All 33 cases in this volume share one surface feature: data is present. Graphs, literature citations, and comparative trial results are all real. The problem lies in how they are displayed. Axis scales are stretched, control-arm data are removed, primary endpoints are swapped for secondary ones, domestic trial results disappear while a favorable overseas study remains. None of these actions would be called falsification outright, but the impression conveyed to the reader diverges sharply from the source document. The auditor's question must shift from 'Is there data?' to 'What is missing compared with the original publication?'

Mio"When you look at a graph in a promotional resource, what do you check first?"

Yui"Whether the numbers are accurate. But in this volume's cases, the numbers are correct and the resource is still flagged..."

Mio"That's exactly the starting point. Stretching the vertical-axis scale makes an effect look larger without changing a single value. Compare the resource graph side-by-side with the source figure and you'll see it immediately. PMD Act Article 66 prohibits exaggeration—that includes visual manipulation, not just fabricated numbers."

Yui"What about cases where the control-arm data were removed? If the company says they couldn't get permission to reprint the competitor's data?"

Mio"No permission means don't use that comparative figure at all. Leaving only your product's bar while keeping the comparative layout lets the reader infer superiority that was never demonstrated. If data can't be shown, remove the whole figure or find a different presentation method."

Yui"I saw a resource where a trial's primary endpoint showed no significant difference, but the material used a secondary endpoint and stated 'efficacy was demonstrated'—without mentioning it was secondary. How would you raise that?"

Mio"Confirm first, then escalate. Ask: 'Is this endpoint primary or secondary?' Once they acknowledge it's secondary, follow up: 'Where is the primary endpoint result shown?' When the representative has to explain what's missing, the gap becomes undeniable."

Yui"There was also a case where data from the open-label phase were removed so that only the blinded phase remained. Why is that a problem?"

Mio"Both arms tend to show improvement during open-label periods. Strip that out and the gap between arms artificially widens. The processed dataset describes a different trial design than the one the authors published. Cutting numbers without cutting the design context produces a picture the original investigators never intended."

Yui"For cases where domestic trial results were omitted and only favorable overseas data were shown—where's the lever?"

Mio"Ask: 'Do domestic clinical trial data exist?' If yes: 'Why are they absent from this resource?' If domestic data are unfavorable to the product and were left out, the resource fails the Principle (1) requirement to provide a comprehensive summary of the latest scientific evidence. One-sided selection is the violation."

Yui"Do you think the representatives acted with deliberate intent? Some of them might have just thought they were making the graph cleaner."

Mio"Intent doesn't determine the outcome. The standard is whether the reader's impression diverges from the source. Our job as auditors is to document the gap between the resource and the original, not to speculate about motive. Gap exists—send it back. That's the whole role."

