Ch.2 Comprehensive product information summary | JPMA Guide Commentary | Advertising Regulation in Japan (Promotional Material Review) and Artificial Intelligence

Chapter 2 covers the Comprehensive Product Information Summary, the main body of Part I. It gathers the full picture of a single prescription drug — efficacy, safety, and handling — into one document so that healthcare professionals can take in, at a glance, everything they need for a prescribing decision. It is the thickest and most rigidly structured material of all. Where Chapter 1 set out ground rules that apply to any material, this chapter has one defining feature: the Guide dictates what must be included, and in what order.

The author cannot pick and choose which items to include. The freedom to "play up the favorable sections and thin out the unfavorable ones" is foreclosed in advance by a fixed set of items and a fixed sequence. This builds the spirit of the foreword — the package insert is the original, the summary merely "supplements" it; inform accurately without misleading — directly into the page structure itself.

01Why the fixed items and order are a design philosophy

There are two forms. The Comprehensive Product Information Summary covers the whole picture; the Specific-Item Product Information Summary narrows to selected items. Even the latter cannot escape Chapter 1's basic considerations, and for the items it does cover it must follow the relevant Chapter 2 rules. In other words, "keeping it short" is allowed; "diluting the rules" is not.

Sixteen items are prescribed: (1) items on the cover, (2) development history, (3) features/characteristics, (4) product information (DI), (5) clinical results, (6) pharmacokinetics, (7) pharmacology, (8) safety pharmacology and toxicity studies, (9) physicochemical properties of the active ingredient, (10) pharmaceutical matters, (11) handling precautions, (12) packaging, (13) related information, (14) principal references, (15) the name and address of the marketing authorization holder, and (16) the date of preparation or revision.

What the sequence itself says

These sixteen items are not a loose bullet list. Read in order, they form a single arc. Development history explains why the drug came to be; features and clinical results present the core of efficacy; pharmacokinetics, pharmacology, and toxicity establish the drug's profile; handling and packaging carry the practical detail; and the document always closes with principal references (where the evidence lives) and the date of preparation or revision (which version this is).

That the ending is sealed with "references" and "date" is no accident. State a claim, then always end by attaching its source and its point in time — this implements, at the level of the table of contents, the foreword's idea of guaranteeing verifiability through structure. By looking at the final two items, a reader can always return to the primary source and check whether the version is current.

02The cover — placing the most critical safety information in first sight

Item (1), the cover, is the product's "face," one of the few places where the Guide reaches into how things look. It carries the Japanese Standard Commodity Classification number (down to the detailed subclass), the therapeutic classification name consistent with the product title, the regulatory category, the name, and whether the drug is listed in the NHI price standard. For the regulatory category, the full text of the applicable designation — poisonous/powerful drug, narcotic, psychotropic, prescription-only drug, conditional approval, and so on — is appended to the name.

Warnings and contraindications reproduce the full text of the package insert, broken into sections, placed in a visible spot on the cover with attention to framing, background color, and text color, and shown clearly in gothic typeface at 10 points or larger. Only when there are so many contraindications that a prominent layout becomes difficult may the reasons for the setting be omitted — but the items themselves can never be dropped.

A new drug subject to post-marketing surveillance carries the unified mark on the cover for six months after launch. A product with a Risk Management Plan (RMP) is marked "Risk Management Plan target product" on the cover. Specifying point size, typeface, and framing may look excessive. But if "inform accurately without misleading" is taken seriously, then the size and placement of text are exactly what determine the quality of communication. Safety information that is small and tucked into a corner has merely been "included," not "communicated."

03Development history and features — brakes that keep the story from running away

(2) Development history: tell the story, but never depart from the approved facts

You may describe the background to development, the development process, the clinical positioning, and the overseas approval and marketing status. It is one of the few columns where a narrative is allowed — which is precisely why brakes are fitted. When efficacy, indications, or dosage differ between overseas and Japan, the domestic approval content must be written out accurately, including any restrictive wording. References to existing drugs must not become disparagement of competitors. If improved safety was the main aim of development, that may be stated, but it must not turn into emphasis or a guarantee of safety.

(3) Features/characteristics: the column most prone to exaggeration

The features column is where the writer most wants to "sell," and therefore the one most prone to exaggeration. The Guide lays a dense set of constraints here. Write efficacy and safety in balance; make no false or exaggerated claims of efficacy and no emphasis or guarantee of safety. Do not contradict the warnings and contraindications. State indications and dosage accurately, restrictive wording included. Do not put "reference information" in this column. If you state efficacy or safety, carry the supporting data within the material and cite the page where it appears.

