Linguistic Validation work is usually carried out by commercial LV companies. The process usually takes between 18 - 20 weeks, depending on the language. In exceptional circumstances translations may be done elsewhere, but a rigorous protocol must be adhered to in order to maintain the standards required.
The following process is important in order to ensure a high standard of translation into a new “target” language and its linguistic validation. The comments and discussions at the various stages are needed for our records of the translation history, so that we can refer to reasons why certain decisions and choices were made. This record is especially important if subsequent psychometric evaluation suggests that improvements to the translation are needed.
The linguistic validation process is divided into 3 main phases:
Phase 1 is the main forward and back translation stage
Phase 2 is the piloting stage
• Review by psychologist and/or clinician
• Cognitive Debriefing with patients
Phase 3 is the finalisation stage
When working with Mapi on translations, as shown above, we would normally review the translations after:
• the back-translation step
• any subsequent retranslation steps
• clinician / psychologist review
• cognitive debriefing
Phase 1 is not needed when adapting a language for use in another country or community. However, the existing language version must be reviewed first by someone who is a native speaker of the required language and resident in the relevant country. They should also be fluent in English. For English adaptations (e.g. for Australia, India) it should be sufficient for the person wishing to use the questionnaire to carry out this review, as long as English is their first or principal language (e.g. in India, someone who has been educated throughout in English).
Phases 2 and 3 are still required.
NOTE: Languages which require reformatting of the questionnaires
For languages which are written other than from left to right, LV teams must include a member who can reformat the questionnaires, in Word, to the required standard of HPR. Target language versions of the questionnaires are required from the FT stage until formatting of the final versions.
Phase 1 – Forward and Back Translation
Forward Translation (FT)
1.1) Two parallel forward translations by native speakers of the target language, fluent in English and resident in the country where the translation is to be used. Ideally one person should be a professional translator and the other a psychologist with experience of designing psychological questionnaire measures. Familiarity with the subject of the questionnaire would be an advantage.
1.2) Review of discrepancies and reconciliation of the two forward translations into 1st intermediate forward translation by a third person, who is a native speaker of the target language and fluent in English.
1.3) Notes on difficulties found, decisions made and reasons for those decisions during Step 2 for the forward translation report.
Back Translation (BT)
1.4) Two parallel back-translations of the intermediate forward translation, done completely “blind” (i.e. with no help or information about the original English questionnaires) and independently of each other. These should be done by people who are native speakers of English and fluent in the target language.
1.5) Review and discussion of discrepancies with back-translators.
1.6) Notes on issues raised during Step 1.5 for back-translation report.
1.7) Review of back-translation report by a member of HPR, who highlights any concerns.
Any revisions to Forward Translation
1.8) Review by forward translators / translation co-ordinator of any concerns or discrepancies noted during the BT. Retranslation where necessary.
1.9) Notes on decisions made in revisions, together with notes on any items needing particular attention and possible testing of alternatives during clinician review and cognitive debriefing.
1.10) Review of revisions and notes by a member of HPR.
Phase 2 – Piloting
Review by psychologist and/or clinician
2.1) By psychologist:
This step may be particularly useful if the questionnaire is a generic one (e.g. the WBQ12) and there is no reason for a clinician review. Where one of the forward translators is a psychologist (particularly one with experience of questionnaire design) this step would not be necessary
2.2) By clinician:
Where the questionnaire is disorder-specific, review of the revised intermediate translation by a clinician practising in the relevant field (e.g. diabetes, renal disease, ophthalmology) is important. In particular, review of any phrases relating to the nature of the treatment or name of the disorder or associated symptoms. It is important to identify the name the clinician gives to the disorder when talking to patients (i.e. not necessarily the terms used between health professionals). If one of the forward translators is just such a clinician, this step would not be necessary. If the clinician is simply providing a brief review of the questionnaire and not acting as one of the translators themselves, s/he may not wish to spend time studying the concept translation guidelines, but these should nevertheless be made available to him/her along with the translation, in case s/he has questions about an item. Many questions can be answered by these documents.
2.3) Notes on suggestions and decisions resulting from both the above steps and notes on any items needing particular attention (or testing of alternatives) during cognitive debriefing need to be included in the psychologist/clinician report.
2.4) Review by the HPR and discussion of any proposed revisions.
Phase 2 - Piloting
2CD.1) Interviews of approximately five patients with as good a mix as possible, e.g. for the DTSQs & c, ADDQoL, ADKnowl we would aim for the following:
- a) a minimum of 5 people
- b) all with diabetes
- made up of a mixture of the following:
- c) men and women
- d) across a good age range
- e) With Type 1 and 2 diabetes
- f) the Type 2 patients would ideally include someone on diet alone, someone on tablets and someone on insulin.
2CD.2) Report by person conducting the interviews using the CD report template provided, and listing:
- a) demographic and known clinical / treatment details of interviewees (these may have a bearing on any notes relating to (b) below);
- b) account of patients’ understanding of all items, even where the interviewer doesn’t think there is a problem. If there are any problematic items, a note of which patient(s) found it problematic and in what way.
- c) selected scores for any of these patients where particularly relevant (e.g. on the DTSQ, if the patients’ scores appeared to contradict what they were saying in the interview).
2CD.3) Review by translation co-ordinator:
- a) Review of any problematic words/phrases/items with forward translators following CD interviews.
- b) Retranslation and back-translation where necessary.
- c) Notes on changes, plus explanations to be given on the CD report template.
2CD.4) Review by HPR and discussion of any proposed revisions.
Phase 3 - Finalisation
Please ensure you keep to the original format. We will give you a Word document of the original English that you can use as a template. As you will see from your agreement, you need to keep strictly to this template and not add extras into the header or make any changes to the style. This includes font and font size, unless you are producing e.g. Chinese and you have to produce a translation in a different script. If you need to add e.g. patient number and completion date somewhere, we suggest that you do as others have done and have a separate cover sheet in front of the questionnaire.
3.2) DTSQc – generic and trial specific versions
If you are translating the DTSQc, please ensure that you start by producing the generic version of the introduction. If you need to change the wording to be specific to your study, please produce this as a separate version and send both questionnaires to us on completion. Trial Specific Instructions need to be agreed with HPR before linguistic validation starts.
An important step sometimes forgotten! A new translation/adaptation needs to be proofread by at least one native speaker of the target language.
3.4) The questionnaire also needs to be proofread for formatting issues, using the original English format as a template and checked subsequently by HPR. HPR will finalise the footer and copyright statement.