33 real cases from the reports

03-01FY2016気管支拡張剤製薬企業が主催する Web セミナーの図表
What happenedIn a web seminar hosted by Company E, a slide was used in which the vertical axis of the primary endpoint graph (disease exacerbation count) was partially enlarged, creating a visually different impression from the standard graph in the product information document. Rapid screen transitions further increased the risk that viewers could not accurately read the data.
MHLW viewThe efficacy and effects of the drug are presented in an exaggerated manner through manipulated graphs that are likely to cause misunderstanding.
Competency lostRisk Detection, Knowledge
Next movePlace the slide graph and the product information graph side by side, verify axis scales, identify any enlargement or cropping, and request correction to match the original publication.
03-02FY2016抗ヒスタミン薬製品情報概要
What happenedCompany F's product information document contained a graph showing only the 24-hour post-dose data point from a double-blind study that reported results at three time points (24 hours, 1 week, and 6 weeks), making it impossible to assess the drug's full efficacy profile.
MHLW viewBy extracting and manipulating only a portion of a figure, the efficacy and effects of the drug are presented in an exaggerated manner. (The 'Guidelines for Preparing Pharmaceutical Product Information Summaries' issued by the Japan Pharmaceutical Manufacturers Association [JPMA] state that 'when citing data from original articles, the content must be presented accurately; data must not be selectively extracted to favor the company's own product, and care must be taken not to distort the intent of the original article, with the source clearly indicated.')
Competency lostRisk Detection, Knowledge
Next moveCompare all three time-point data from the cited publication against the printed graph, and request justification for the selective extraction and addition of the omitted time points.
03-03FY2016抗生物質製剤パンフレット（2016 年 8 月作成）
What happenedCompany G's pamphlet used the claim 'well-balanced and superior antibacterial activity' while omitting data on other oral antibiotics and all intravenous agents that appeared in the original study, without disclosing that the data were selectively extracted, suggesting an intent to steer prescribing toward the company's own product.
MHLW viewBy extracting and manipulating only a portion of a figure, the efficacy and effects of the drug are presented in an exaggerated manner. [Cross-reference: case involving recommendation of an initial dose outside the approved dosage and emphasis in data comparisons]
Competency lostRisk Detection, Intelligence
Next moveCross-check all agents and all routes of administration from the original study against the pamphlet content, obtain a written explanation for omissions, and require either disclosure of selective extraction or inclusion of all relevant data.
03-04FY2017抗菌薬MR 提供資料及びホームページの臨床試験結果紹介
What happenedOn Company D's website and in MR-distributed materials, control-arm data were removed from all comparative trial graphs, and the non-inferiority/superiority judgment results based on 95% confidence intervals—described as the basis for the primary endpoint evaluation—were not displayed.
MHLW viewBy extracting only a portion of the data, the efficacy and effects of the drug are presented in an exaggerated manner.
Competency lostRisk Detection, Knowledge
Next moveSystematically compare the cited publication with every graph in the materials, compile a list of deleted control-arm data and missing statistical outcomes, and require correction and addition across all items.
03-05FY2017糖尿病治療薬簡易版製品パンフレット
What happenedCompany E's diabetes drug pamphlet presented only the 100 mg arm from a three-arm non-inferiority trial (comparator, 100 mg, 300 mg) under the heading 'superior glycemic lowering effect,' while the actual difference from the comparator at the primary endpoint (week 52) was only 0.01%.
MHLW viewBy extracting only a portion of the data, the efficacy and effects of the drug are presented in an exaggerated manner. [Cross-reference: case involving selective extraction of favorable data and emphasis on pharmacodynamic evaluation that was not the original primary endpoint]
Competency lostRisk Detection, Knowledge
Next moveVerify all three arms and the non-inferiority design against the pamphlet heading, then require addition of comparator and 300 mg arm data and correction of the title.
03-06FY2017プロトンポンプ阻害薬パンフレット（2018 年 2 月作成）
What happenedCompany F's pamphlet showed only baseline and day-1 data from a gastric pH study while omitting day-7 results, displaying only the time point most favorable to the company's product. A title describing time-to-pH thresholds on day 1 was added despite the study's actual primary endpoint being 24-hour holding time ratios.
MHLW viewBy extracting only a portion of the data and attaching a title inconsistent with the original purpose of the cited article, the efficacy and effects of the drug are presented in an exaggerated manner. [Cross-reference: case involving manipulation of data to include only the blinded period, excluding the open-label period, thereby exaggerating the difference from the control group]
Competency lostKnowledge, Risk Detection
Next moveCompare all time points and the primary endpoint from the original study with the pamphlet's content and title, then require addition of day-7 data and revision of the title to accurately reflect the study's primary endpoint.