In places the discipline of science has become regulatory text outright. If a comparison against another company's drug is shown, the trial title must be given and only the result of the primary endpoint (the confirmatory analysis item) may be presented. A "good number" picked up post hoc from a secondary endpoint or a subgroup cannot be made the lead. This translates into rules of expression the statistical practice of strictly separating "confirmatory analysis that tests one pre-specified hypothesis" from "exploratory analysis that searches for hypotheses."

Detailed rule: brakes that reach even how figures are drawn

Figures in the features column carry a concrete fifth-level rule. If the number of cases is fewer than ten, do not graph the response rate or express it as a percentage; show it as raw counts ("● of ■ cases"). Turning a handful of cases into a percentage makes the result look more certain than it is. When there is no significant difference, the hazard ratio may be shown, but the relative risk reduction must not be written. Differences from a control must not be visually emphasized with arrows and the like. Safety must be consistent with the package insert; note the serious adverse reactions, the major adverse reactions, and the fact that the package insert and the clinical-results safety findings are to be consulted. The adverse reactions of placebo or the control drug are not written in this column. Animal data are labeled "(animal species)" and in vitro data "(in vitro)," and are not connected directly to the clinic.

04Product information (DI) — keeping the window onto approved facts in sync with the latest version

Item (4), product information (drug information), is the "window" onto approved facts placed inside the material. At the top of the relevant page, state prominently that one must pay close attention to revisions of the precautions and related information, including warnings and contraindications; write in accordance with the latest package insert; and note that package insert's date of preparation or revision. A material can begin to go out of date the moment it is printed. So it declares of its own accord, "as of when did I transcribe the approved information." This too is part of verifiability.

05Clinical results — the thickest chapter in the Guide

Item (5), clinical results, presents the core of efficacy and is the most voluminous item in the Guide. Yet the gateway for "data that may be carried" is tightly narrowed.

The range of trial results that may be carried

Results evaluated during the domestic approval review
Evaluation materials used at the time of package insert revision
Original papers that have undergone rigorous peer review (conference presentation data and review articles excluded)
Results evaluated as re-examination application materials

Note that peer review is the watershed of quality. In the evidence hierarchy, meta-analyses, systematic reviews, and RCTs sit at the top, while non-peer-reviewed conference presentations and reviews are hard to use as evidence as they stand. A comparative clinical trial may be carried only if it is double-blind, randomized, or material evaluated during the approval review as an accepted substitute for a double-blind design. Randomization evens out known and unknown biases; blinding removes the bias of expectation and subjectivity. The quality of the design is itself the condition for inclusion.

The four conditions for RWE (real-world evidence)

To carry real-world data, four conditions must be met: it reinforces or supplements the approved scope; the fairness of control, concomitant, and reference drugs is secured; it is presented alongside the confirmatory trial (usually an RCT) that is the principal basis of approval; and the important limitations from bias are stated prominently. RWE reflects real practice but is prone to confounding. Hence the stance required: do not let it stand alone as the lead, and do not hide its limits.

Items to record and the "explicit positioning"

Clinical results record the trial title (phase; the control drug by generic name; own-company drugs may be shown by brand name), the type of trial (note overseas data or international joint trials), the trial method (objective, subjects, number of cases, dosing, endpoints, analysis plan, judgment criteria), the source, and conflicts of interest. Most important is distinguishing the positioning of endpoints: make clear which is a confirmatory analysis item and which is no more than a nominal p-value.

A p-value is only "the probability that, assuming no difference, this difference would arise by chance"; it says nothing about the size of the effect or its clinical value. A nominal p-value — any p-value from analyses other than the pre-specified confirmatory analysis — cannot be used for a confirmatory conclusion. That is why the Guide asks that the favorable parts not be singled out for emphasis, that figures and arrows not oversell, and that the writing let the reader tell a confirmatory result from a nominal p-value.

Detailed rule: safety is handled differently depending on the column

For the very same drug, how safety is handled changes with the column it is written in. This is an easy place to err.

