Asked whether a confirmed diagnosis was required, the MR replied that it could be used in undiagnosed patients and that efficacy could be expected in delirium in other psychiatric conditions. No evidence exists outside the indication. A hospital MR, a specialty MR and a medical affairs staff member were all present, and none recognised the impropriety.
- Respiratory agent and vaccine (company-sponsored lectures) ── Numerous instances of speakers referring to efficacy outside the approved indication — antiviral effect for a product with no viral indication; reduced dementia and cardiovascular event risk for a vaccine. Similar cases were found across several physicians and several lectures, totalling a considerable number.
- Metabolic agent (company-sponsored lecture) ── For a product whose use in untreated patients is approved in Europe and the US but not Japan, overseas guideline recommendations were presented, conveying unapproved content. The company had asked the speaker to include that section and supplied the slides. Listing the Japanese indication alongside does not make it acceptable.
- CNS agent (online group meeting) ── "In the European guideline, ours is the only agent of its class listed." Asked why, the MR answered immediately: "because no other agent of the class is marketed in Europe." Had the pharmacist not asked, only the impression of superiority would have remained.
- Ophthalmic agent (online individual meeting) ── Asked for safety data, the MR prefaced with "we are not allowed to explain anything other than the primary endpoint" and then said incidence could not be compared. This suggests a misunderstanding of the guideline.
Reports fell sharply in both channels, with disparagement of competitors down in particular. What remains concentrates in three places.
First, unapproved indications by way of the lecture hall. In one case the company supplied the slides and asked the speaker to include the overseas guideline section. The speaker's status as an outside physician is no longer a shield.
Second, a wider cast. A wholesaler's MS appears in the record. Panel members noted institutions where MR visits fell and information from MSs rose.
Third, what MHLW named "theatrical" promotion: in an online briefing, after the MR's explanation, a medical affairs staff member who had been off camera turns the camera on and answers "at the healthcare professional's request", while the materials in fact run as one continuous set. The form of the guideline is met precisely in order to work around it.
The ophthalmic "we cannot explain anything but the primary endpoint" is inhibition and misunderstanding fused. Safety information withheld in the name of the guideline is the outcome furthest from its purpose.
On the fall in reports, the report concludes that the guideline's diffusion appears to have advanced the proper conduct of promotional activity. The panel also recorded concrete positives: providing risks unprompted; explaining one's own product's drawbacks; separating MR and medical affairs activity clearly; a supervisor stopping a subordinate from providing off-label information.
On lectures, however, reference to efficacy outside the approved range is stated to breach not only the guideline but Article 68 of the Pharmaceuticals and Medical Devices Act (prohibition of advertising unapproved products) directly. And because a company that sponsors and commissions a lecture is normally involved to a considerable degree, including prior review, such lectures are in principle treated as the company's own promotional activity and advertising, and may constitute a violation by the company. Where the approved range differs between Japan and abroad, or where results of large overseas trials are introduced, the likelihood of violation is high.
- Enforcement of the guideline's principles, including the RMP. Explaining only efficacy because time is short, and never reaching safety, is called "dishonest and improper" — especially for products where risk has been flagged and careful administration is required.
- Recognition of corporate responsibility for lectures. Compliance obligations extend beyond the company's own MRs and MSLs to lectures commissioned from outside physicians, and demand greater care than before.
- Outreach to healthcare professionals. Awareness of the guideline and of the general reporting channel is low. Training sessions through associations and academic societies, concrete examples of improper conduct with model responses, and dissemination from prefectures to institutions are all sought. Asking a question on the spot when something seems wrong — and building an environment where that happens — is treated as a means of correction in itself.
- Care with email and press releases. The guideline applies regardless of medium. A press release is not itself advertising, but using one in sales activity may constitute promotional activity.
- Education that does not stop at technique. Internal training risks confining itself to the technical. Representatives should understand the guideline's purpose and provide honest information including what is unfavourable to their own product. Such conduct, the report concludes, builds trust in the company — and the project itself should begin collecting and publishing examples of it.
13What ten years show
Set the years side by side and the forms of violation change while several structures hold.
First, the movement toward wherever review does not reach. Once product information summaries were correct, the deviation moved to briefing slides; once slides were reviewed, to oral explanation; once face-to-face declined, to speech online; then to email and the company-sponsored lecture. The decade is a record of stepping outside each net as it is cast. "Theatrical" promotion, named in FY2025, is the terminus so far.
Second, selection rather than fabrication. The largest category over ten years is "explanation without evidence", followed by "efficacy emphasised alone". Cases of invented numbers are almost absent. Which real data is shown and which is withheld moves the conclusion. That is why review must operate on composition, not wording — a point the FY2023 renal anaemia pamphlet makes more plainly than any other case.
Third, individual deviation and organisational design keep appearing together. Disparagement in front of a supervisor (FY2021); identical explanations across institutions (FY2018, FY2021); a lecture whose slides the company supplied and whose content the company requested (FY2025). MHLW reaches for the word "organisational" when it sees these signatures. The same reports also record a supervisor stopping a subordinate from offering off-label information. Organisations amplify deviation and they suppress it.
Fourth, inhibition is also treated as a failure. The reports return repeatedly to the risk of companies becoming too cautious to say anything. What is asked for is not silence but grounded honesty. The FY2025 misunderstanding — "we cannot explain anything but the primary endpoint" — shows that when a rule is understood without its purpose, the rule can end up depriving healthcare professionals of information.
Read from the perspective of material review, ten years reduce to three questions. Does this document preserve the structure of what it cites? Do favourable and unfavourable information carry the same weight? And how will this reviewed document actually be spoken? The third has no answer inside the document. That is precisely why it has remained the most common category of violation for ten years.
Advertising regulation for pharmaceuticals (MHLW, in Japanese) ── the "summary" and "report" PDFs for each fiscal year are available from this page.
Summaries and reports for FY2016, FY2017, FY2018, FY2019, FY2020, FY2021, FY2022, FY2023, FY2024 and FY2025. Secretariat: Mitsubishi UFJ Research and Consulting Co., Ltd. (commissioned by the Compliance and Narcotics Division, Pharmaceutical Safety Bureau, MHLW).