03-07FY2017抗がん剤製品情報概要
What happenedCompany G's product information document published a graph showing the placebo response rate as 1% by excluding 12 cases that responded during the open-label period from the numerator, whereas the cited publication reported 13.0%, causing the apparent efficacy gap between the company's drug and placebo to appear far larger than the actual data warranted.
MHLW viewBy modifying the data, the efficacy and effects of the drug are presented in an exaggerated manner.
Competency lostKnowledge, Risk Detection
Next moveCross-reference the response rate data in the interview form and cited publication with the figures in the product information document, and obtain a written explanation of the numerator and denominator adjustments before requiring correction.
03-08FY2017抗がん剤MR によるプレゼンテーション（スライド・口頭説明）
What happenedA PFS graph in Company H MR's presentation to the pharmacy department contained a reference line at the 12-month time point—present in neither the interview form nor the product information document—where only 11 and 5 cases remained observable; when asked for a scientific rationale, the MR's explanation was internally contradictory.
MHLW viewBy adding a reference line and accompanying explanatory text, the efficacy and effects of the drug are presented in an exaggerated manner.
Competency lostSixth Sense, Risk Detection
Next moveRequest a written scientific rationale for the reference line (pre-specified criteria or relationship to the median follow-up), document it as a discrepancy from approved materials, and report to the review team.
03-09FY2017糖尿病治療薬パンフレット（2017 年 5 月作成）
What happenedCompany I's pamphlet added color-coding on the time axis and a reference line at 160 mg/dL to a postprandial blood glucose graph, neither of which appeared in the original publication, resulting in a visually exaggerated impression of the difference from competing products.
MHLW viewBy adding a reference line and applying color coding, the efficacy and effects of the drug are presented in an exaggerated manner.
Competency lostRisk Detection, Knowledge
Next moveOverlay the original and pamphlet graphs to identify all added coloring and reference lines, then require either a documented scientific basis for each addition or their removal.
03-10FY2017多発性硬化症治療薬パンフレット
What happenedCompany J's pamphlet changed the vertical axis of an efficacy graph from the original 0–100% to 0–50%, and added baseline data absent from the cited publication to a separate graph, both changes making between-group differences harder to assess visually.
MHLW viewBy altering the scale of the axis, the efficacy and effects of the drug are presented in an exaggerated manner, and the original purpose of the cited article is made difficult to understand.
Competency lostRisk Detection, Knowledge
Next moveCompare axis scales and added data points against the cited publication, compile a list of all discrepancies, and require restoration of the original scale and written justification for any added data.
03-11FY2018脂質異常症治療薬プレゼンテーション用スライド
What happenedMultiple graphs in hospital presentation slides showed only the low-dose and placebo arms of a three-arm trial, omitting the high-dose arm and thereby concealing the regulatory assessment finding that increasing the dose produced little additional effect.
MHLW viewIn a three-arm comparative trial, graphs were created by extracting results from only one or two of the three arms.
Competency lostRisk Detection, Intelligence
Next moveCross-reference all three arms from the regulatory review report against the slides, identify graphs lacking the high-dose arm, and require inclusion of complete three-arm data.
03-12FY2018乾癬治療薬新薬ヒアリング用資料
What happenedA safety comparison document submitted at a new drug hearing translated most adverse event data from the original publication, but deliberately omitted the 'malignant tumor' category—the only serious adverse event that occurred in the company's drug arm but not in the comparator arm.
MHLW viewRegarding safety, only information that was particularly serious and unfavorable to the drug in question was omitted.
Competency lostRisk Detection, Sixth Sense
Next moveSystematically compare all adverse event categories in the original publication against the submitted document, request disclosure of the malignant tumor data, and report the potential intentional omission to the review committee.
03-13FY2018糖尿病治療薬医療関係者向け情報サイト上の製品紹介動画
What happenedIn a product introduction video on a healthcare professional website, the legend for a graph showing blood glucose trends in the drug and placebo groups was displayed in reverse. The graph's aspect ratio had also been altered to make fluctuations appear smaller, creating a misleadingly favorable impression of glycemic control.
MHLW viewThe graph was cited inaccurately, with the legend displayed in reverse, and the aspect ratio was altered.
Competency lostKnowledge, Risk Detection
Next moveCompare the graph's legend, aspect ratio, and data values against the original publication and review report item by item; document discrepancies and request correction from the company.
03-14FY2018抗ウイルス薬製品紹介パンフレット
What happenedIn a product brochure for an antiviral drug, the primary endpoint (duration of illness, with no significant difference versus the comparator) was described in only a few lines of text, while secondary endpoints related to viral titer were presented across two pages with graphs emphasizing the drug's superiority.