Point	Features column	Safety column of clinical results
Adverse reactions of control / placebo	Not written	Record event names, case counts, and incidence for both the drug and the control (placebo included)
Serious reactions / those leading to discontinuation	Serious/major reactions plus a note to consult the DI	Record event names and case counts. Do not write "there were no serious adverse reactions"
Significance tests / confidence intervals	As a rule, not shown	Do not carry between-group significance tests or confidence intervals unless the endpoint is primary and confirmatory (per-group estimates are allowed)

The principle of not emphasizing safety runs throughout. One does not wheel out a significance test showing no difference from placebo to create the impression that the drug "is safe." Control-drug results may be carried as fact, but not evaluated or commented on. This is the asymmetric duty the foreword describes — safety-related information is disclosed even when it is unfavorable to one's own company.

Detailed rule: a presentation that never steps beyond the approved scope

As a rule, present within the approved scope. Even where dose adjustment is permitted, do not carry efficacy data above the approved dose. Do not use trial data inconsistent with the starting dose or the method of adjustment. A dose-finding trial that includes off-label dose groups must be labeled "dose-finding trial," with the approved dosage noted. When presenting results that include off-label content, state at the outset "the fact that part is off-label and the reason for inclusion," and note the corresponding indication and dosage. Results in which a dose group contains off-label cases may not be conveniently re-analyzed. Subgroup analyses are not carried, except those that were in the original plan and are scientifically valid.

At the top of the first page of clinical results, place "warnings, contraindications, etc. on page ○○" in a larger point size than the body text. Precisely because it is where the eye goes to the efficacy figures, the route to the safety information is shown first.

Detailed rule: case presentations are "as a rule not made"

Case presentations easily become a gateway to overselling exceptional data, and as a rule are not made. The exceptions are limited to alerting to adverse reactions, a small number of cases of a rare disease, and situations such as contrast agents that can only be shown through images (fictional model cases are treated the same). When made, cite the source, name competitors generically, do not emphasize or guarantee efficacy or safety, and do not extend from a case to an evaluation of the drug as a whole. At the top, in larger type than the body, place the note that "the presented cases are only a part and not all cases show the same result," and always record any adverse reactions.

06The items that show the drug's profile — pharmacokinetics, pharmacology, toxicity

(6) Pharmacokinetics

Record absorption, distribution, metabolism, and excretion (ADME) in humans. If there are no human data, animal and in vitro data may be carried as a supplement, labeled "(animal species)" and "(in vitro)." State the subjects (healthy people/patients, adults/children), and label foreign data "(foreign data)" in the title. Pharmacokinetics in special-condition patients — the elderly, those with renal or hepatic impairment, those on dialysis — may be recorded where reference data exist, and where there is support, a discussion of dose and interval according to organ function is allowed. But this must not turn into an emphasis on safety for these patients. For drugs requiring TDM, show the key parameters such as blood concentration, the principal elimination route, and drug metabolism.

(7) Pharmacology

Based on clinical and non-clinical pharmacology studies, record the pharmacological action and mechanism that support the approved indication. When results that deviate from the approved dosage are carried, the approved dosage must be noted alongside. Comparison with a control drug is fact only — no commentary — and the comparison must not be emphasized in titles or figures. For combination drugs, do not mislead about efficacy through the pharmacology of individual components, and show synergy only with objective data. Non-clinical data are labeled "(animal species)" and "(in vitro)" and must not emphasize or guarantee clinical efficacy or safety. A pharmacological action whose relation to the indication is unclear is shown as "reference information," without emphasis.

(8) Safety pharmacology and toxicity studies

Record safety pharmacology and toxicity studies on the central nervous, cardiovascular, and respiratory systems and so on, based on animal and in vitro results. As a rule, only the facts of the drug's own study results are given; competitors' products are not recorded. No expression that leads to emphasizing clinical safety is used. And — if there is knowledge of a pharmacological action or toxicity suggesting the possibility of an adverse reaction in the clinic, it must be recorded. Even when unfavorable, information on the cautionary side is not buried. Item (8) is where that asymmetric duty appears in its plainest form.

07The later items that carry practical detail and verifiability

Item (9), physicochemical properties of the active ingredient, presents the generic name, chemical name, molecular formula, structural formula, and the like as objective fact. Item (10), pharmaceutical matters, records stability and compounding changes as the facts of study results only, without evaluative words like "suitable for compounding." Item (11), handling precautions, sets out storage, shelf life, and expiry under small headings; for specified biological products, it states under the package insert that usage records are to be kept for at least 20 years. Item (12), packaging, is a short column but practical information that helps distribution and prevents mix-ups.

