Dosage units, contraindications, and adverse reaction lists look straightforward on paper. In practice they are among the most tightly scrutinised strings of text in any regulated industry. A single mismatched unit, an incomplete list of interactions, or a non-standard MedDRA term can turn a routine multilingual package leaflet into a regulatory rejection—or worse, a patient safety incident.
Regulators do not treat these documents as ordinary content. The European Medicines Agency’s Quality Review of Documents (QRD) templates set the structure, mandatory phrasing, and formatting for the Summary of Product Characteristics (SmPC), package leaflet, and labelling. Current practice centres on version 10.4, with version 11 having completed public consultation to improve the package leaflet’s key information section. Every language version must track the approved English source line-for-line. Standard statements cannot be paraphrased; Member State linguistic review checks fidelity and national “blue-box” requirements. Timelines are unforgiving: after a positive CHMP opinion, translations typically need to land within days for the subsequent Day 215/235 checkpoints.
MedDRA sits at the centre of the terminology problem. The Medical Dictionary for Regulatory Activities supplies the controlled vocabulary for adverse events, signs, symptoms, and indications that regulators expect to see. It is updated twice a year and maintained in multiple languages. Translators who invent near-synonyms or rely on general medical dictionaries risk breaking the hierarchical mapping that safety databases and electronic submissions depend on. Consistency across the SmPC, leaflet, and labelling is not a stylistic preference—it is a review criterion.
The consequences of getting it wrong are measurable. Analysis of FDA drug recall data from 2012–2023 attributed roughly 19 percent of recalls to labelling or packaging issues. A separate look at 2017–2019 recalls put labelling errors at about 15 percent of the total. On the patient side, a well-documented study of computer-generated Spanish prescription labels in New York pharmacies found a 50 percent error rate. One recurring failure left the English word “once” untranslated; Spanish-speaking patients read it as the number eleven and took eleven doses instead of one. Similar patterns appear in other languages when dosage instructions, route of administration, or contraindications drift even slightly from the source.
These failures are rarely the result of careless freelancers. They arise when projects lack a locked termbase built from MedDRA preferred terms and QRD-approved phrases, when review cycles skip native-speaking medical reviewers, or when machine output is treated as final rather than as a draft requiring specialist post-editing. Dosage units are especially brittle: milligrams versus millilitres, international units versus micrograms, frequency expressions that do not map cleanly across languages. Contraindications and interactions demand the same precision; an omitted or softened warning can expose both patients and the marketing authorisation holder.
What works in practice is narrower than most general translation processes allow. Successful teams maintain version-controlled glossaries that distinguish QRD standard statements from product-specific language, map every adverse reaction to the current MedDRA release, and run a final linguistic check against the exact template version in force. They also treat the package leaflet’s readability requirements as non-negotiable: the text must remain accessible to the intended patient population while still satisfying the regulator’s technical criteria. Cross-checking SmPC, leaflet, and outer packaging for terminology alignment before submission catches many of the discrepancies that later surface in Member State comments.
Pricing for this work reflects the risk profile. Multilingual QRD-compliant packages covering all official EU languages plus Norwegian and Icelandic are not priced like marketing copy. The cost is driven by specialist linguists, medical reviewers, terminology management, and the tight procedural windows rather than by word count alone. Agencies that treat the work as commodity translation routinely underestimate the rework that follows linguistic review findings.
The same discipline applies outside Europe. FDA labelling under 21 CFR Part 201, Health Canada’s bilingual requirements, and national authority expectations in other major markets all demand equivalent control over dosage language, contraindications, and adverse event terminology. Global dossiers therefore need a single terminology backbone that can be adapted to local templates without introducing divergence.
Artlangs Translation has spent more than twenty years operating inside these constraints. With proficiency across 230-plus languages and a network of more than 20,000 professional linguists, the company has delivered QRD-compliant product information, MedDRA-aligned safety terminology, and full labelling packages for pharmaceutical clients worldwide. Its work extends beyond static documents into video localisation, short-drama subtitle localisation, game localisation, multilingual dubbing for short dramas and audiobooks, and large-scale data annotation and transcription—capabilities that support the broader content ecosystems surrounding modern medicines. The through-line remains the same: terminology that regulators recognise, units that patients can act on correctly, and processes that keep submissions on the critical path rather than in the revision cycle.