MHLW viewOnly the results of secondary endpoints that were favorable to the product were presented in detail.
Competency lostRisk Detection, Knowledge
Next moveVerify the significance of the primary endpoint in the review report, then examine whether the space and detail allocated to primary versus secondary endpoints in the brochure are disproportionate.
03-15FY2018抗がん剤製品紹介パンフレット
What happenedA Phase I trial result in an oncology drug brochure showed a markedly lower adverse event rate than a competing product, but the review report noted the use of premedication that was absent from the brochure. Additionally, the trial objective of "comparative safety evaluation" had been removed, leaving only "verification of pharmacokinetic equivalence."
MHLW viewInformation on pre-medications necessary for evaluating safety was not included.
Competency lostRisk Detection, Knowledge
Next moveCross-check the brochure against the review report for premedication use and stated trial objectives; flag omissions that affect safety interpretation and request that the company add the missing information.
03-16FY2018気管支喘息治療薬プレゼンテーション用スライド、製品紹介パンフレット
What happenedAt an in-hospital product presentation for a bronchial asthma treatment, the primary endpoint (annual asthma exacerbation rate) from the overall analysis (approximately 250 subjects per arm) was not shown; only a Japanese subgroup analysis (approximately 15 subjects per arm) was presented. Secondary endpoints were reported using the full population, making the primary endpoint the sole exception.
MHLW viewFor the primary endpoint, only the subgroup analysis results based on a small number of cases were presented.
Competency lostRisk Detection, Knowledge
Next moveRequest that the company present the full-population analysis for the primary endpoint, and verify the statistical reliability and rationale for using only the subgroup analysis.
03-17FY2018抗がん剤プレゼンテーション用スライド
What happenedAt an in-hospital meeting, a comparison of overall and Japanese-patient response rates was used to claim that the drug is more effective in Japanese patients, but the review report revealed that the prior treatment regimens (proportions of Agent A and Agent B) differed substantially between the two groups—a key confounding factor that was not disclosed.
MHLW viewThe difference in baseline conditions between the two groups was not disclosed, yet the product was described as 'more effective in Japanese patients' compared with the overall population.
Competency lostRisk Detection, Communication
Next moveConfirm the distribution of prior treatment regimens in the review report and, citing the imbalance in baseline conditions, request that the company withdraw the claim of superior efficacy in Japanese patients.
03-18FY2018糖尿病治療薬医療関係者向け情報サイト上の座談会記事
What happenedA roundtable article on a healthcare professional website listed adverse events from a Phase III trial, but several items documented in the review report as frequently observed—including ketoacidosis-related events—had been omitted from the list.
MHLW viewAdverse events that were frequently observed in clinical trials were omitted from the adverse event summary table.
Competency lostRisk Detection, Knowledge
Next moveCompare the adverse event list in the article against the review report, identify omitted items, and request that the company include them as required safety information.
03-19FY2018抗リウマチ薬製品紹介パンフレット
What happenedA brochure for an antirheumatic drug contained multiple deficiencies in citations from the original publication: individual serious adverse event entries were omitted or merged, certain numerical values did not match the source, and the exclusion of patients with unknown responses from the patient satisfaction survey was not disclosed.
MHLW viewCitations from the original publication contained inaccuracies, and certain information was deleted or consolidated inappropriately.
Competency lostKnowledge, Risk Detection
Next moveConduct a one-to-one comparison of figures, adverse event descriptions, and analysis conditions between the brochure and the original publication; enumerate all discrepancies and request corrections from the company.
03-20FY2018血友病治療薬プレゼンテーション用スライド
What happenedA hemophilia drug presentation placed alongside the overall non-bleeding rate a separate figure calculated using only patients who completed the weekly dosing regimen, excluding those who discontinued. This denominator-restricted figure was absent from the approved product information and the review report, creating a misleading impression by omitting dropout bias.
MHLW viewAn analysis result that was arbitrary and capable of creating a false impression of superiority was presented.
Competency lostRisk Detection, Knowledge
Next moveVerify the denominator definition of the presented data against approved product information and the review report; confirm whether dropouts were excluded, then flag the use of undocumented data to the company.
03-21FY2018抗精神病薬企業担当者による口頭説明
What happenedWhen asked at an in-hospital session about QT prolongation concerns, the company representative stated that no QT prolongation was observed in an overseas trial at 12 mg/day. The review report, however, documented QT prolongation in a domestic trial at a lower dose.