(13) Related information: leaving the facts of version and reimbursement on record

Record the approval number and date (additional approvals up to the latest), the NHI listing date, the launch date, approval conditions, notes on insurance benefits, and the publication dates of re-examination and re-evaluation results. Even the handling of blanks is prescribed — for example, if there is no information, the item name itself is not listed. None of this is flashy, but it is the information that lets one later trace "when, and under what conditions, the drug was approved."

(14) Principal references, (15) marketing authorization holder, (16) date of preparation or revision

Item (14) is the column showing the references that back up the entries: if the clinical results are approval-time evaluation materials, say so; for an original paper, give the bibliographic details; for internal material, write it so the specific content is clear. References relating to clinical results are also carried, with bibliographic details, on the relevant clinical-results pages. Item (15) is the name and address of the marketing authorization holder (including the reference-request and inquiry contacts); for a drug held under a foreign special approval, the name of the selected marketing authorization holder is given along with the holder's name and country.

That the sixteenth item is "date of preparation or revision" itself expresses the character of the Guide. The close is no flashy prohibition. It keeps a record of the version and keeps the material in a state that can later be verified and recalled — and the accumulation of that plainness is what supports trust. A structure that begins with the cover (the face) in item one and ends with the date (the version) in item sixteen is the final bolt of a consistent design: "inform accurately, and let it always be checked."

In closing

The heart of Chapter 2 lies in the fixed set of sixteen items and their order. It narrows the author's editorial discretion, places dense brakes on the columns that present the core of efficacy (features and clinical results), and always puts the evidence and the version at the very end. This is the result of folding the foreword's spirit — the package insert is the original, the summary supplements it; inform accurately without misleading — into the page structure itself.

The detailed rules of each item — raw counts when fewer than ten cases, handling control-drug adverse reactions differently by column, always disclosing unfavorable toxicity findings — are concrete measures for blocking expression that is "factual yet misleading" and for protecting verifiability through structure. The detail of each item is explored further on the pages below.

Pages in this level

Case studies — the "it isn't written, so…" mindset, and how each lands

Chapter 2 constrains editorial latitude through 16 fixed items in a fixed order. The Characteristics and Clinical Data sections attract the most pressure toward exaggeration, and the supplementary rules concentrate on case counts, the designation of endpoints, and the handling of comparator data. The cases below focus on the sections where creative misreading is most tempting.

GRAY 1 — Straying from the spirit (untouched by the wording, but against the foreword's spirit)

The scheme: "In the Development Background section I noted that overseas approval covers indications A and B. Domestically only A is approved, and I put the qualifying condition in a small footnote at the end. Nothing I wrote is false, and the qualifying condition is technically present."

Verdict: The rule requires that when domestic and overseas indications differ, the domestic approval — including its qualifying conditions — must be "accurately distinguished." Foregrounding the broader overseas scope in the body text while burying the domestic qualifying condition in an out-of-sight footnote does not constitute accurate distinction. Readers will infer that indication B is also approved domestically. "Technically present" is not a guarantee against misreading.

GRAY 2 — Exploiting the gap (deliberately working a loophole to invite misreading)

The scheme: "The supplementary rule on the Characteristics section says 'if the case count is fewer than 10, do not use graphs or percentage notation for response rates.' My cohort has exactly 10 cases — not fewer than 10 — so I can write '80% response rate' and include a bar chart."

Verdict: The rule exists because presenting a small-sample percentage makes results look more reliable than they are. The risk of misreading does not change qualitatively between 9 and 10 cases. The threshold of fewer than 10 is a practical marker, not a line that makes 10 cases categorically safe. An 80% response in 10 patients shifts to 70% or 90% with a single case change; graphing it creates the impression of an established rate. The spirit of the rule applies to all small cohorts.

GRAY 3 — Violating higher norms (arguably "not in the text," yet almost certainly out under the PMD Act, Fair Advertising Standards, and the JPMA Code)

The scheme: "I put only RWE in the Clinical Data section. I know the rule requires RWE to be shown alongside the confirmatory trial, but the approved-study results appear separately in the DI section. A one-page RWE summary should be able to stand alone."

