How to translate Clinical Outcome Assessments (COAs)
Preparation and teamwork are required to translate Clinical Outcome Assessments (COAs). As you know, creating and validating a measurement instrument in one language can be a complex process. Translation requires a similar standard of care. In addition, internal review boards (IRBs) and/or regulatory agencies will require proof that qualified professionals translated and validated multilingual instruments.
The translation and validation process takes time. To avoid delays, establish a workflow and clarify the roles of all participants. Below, we describe the process and procedures followed when we translate Clinical Outcome Assessments for patients (PROs), clinicians (ClinROs), observers (ObsROs), and performers (PerfOs).
Identify the needs of the target population
- What are the demographic characteristics of the target population? Should all speakers of a target language be able to read the instrument? Alternatively, will it be tailored to a particular region or country? Remember, people speak different dialects of Arabic, English, French, Spanish, and other languages around the world.
- Will an interviewer ask the questions verbally, or will the subjects read the questions and write the answers? Translators will always strive for simplicity of language. However, they also need to know if they should write for a particular level of literacy or in a specific formal register.
Establish the internal review team
- Carefully choose and educate the members of your internal review team at the start of the project. Internal review commonly becomes a bottleneck. Avoid delays by choosing the reviewers carefully and engaging them early in the process. We have written elsewhere about best practices for internal review. To summarize:
- Choose reviewers with the appropriate language ability. Even if a reviewer is a native speaker of the target language, he or she may not be familiar with the target locale (for example, a European Portuguese speaker will speak and write differently than a Brazilian Portuguese speaker).
- Choose reviewers with relevant subject matter expertise.
- Choose reviewers who have the time, motivation, and temperament to give adequate attention to the materials and collaborate productively with the rest of the team.
- Be clear about the reviewer’s role. The reviewer should not try to “make it better.” Instead, they should only flag instances of truly confusing or inaccurate text. The reviewer should know the difference between substantive and preferential changes. Preferential changes will bog down the process significantly.
Conceptual analysis and translation
- Before beginning the translation, perform conceptual analysis of the English source. At this point, a conversation between the internal review team and the translation team helps clarify the meaning, context and goals of the questionnaire. Engage a bilingual cultural consultant who has deep familiarity with the culture of the target population. He or she will identify unfamiliar or confusing items, terms, and concepts for clarification.
- If you’ve already chosen internal reviewers, ask them to weigh in before the translation process begins. Include notes of clarification in the source to provide context for the translation team.
- Send the adjusted and annotated source document for translation. The process includes translation by a native-language professional linguist, bilingual review by a second native-language linguist, and in-house quality assurance. If required, more than one “forward” translation can be created. These will be reviewed and reconciled during linguistic validation.
- At this point, the review team will engage in linguistic validation of the translated instrument. Problems will need discussion and resolution.
- As mentioned above, the reviewer’s job is to confirm that the meaning of each item is clear. A reviewer should not try to “improve” the writing. Limit suggested changes to:
- Questioning consistency of meaning between the source and target instruments
- Identifying confusing or inappropriate words or expressions.
- Every suggested correction should include a note to the translator explaining why the wording will obstruct the subject’s understanding of the question, and how it should be fixed.
- Changes recommended by the review team will be incorporated by the translation team to create a finalized translation.
- Back translation provides conceptual validation of the translation. A native English-speaking independent professional translator who has no knowledge of the original source document will “back-translate” the finalized translation into English.
- Back translation validates conceptual equivalence, not linguistic equivalence. Substantial differences between the meaning of the original source text and meaning of the back translation must be discussed and reconciled. This may require meetings between the translator, back-translator, and reviewers.
- Some studies require not one, but two independent back translations followed by reconciliation. The two back translations must be conceptually equivalent, not linguistically identical.
Pre-testing and cognitive review
- Pre-test the translated instrument with individuals who are representative of the target population. Each subject will be asked to repeat the question in their own words. They’ll explain what they think the question is asking and how they decided on their response. The testers will flag consistently confusing or misleading elements for further review. These debriefing sessions should be closely documented (ideally, recorded and transcribed) and the findings reported to the review committee.
Finalizing the questionnaire
- If pre-testing indicates that significant changes are required, a second back-translation and reconciliation session may be warranted. However, if everyone stays apprised of the process as it unfolds, and the translation team and reviewers understand the needs of the target population, changes at this point will probably not be necessary.
Documentation should include
- The initial forward translation
- A summary of recommendations by the review team
- The back-translation(s)
- A summary of any modifications proposed during pre-testing
- A linguistic validation certificate if required
- Any other requirements specific to the IRB or regulatory agency
Variations on this process apply to questionnaires, surveys, and informed consent forms. We also follow similar procedures when translating for medical devices and pharmaceuticals. For example, IFUs (instructions for use) and any other high-risk instructions require the highest level of accuracy.
In summary, prepare to translate clinical outcome assessments by establishing a schedule and organizing your team. Following best practices simplifies and streamlines the translation process. As an ISO 17100-certified language service provider, Scriptis follows processes that meet all regulatory and compliance standards. We have the resources to meet your language needs for clinical research.