MHLW viewResults from domestic trials that were unfavorable to the product were not disclosed; only overseas trial results were cited to emphasize safety.
Competency lostRisk Detection, Knowledge
Next moveReview QT prolongation findings from the domestic trial (dose levels and incidence) in the review report, then request that the company representative disclose the domestic data and verify that complete safety information is being provided.
03-22FY2018抗がん剤企業担当者による口頭説明
What happenedDuring a new drug hearing, despite the appropriate use guide explicitly stating that superiority had not been confirmed and that overall survival was shortened in hematopoietic stem cell transplant patients compared with controls, the company representative selectively emphasized the portions of the Kaplan-Meier curve where the drug group exceeded the control group to argue for superiority.
MHLW viewDespite the absence of confirmed superiority for the product, only the portions of the Kaplan-Meier curve where the overall survival rate exceeded that of the comparator were highlighted to assert superiority.
Competency lostRisk Detection, Knowledge
Next moveHave the appropriate use guide at hand when reviewing the full Kaplan-Meier curve during the hearing; if explanations are restricted to a subset of the time axis, request that the company address the complete time course.
03-23FY2018糖尿病治療薬製品紹介パンフレット
What happenedIn a diabetes drug brochure, the vertical axis of the graph presenting the primary endpoint result had been changed to absolute HbA1c values. The primary endpoint in the original publication and review report was defined as the change in HbA1c, making the substitution of absolute values an inaccurate citation.
MHLW viewThe primary endpoint, which was expressed as change from baseline, was converted to absolute values when creating the graph.
Competency lostKnowledge, Risk Detection
Next moveCompare the vertical axis definition of the brochure graph against the original publication and review report; confirm that the switch from change scores to absolute values constitutes a misrepresentation of the primary endpoint and request correction.
03-24FY2018鎮痛薬製品紹介パンフレット
What happenedA pain relief drug brochure cited a graph showing no change in blood pressure for the study drug, but the original publication also included a graph for the comparator drug, which likewise showed no significant blood pressure increase. By omitting the comparator graph, the brochure created the impression that only the study drug carried no risk of blood pressure elevation.
MHLW viewBy presenting only the graph for the investigational product without showing the comparator's graph, the impression was created that only the investigational product carried no risk of blood pressure elevation.
Competency lostRisk Detection, Knowledge
Next moveVerify the comparator blood pressure graph in the original publication and, on the grounds that omitting it distorts the comparative context, request that the company include the comparator data alongside the study drug graph.
03-25FY2019抗菌薬企業担当者による説明資料
What happenedAn MR's product presentation slides merged results from multiple clinical trials with differing eligibility criteria and objectives, listed primary and secondary endpoints without distinction, and omitted control-arm data from non-inferiority trials, creating an overall impression of superiority. Safety information also lacked any RMP explanation and was insufficient regarding hepatic-impairment patients.
MHLW viewData was excerpted, processed, and presented in a manner that created the impression that the product is superior to other drugs, including by combining the results of multiple clinical trials into a single aggregated figure.
Competency lostKnowledge, Risk Detection
Next moveRequest that the material separately tabulate each trial's eligibility criteria, objectives, and design, and verify in writing the basis for any aggregated figures. Also ask for supplementary information on the RMP and safety in hepatic-impairment patients.
03-26FY2019利尿剤企業担当者による説明用資料
What happenedA diuretic brochure set the adverse-event reporting threshold at 1.0% or above—higher than the source paper's 0.5% cutoff—causing three adverse events listed in the original article to be omitted from the table.
MHLW viewBy showing only adverse reactions occurring at or above a certain frequency, the excerpt of data from the original publication created the appearance that no other adverse reactions existed.
Competency lostRisk Detection, Knowledge
Next moveCompare the adverse-event thresholds between the source article and the material in a side-by-side table; if thresholds differ, ask for the rationale and a complete list of omitted events.
03-27FY2020抗リウマチ薬対面の面談にて、企業担当者による口頭説明
What happenedA clinical trial for an anti-rheumatic drug involved switching between the product and a comparator, but the trial design diagram used in the interview form and product summary omitted any reference to the comparator, creating a misleading impression of a monotherapy switch.
MHLW viewAlthough the study involved concomitant use of another drug, figures and tables were manipulated in a way that could mislead readers into believing no concomitant drug was used.
Competency lostRisk Detection, Knowledge
Next moveCross-check the trial design diagram against the regulatory review report and request replacement with a figure that explicitly shows the comparator treatment arm.