Verdict: One of the four conditions for including RWE is that it must appear "alongside the confirmatory trial that forms the primary basis for approval." Chapter 2's clinical-data rules apply even in focused materials. Presenting RWE without the confirmatory trial implies that a dataset inherently prone to confounding and bias constitutes the primary efficacy evidence. This falls under the Fair Advertising Standards' prohibition on implying superiority without adequate scientific basis, and conflicts with the JPMA Code's hierarchy-of-evidence requirements.

BLACK — Breaching the explicit text (a plain violation of the provision itself)

The scheme: "I included QoL score improvement data — a secondary endpoint — in the Characteristics section. It's clinically meaningful and compelling for patients. I know there are rules about reference information, but the Characteristics section is precisely the place to communicate what makes this drug special, and QoL improvement is one of those things."

Verdict: Chapter 2's supplementary rules state explicitly that "'reference information' must not be included in this section [Characteristics]." Chapter 1 designates QoL data as reference information as a general rule, and Chapter 2 carries that forward by barring it from the Characteristics section. The argument that "it is a characteristic so it belongs here" ignores the structure of the provisions. The material must be revised; QoL data must be moved to a clearly labelled reference-information section.

Improper intent by implication — three scopes of "insinuating" without stating it

Chapter 2 uses a fixed list of 16 items in a fixed order to prevent selective editorial weighting — the freedom to make favourable items thick and unfavourable ones thin is pre-empted by structure. Yet three patterns of implication can slip past line-by-line review of each section's wording: a choice of symbol, a two-sentence pairing, and a narrative arc running through the whole document. None of them contains a prohibited word when examined in isolation.

Implication 1 — Local: carried in a single word or sentence

How it is built: "Each entry in the Features section is prefixed with a filled circle (●). Format used: '● Clinical efficacy: statistically significant difference achieved on the primary endpoint (p=0.008).'"

Intent that can be read: The filled circle functions culturally as a mark of approval, excellence, or pass. Prefixing a factual statement with this symbol conveys the evaluative judgment "this feature is superior" without using a prohibited evaluative word. No explicit rule names the prohibition of a bullet symbol.

Verdict: A symbol is part of the expression. That "●" communicates a favourable evaluation in the reader's cultural context is a foreseeable cognitive effect, not an incidental one. Chapter 2, item 3 (Features section), prohibits false or exaggerated claims about efficacy; attaching a symbol that implies superiority to a factual sentence falls within that prohibition regardless of the accuracy of the underlying statement. The correct approach is to remove the symbol and present facts in plain text.

Implication 2 — Adjacent: read across neighboring clauses

How it is built: "Efficacy on the primary endpoint was confirmed in a confirmatory analysis (p=0.008). Signals in the same direction were observed in multiple analyses for secondary endpoints (QoL score, biomarker change)." — these two sentences appear consecutively in the Features section.

Intent that can be read: The first sentence establishes confirmatory evidence, and the second immediately offers secondary endpoint data as corroboration. The modest phrasing "signals were observed" does not block the reader from inferring that the efficacy has been confirmed at multiple levels.

Verdict: Chapter 2 restricts the Features section to results from the confirmatory analysis of the primary endpoint; secondary endpoint and exploratory results may not be mixed in without being designated as reference information. Placing secondary data immediately after a confirmatory result, without labelling it as reference information, conceals the difference in evidential standing from the reader. The correct approach is to separate the two sentences, move secondary and exploratory findings to a designated reference information section, and label them clearly.

Implication 3 — Whole-context: arising from the document as a whole

How it is built: "The Development Background section is written as an answer to 'Why was this drug needed?'; the Features section presents 'The answer was found'; the Clinical Data section frames the content as 'Data that prove it.' At the end of Clinical Data, a post-hoc subgroup analysis is displayed with a single-line note reading 'For reference.'"

Intent that can be read: Structuring the whole document as a 'problem → solution → proof' narrative gives everything that appears in the proof section — including the post-hoc data — the standing of evidence confirming the solution. The "For reference" note provides nominal distancing but does not interrupt the narrative logic that the reader has been following through three sections.

Verdict: Chapter 2 allows post-hoc and subgroup analyses only when they were pre-specified and scientifically justified; when included, the opening of that passage must state that the analysis is post-hoc and explain the reason for inclusion. A single-line end-of-section note cannot neutralise the evidential weight that the document's overall narrative has already lent to the data. The narrative frame constitutes a structural promotion of the post-hoc findings. The correct approach is to satisfy the conditions for including post-hoc analyses in full, and to present them in a clearly delimited section separated from the confirmatory narrative.