Translation and linguistic validation of 24 PROMIS item banks into French

The 24 adult PROMIS item banks and scales that underwent English to French translation are listed in Table 1. Domains were prioritized by patients and clinicians from chronic pain management programs and represent the most important domains selected across chronic conditions [20]. All item banks use a 5-point Likert scale and query the past 7 days except for the physical function and participation item banks, which do not use a recall period, the self-efficacy item banks, which ask about the respondent’s current level of confidence and the psychosocial illness impact - positive and negative banks, which ask respondents to consider how their illness has affected them and to rate how true statements are for them before their illness, again now and since their illness.

The PROMIS Health Organization (PHO) in 2018 provided authorization to the Canada PROMIS National Center (PNC) representatives (SA, SB) to translate 24 adult PROMIS item banks into French. We followed translation recommendations by ISPOR [21] and used best practice methods of FACIT [22, 23] and PROMIS. The translation was conducted in partnership between FACITtrans and the Canadian PNC using the Functional Assessment of Chronic Illness Therapy (FACIT) translation methodology (Table 2) which targets semantic/cultural, content, and conceptual equivalence [22]. Translation methods were based on consensus-based best practices [23, 24]. To meet PROMIS methodological standards, a universal approach to translation was adopted. We aimed to create one French version of item banks that can be used in all French-speaking countries. A benefit of this approach is that the final translations should not require further adaptation for other countries, although additional testing in new countries is recommended.

Interested parties including clinicians and members of the general population were involved in the original development and validation, and during cultural and linguistic validation (through cognitive debriefing). All members of the French translation team met International Organization for Standardization (ISO) 17,100 standards for professional competencies and translator qualifications for linguists and were under contract with FACITtrans. All were native speakers of French with the exception of the Back Translator who was a native speaker of English fluent in French.

The translation process was iterative and included native French speakers from Belgium, Canada, France, and Switzerland. Forward translation into French by two native French-speaking translators was followed by reconciliation by a third native French-speaking translator. A native English-speaking translator fluent in French then completed a back translation of the reconciled version from French into English. The source and back-translated versions were then compared by FACITtrans to identify discrepancies and evaluate translation issues, providing guidance to further refine the translations during the subsequent review phase. Three independent reviews by bilingual translators from Belgium, Canada and France were completed to assess the clarity and consistency of terminology and equivalency across the English source and French translations. The translations were then finalized by the lead French Language Coordinator (LC), a native of Canada. FACITtrans conducted a final review of all French translations prior to providing all documentation to the Department of Medical Social Sciences (MSS) at Northwestern University (NU), Feinberg School of Medicine which conducted a quality review of the entire translation process and approved the French translations for linguistic validation. The translations were then formatted into the electronic layout and proofread by two translators, one from Canada and the other from France, working independently from one another. Reconciliation of the proofreading commentary followed resulting in translations which were ready for cognitive interviews during the linguistic validation phase. All phases of the translations’ development from forward translation to finalized translation after cognitive interviews is documented in the French Item Histories housed within FACITtrans’ document management system.

The interview guide was created using FACIT testing administration procedures with a debriefing script to provide interviewers with instructions and cues to administer to participants for the cognitive interviews. Three Canadian research team members from McGill, who were fluent in French, were trained on conducting the interviews and following a standardized process. Eighty members of the general population answered the French version of the paper questionnaire and then were cognitively interviewed using a script developed by FACITtrans and the Department of Medical Social Sciences (MSS) at Northwestern University’s (NU) Feinberg School of Medicine. Participant comments and critiques were used to further refine the translations. Changes made as a result of the interviews were proofread for accuracy. Recruitment and cognitive interviews were carried out by the Canadian PNC at McGill University, located in Montreal, Quebec, the second largest French-speaking city in the world after Paris, France.

Pre-final versions of the French translations were completed by a small sample of French-speaking community participants in Canada (n = 80). Respondents country of birth included Algeria (n = 1), Bulgaria (n = 1), Canada (n = 45), France (n = 1), Haiti (n = 1), India (n = 1), Mauritius (n = 21), Mexico (n = 2), Morrocco (n = 3), Romania (n = 1), Venezuela (n = 1). Two participants did not report their country of birth during the cogntive interview. The participants then took part in individual cognitive interviews conducted in French to assess the relevance, understandability, and cultural appropriateness of the translations.

A qualitative analysis of cognitive interviews for each translated item was used to identify conceptual and linguistic differences between cultures. The synthesis included documenting when during the translation processes the potential issues were found, the frequency with which they occurred, and the manner in which the issue was resolved.