03-28FY2021関節機能改善薬オンライン製品説明会時に企業担当者から示されたスライド・口頭説明
What happenedDuring an online product presentation for a joint-function drug, the secondary-endpoint graph was displayed larger than the primary-endpoint graph, and the presenter explained efficacy using the secondary endpoint without disclosing its secondary status.
MHLW viewIn the slides used during the presentation, a secondary endpoint graph was displayed larger than the primary endpoint graph on a single slide. Primary endpoint results are the most important and reliable, as they satisfy the various design requirements for verification, and explanations should center on the primary endpoint. Displaying the secondary endpoint more prominently than the primary endpoint in promotional materials is inappropriate, as it may cause misunderstanding. Furthermore, when presenting and explaining secondary endpoint results, it is necessary to clearly state that the data are from a secondary endpoint and are for reference purposes only.
Competency lostKnowledge, Communication
Next moveRequest that slides be revised to give greater visual prominence to the primary endpoint, and that the presenter explicitly state each endpoint's status at the start of the verbal explanation.
03-29FY2021皮膚炎用薬オンライン面談時に企業担当者が提示した資料
What happenedAlthough the trial for a dermatitis drug did not include a direct comparison between the product group and Group A, the presented graph included an inter-group difference estimate between the two—absent from the original publication—implying superiority through a comparison that was never conducted.
MHLW viewData of questionable reliability was found in the graph presenting the primary endpoint results of the clinical trial. Despite the absence of a direct comparison between the product group and Group A, an between-group difference estimate not present in the original publication was added to the same graph, creating the impression that the product is superior to Group A in a direct comparison. Adding numerical values to a graph that do not appear in the original publication constitutes overstatement, and care must be taken when creating figures.
Competency lostRisk Detection, Knowledge
Next moveVerify the source of the inter-group difference estimate added to the graph; if it is absent from the original article, request a revised version with that figure removed.
03-30FY2022糖尿病・慢性心不全治療剤企業の WEB 製品説明会
What happenedA web-based product presentation for a diabetes and chronic heart failure drug used a table labeling the company's product as 'standard dose' and a competitor's as 'high dose'—based on diabetes dosing conventions—despite both products being approved only at 10 mg for the heart failure indication.
MHLW viewFigures and tables that could cause a factual misunderstanding were presented, describing the company's own product as being at the "standard dose" and a competitor's product as being at a "high dose."
Competency lostRisk Detection, Intelligence
Next moveVerify approved doses per indication against the package inserts and request replacement with a dose comparison table scoped strictly to the heart failure indication.
03-31FY2023高脂血症治療薬製薬企業担当者（オンライン）
What happenedDuring an online presentation for a hyperlipidemia drug, survey data from a study of severe asthma patients overseas was used to support claims about dosing preference advantages, despite the target disease and healthcare system differing substantially from the drug's actual patient population.
MHLW viewInformation was provided based on materials created by selectively excerpting portions of a graph from a foreign publication that were favorable to the company's own product, despite differences in healthcare systems and patient characteristics.
Competency lostKnowledge, Risk Detection
Next moveCheck the referenced study's target disease, country, and healthcare system; if these do not match the approved indication, request replacement with evidence specific to the actual patient population.
03-32FY2023高脂血症治療薬製薬企業担当者（オンライン）
What happenedThe graph shown at an online presentation for a hyperlipidemia drug extracted only 2 of the 5 items from the source document—selecting specifically those that portrayed the product favorably compared to conventional treatments.
MHLW viewExcerpting, modifying, or integrating data when citing figures and tables from original publications or other sources constitutes a violation of the MSA Guidelines.
Competency lostRisk Detection, Knowledge
Next moveCompare all five items in the source graph against the two shown in the material and request written justification for the selection. If none is provided, ask for a version showing all items.
03-33FY2023抗ウイルス剤電子メールによる DM
What happenedA mass email promoting an antiviral drug included an interaction comparison graph from which the 'non-interaction' segment—patients for whom neither drug showed a difference—had been removed, contrary to the original publication.
MHLW viewExcerpting, modifying, or integrating data when citing figures and tables from original publications or other sources constitutes a violation of the MSA Guidelines.
Competency lostRisk Detection, Knowledge
Next moveDirectly compare the distributed graph against the original publication figure, confirm the deleted segment, and request replacement with material that accurately reflects the source.

