How Do You Coordinate PSUR Translation Before the EMA Deadline
- 1 day ago
- 6 min read
You have five days. Once the CHMP adopts its opinion on a PSUR single assessment, the amended product information has to reach member states in every affected EU language on the linguistic-review clock, and the version you started translating too late is the one that slips. The report behind that assessment runs on its own earlier deadline. A periodic safety update report, the recurring benefit-risk review a marketing authorisation holder files on a fixed schedule, is due within 70 calendar days of the data lock point for intervals up to 12 months, and within 90 days for longer intervals.
Coordinating the translation is one sequence: freeze the scope, lock the term base, produce every language in parallel, check the versions against one master, then hand off for submission. Run it in that order and the deadline holds. Skip a step and you pay for it at review.
The two deadlines you're translating against
Two clocks matter here, and they measure different things. The report itself is submitted in English through the EMA PSUR repository, mandatory since 13 June 2016, so the core document is not where the language work sits. The frequency and data lock points come from the EURD list under Article 107c of Directive 2001/83/EC, and the single assessment runs under Article 28 of Regulation (EC) No 726/2004.
The translation load lands after assessment. When a single assessment recommends changing the summary of product characteristics, the package leaflet, or the labelling, those annexes have to be produced in up to 24 official EU languages, plus Icelandic and Norwegian for the EEA. The window is short. Under the centralised linguistic review, marketing authorisation holders submit translations to member states by day five after the CHMP opinion, and the full amended product information is due on the date in the translation timetable issued with that opinion.
The workflow, in five steps
Run the coordination as a fixed sequence. Each step exists to protect the one after it.
Freeze the scope. Lock the final English source and the exact annex sections the assessment changed, then stop accepting edits. Every downstream language references this one version.
Lock the term base. Fix the MedDRA terms, product-name conventions, and QRD standard statements before translation starts, so all languages draw from the same approved glossary.
Produce every language in parallel. Assign all target languages at once against the frozen source, not one after another, so the slowest language sets your timeline instead of the sum of all of them.
Check across versions. Reconcile each language against the frozen English master and the QRD template, confirming a warning phrased one way in German reads the same way in Finnish.
Hand off for submission. Deliver the reviewed annexes to regulatory affairs in the format the variation and the repository need, with the audit trail attached.
The order isn't cosmetic. A term base locked after step three means re-editing every language you already finished.
Freeze the source before anyone translates
The single biggest cause of blown PSUR translation deadlines is a source change after languages are in production. One edit to an English warning, made on day three of a five-day window, ripples into every version already drafted. This is the same mechanic that makes a single SmPC change cascade into retranslation across every national version: the source is one document, the outputs are two dozen.
So freeze first. Version-lock the English annexes, agree the scope in writing with regulatory affairs, and treat any later change as a controlled exception with its own re-review, not a quiet swap. Then lock terminology. A term base that fixes MedDRA coding, product names, and the QRD standard phrases before translation starts is what keeps 24 versions saying the same thing.
The bottleneck is single-threaded translation
Most teams miss the deadline for one structural reason: they run languages as a queue. If Finnish waits for German to clear review, and Portuguese waits for Finnish, the day-five window closes long before the last language moves. Parallel production breaks that. You assign the whole language set at once against the frozen source, gated by the shared term base so parallel doesn't drift into divergent.
This is the coordination layer that separates EMA pharma regulatory translation delivered on time from delivered late. The scientific content is fixed by the assessment. What you control is whether 24 languages move together or one behind another.
How AD VERBUM coordinates PSUR translation
AD VERBUM runs PSUR and product-information translation as term-base-driven parallel production. We ingest your translation memory and term base first, generate constrained output on client-tuned models through our LangOps System, then route every language through certified subject-matter review. Our process holds ISO 17100, which requires an independent second linguist to revise each translation, so no safety text ships on a single pair of eyes.
We host on EU infrastructure certified to ISO 27001, which keeps special-category safety data inside the EU as GDPR Article 9 expects. With 3,500+ subject-matter linguists across 150+ languages, all 24 EU versions move on one clock. The same discipline applies when you need certified translation of clinical trial documents under CTR 536/2014, where the source is fixed and the deadline is fixed with it.
Where it goes wrong
Most missed deadlines trace back to four avoidable failures.
Late source changes. An edit to the English annex after languages are in production forces rework across the whole set, and the later it lands, the more it costs.
Unmanaged terminology. Without a locked term base, MedDRA terms and QRD standard statements drift between languages and get flagged at linguistic review.
Single-threading. Running languages one after another through a single reviewer queue guarantees the day-five window closes before the last version clears.
Starting after the opinion. Waiting for the CHMP opinion to begin translation wastes the assessment weeks you could have used to stage the term base and translation memory.
Fix the first two and the deadline stops being the enemy.
Our pharmacovigilance translation services
Our translation services for regulated sectors run on ISO 27001 and ISO 42001 certified, EU-hosted infrastructure, with no reliance on public cloud tooling for core processing. Every project runs through our AI+HUMAN hybrid workflow: we ingest client Translation Memories and Term Bases first, our proprietary LLM-based LangOps System generates output constrained by client terminology on client-tuned open-weight models, and our certified subject-matter experts review for technical accuracy and regulatory compliance. Our QA is aligned to ISO 17100 and ISO 18587, with sector-specific requirements such as Directive 2001/83/EC pharmacovigilance obligations and the EMA linguistic-review timetable applied where relevant. We serve Life Sciences, Legal, Finance, Defense, and Manufacturing clients across 150+ languages with 3,500+ subject-matter linguists. For teams managing audit-sensitive content, contact us to discuss your security and compliance requirements directly.
FAQ
What is the PSUR submission deadline?
A periodic safety update report is due within 70 calendar days of the data lock point for reporting intervals up to 12 months, and within 90 calendar days for intervals over 12 months. The date is legally binding and published in the EURD list under Article 107c of Directive 2001/83/EC. Ad hoc reports requested by a competent authority follow the date in the request, or 90 days by default.
Is the PSUR itself translated into every EU language?
No. The report is submitted in English through the EMA PSUR repository, mandatory since 13 June 2016. The multilingual work comes after assessment, when a single assessment outcome changes the summary of product characteristics, package leaflet, or labelling and those annexes have to be produced in each affected national language.
What triggers the tight translation deadline?
The CHMP opinion on a PSUR single assessment. Under the centralised linguistic review, marketing authorisation holders submit product-information translations to member states by day five after the opinion, with the full amended annexes due on the translation timetable issued with it. That five-day window is why the term base has to be locked in advance.
What does ISO 17100 add to safety translation?
ISO 17100 requires an independent second linguist to revise every translation, so no language version ships on one reviewer's judgment. For safety text, where a single mistranslated contraindication is a compliance event, that revision step is the control. We hold ISO 17100 across our pharmacovigilance work at AD VERBUM.
How do you keep 24 language versions consistent?
A locked term base and a version-locked English source. Every language references one frozen master and one approved glossary of MedDRA terms and QRD standard statements, then each version is reconciled against that master before handoff. Without that step, small wording differences between languages surface at linguistic review.
Does GDPR affect PSUR translation?
Yes, where reports or patient-facing material carry personal data. Pharmacovigilance and clinical data are special-category data under GDPR Article 9, so translation has to run inside controlled, EU-hosted infrastructure. At AD VERBUM we process this content on ISO 27001 certified systems hosted inside the EU.
