Before writing any section of a product information summary, the author must know the precise contour of the approval. The electronic package insert, the approved indications, and the approved dosage are the reference lines against which every line in the material is measured. Beyond these fixed points, post-approval evidence — RMP, early post-marketing phase vigilance, re-examination and re-evaluation, and real-world evidence — does not widen the approval; it reinforces the evidentiary foundation that the approval rests on across time. Peer-reviewed original articles underwrite the quality of that foundation.

The three pillars of the foreword — ① blocking misunderstanding, ② implementing balance on the printed page, ③ ensuring verifiability — all presuppose an accurate reproduction of the approval. Without knowing the contour of the approval, it is impossible to judge whether a given sentence blocks or creates misunderstanding. Balancing efficacy and safety information requires a grasp of both sides as the approval document presents them. And traceability to primary sources depends first on knowing which sources were consulted. Understanding these terms precedes all three pillars.

01Term reference — regulatory and approval vocabulary

TermMeaningNotes and implications
Electronic package insert The primary source for proper-use information for a medicinal product. The product information summary exists to supplement it Any discrepancy between the summary and the package insert converts supplementation into deviation. The package insert must be checked before any material is drafted
Scope of approval The full set of indications, dosage regimens, and patient populations that a regulatory authority has authorized One word beyond this scope is off-label — including hints and suggestions, not only direct statements
Indications and effects The approved indication as it appears in the electronic package insert, including all qualifying wording. The full text — not the condition-free indication alone — constitutes the approved indication The same standard applies wherever the indication is mentioned — development history, features column, anywhere
Dosage and administration The approved route, dose, and schedule. Both the ceiling and the floor are set by the approval document The phrase "adjust as appropriate" confers discretion within the approved range; it is not a waiver of the ceiling
Qualifying (restrictive) wording A limiting condition attached to an indication: "only when [therapy X] is inappropriate," "patients for whom existing agents have been insufficient" A condition, not a footnote. Drop it and the shape of the approval changes
Adjust as appropriate A phrase in the dosage section that permits dose calibration to the patient's condition — within the upper and lower bounds specified in the approval document Not a basis for presenting efficacy data from doses above the approved maximum
Reference information Secondary findings obtained within the approved study: QoL, biomarker changes, exploratory subgroup results. Not primary confirmatory outcomes Must be labeled "reference information" and kept out of the features column. Mixing it with primary endpoints invites exaggeration
RMP (risk management plan) A post-approval framework that systematically identifies and minimizes a product's risks through safety specifications and risk minimization measures Safety concerns identified in the RMP stand alongside package insert warnings as core safety information. No material may emphasize safety in a way that conflicts with the RMP
Early post-marketing phase vigilance Mandatory collection of safety information from healthcare institutions during the period immediately following a new approval The product's safety profile is still being established during this period. Statements confirming safety are especially hazardous at this stage
Re-examination / re-evaluation Re-examination: review of a product's efficacy and safety using post-marketing use-results data. Re-evaluation: reassessment of an already-approved product against current scientific standards While either procedure is ongoing, materials must be based on the current approved content. Anticipating outcomes not yet confirmed is impermissible
Peer-reviewed original article A primary publication that has undergone independent critical review by subject-matter specialists before appearing in an academic journal The highest-quality citable form for establishing verifiability. Abstracts, press releases, and conference slides are not substitutes
RWE (real-world evidence) Knowledge derived from data generated during routine clinical care — electronic health records, registries, claims — rather than controlled trials Carries higher bias risk than randomized controlled trials. The conditions under which RWE can support an efficacy claim are limited. RWE and RCT data must never be merged without clear differentiation

02The electronic package insert — what it means to be the primary source

The product information summary supplements the package insert; it does not precede it. The starting point for every line in any material is the package insert, and nothing may be stated in a material that the package insert does not contain.

The corollary is equally important. The risk information the package insert communicates — warnings, contraindications, serious adverse reactions — exists as a prior fact that no material can ignore. When a material selects only the safety information that favors the product, a substantive discrepancy with the package insert emerges. Recognizing the package insert as the primary source is the foundation of what the foreword calls the "asymmetric obligation" — the duty to disclose information even when it is unfavorable to the company.

03Qualifying wording — reproducing the approval means copying the condition too

Qualifying (restrictive) wording is a limiting condition attached to an indication. "Only when [treatment X] is inappropriate," "patients for whom existing therapies have proved insufficient" — these phrases are not contextual color added to the indication; they are part of the definition of the indication itself.

Dropping the condition alters the approval

What happens when an indication is written in a product information summary without its qualifying wording? The healthcare professional reading it understands that the indication may be used without restriction. What the package insert approved conditionally appears in the material as an unconditional approval. This is not a reproduction of the approved fact; it is an alteration of it.

The same standard applies in every column where the indication is mentioned — the development history column, the features column, and anywhere else. When domestic and overseas indications differ, the domestic approved content must be written separately and include the qualifying wording in full.

04Reference information — the partition from the features column must hold

Reference information consists of findings obtained incidentally within an approved study — QoL measurements, biomarker changes, exploratory subgroup analyses. These are not the confirmatory results of the primary endpoint, which is what the features column is reserved for.

Reference information must be clearly labeled as such. The features column must contain only confirmatory primary results; placing reference information there is impermissible. The features column is the section most prone to exaggeration, and when secondary findings infiltrate it, unconfirmed knowledge begins to carry the same apparent weight as the evidentiary basis for the approval. The foreword pillar of "blocking misunderstanding" is put into practice precisely by maintaining this partition.

05RWE and peer-reviewed articles — evidence quality and verifiability

Real-world evidence (RWE) is knowledge drawn from routine clinical data. Because it is not a randomized controlled trial, it inherently carries risks of selection bias, confounding, and inconsistent observation periods. Placing RWE on equal footing with RCT results is not permitted. When RWE is included in a material, the text must state that it is derived from real-world data and identify the design's limitations.

A peer-reviewed original article — whether it reports an RCT or a real-world study — is the most verifiable form of citation available. When citing, author, journal, volume, issue, and year must be given so that the reader can trace back to the primary source. This is the concrete implementation of the foreword's third pillar: verifiability. Press releases and conference slides are not acceptable substitutes.

In closing

Because the electronic package insert is the primary source, the content of any product information summary begins with the approved facts. Copy qualifying wording together with the indication. Keep reference information out of the features column. Distinguish RWE from RCT data. Present citations in a traceable form. Each of these practices is not a separate compliance exercise but a consequence of a single requirement: communicate the contour of the approval accurately so that readers can return to the primary source. The three foreword pillars all stand on an accurate understanding of the terms defined here.