The Anatomy of Failure ── All 8 categories

  1. 01. Promotion of Unapproved or Off-Label Indications and Dosage (33 cases)
  2. 02. Claims Lacking Evidence or Scientific Basis (69 cases)
  3. 03. Cherry-Picking, Data Manipulation, and Selective Presentation (33 cases) (this category)
  4. 04. Exaggerated and Misleading Expressions (28 cases)
  5. 05. Emphasizing Efficacy While Downplaying Safety (22 cases)
  6. 06. Disparagement and Defamation of Competitors' Products (28 cases)
  7. 07. Undisclosed Conflicts of Interest and Improper Conduct in Lectures and Prescribing Guidance (10 cases)
  8. 08. Cross-Category Violations Rooted in Process Failures (4 cases)
Key points
  1. Visual manipulations—stretched axis scales, added guide lines, selective coloring—can constitute exaggerated advertising without altering a single number. The first audit step is a side-by-side comparison with the source figure.
  2. Removing control-arm data, omitting domestic trial results, and substituting secondary for primary endpoints all share the same logic: show only what favors the product. Verify against the original trial design and citation.
  3. The presence of data does not guarantee accuracy. The core audit question is what was removed or substituted, not whether a citation exists. Documenting and escalating that gap is the auditor's core function.
Sources
  1. MHLW, "Monitoring Project on Promotional Information for Prescription Drugs — Annual Reports" (FY2016–2024).
  2. PMD Act Article 66 (Prohibition of exaggerated advertising)
  3. Promotional Activity Guidelines, Section 1-3, Principle (1): Four requirements for accurate information provision — comprehensive presentation of the latest scientific evidence
  4. Promotional Activity Guidelines, Section 1-3, Principle (2): Seven prohibited acts — prohibition of selective, one-sided, or exaggerated presentation