Clinical Protocols: The Terms a Bad Translation Risks

A clinical trial protocol isn’t read once — it’s consulted throughout the entire study, by the sponsor, the ethics committee, monitors, and the site team itself. A swapped term in an exclusion criterion, a rounded dose, or an adverse event rewritten with a loose synonym doesn’t show up as a “language mistake” — it shows up as an inconsistency between documents, and that’s exactly what sends a submission back for correction.

This guide covers what ICH-GCP requires from a translated protocol, why inclusion/exclusion criteria deserve extra scrutiny, and the checklist any protocol translation should pass before it reaches the ethics committee.

What ICH-GCP requires from a translated protocol

ICH-GCP (Good Clinical Practice, internationally harmonized) doesn’t set translation-specific rules — but it does set the precision standard the protocol itself must maintain, and that standard doesn’t disappear when the document changes language. A translated protocol remains the reference document for how the study is conducted: any ambiguity introduced in translation becomes ambiguity in how the study is actually run, across different sites.

In practice, this means protocol translation can’t be treated as flowing prose: it needs to preserve the numbered section structure, identical terminology across every repeated mention of a technical term, and traceability between the protocol and the other documents in the dossier (informed consent form, case report form, investigator’s brochure).

Inclusion and exclusion criteria: where a wrong term costs the most

Inclusion and exclusion criteria define who enters — and who doesn’t enter — the study, and by far they’re the section where terminological precision carries the highest cost of error. A swapped term here isn’t a style slip: it can change who the criterion is meant to include or exclude, with a direct effect on the validity of that group’s results.

That’s why inclusion/exclusion criteria should always go through term-by-term verification against the source — not a fluent read of the translated paragraph, which tends to sound correct even when a critical qualifier (“uncontrolled”, “within the last 6 months”, “stable”) has been dropped.

Consistency with the rest of the dossier

The protocol is rarely translated in isolation — it lives alongside the informed consent form, the investigator’s brochure, the case report form and, later, the package insert. An adverse event, drug name, or protocol code needs to appear with the exact same wordingacross all of these documents. When it doesn’t, the ethics committee flags the discrepancy and pauses review until it’s corrected — the most common, and most avoidable, delay in translated submissions.

An example of how this shows up in practice, sentence by sentence:

The first version loosely paraphrases the objective criterion (“creatinine clearance ≥ 60 mL/min”) into a vague equivalent (“normal kidney function”) — losing the exact value any eligibility criterion needs to carry.

What can never change in a protocol translation

Checklist before submitting to the ethics committee

How this works in practice

Every protocol processed by SealScribe runs through a glossary built from official sources (ICH-GCP, ANVISA’s Controlled Vocabulary, MedDRA in Portuguese, DCB for drug names), with each term locked consistently across the entire document — including across the protocol, the consent form, and other documents in the same dossier — and review before delivery. When the document contains patient or site identifying data, anonymization runs before anything reaches the cloud, without altering any clinical data — included in every order, at no extra cost.