Introduction

Low back pain (LBP) is undoubtedly the most common musculoskeletal disorder in the world [1, 2]. The average point prevalence of LBP is estimated at 18.3% and the one-month prevalence at 30.8% in the general population [3–5]. It is also estimated that 40–90% of adults will develop LBP during their lifetime [6, 7]. However, the estimation of the prevalence of LBP meets a major methodological limitation in the literature, due to the heterogeneity in the definition of cases [3, 8]: this makes it difficult to make valid comparisons and syntheses of frequency estimates. This difficulty can be observed in particular in systematic reviews on LBP [3, 4, 9]. To find solutions to this problem, 28 researchers from 12 countries worked in 2005–2006 to develop consensus definitions of LBP. As a result of this work, a “minimal” and an “optimal” definition of LBP were adopted by consensus and published [10]. These definitions were developed using the Delphi method [11–13] and were called “Delphi definitions of low back pain prevalence (Delphi DOLBaPP)”. The “minimal” definition includes a question on the characteristics of pain (anatomical site, symptoms, and time of onset of symptoms) and a question on related functional limitations, and is recommended for use in large surveys where space and time are limited. The “optimal” definition includes five additional questions on frequency, duration of symptoms, intensity of pain and sciatica [10] and is more appropriate for studies specifically dealing with LBP. In order to enable practical use of these definitions in different contexts, their presentation was adapted in the form of questionnaires that could be self-administered or administered by an interviewer (referred to as the “Delphi DOLBaPP questionnaires”). These standardized definitions and the derived questionnaires have been developed in English and are suitable for use with English-speaking populations. However, the use of these definitions in languages other than English requires them to be culturally adapted, following a rigorous process, to ensure that their translation will be equivalent to the original version both in its linguistic and cultural dimensions [14]. Such adaptations of the Delphi DOLBaPP definitions have been achieved in Spanish [15] and German [16] and several adaptations in other languages are in progress. However, to our knowledge, no cross-cultural adaptation of the Delphi DOLBaPP definitions in French has yet been made. The aim of this article is to present the process of cross-cultural adaptation of the Delphi DOLBaPP definitions in Quebec French and their pre-test among French-speaking adults in Quebec.

Materials and methods

The cross-cultural adaptation of the Delphi DOLBaPP questionnaires in French was conducted according to the “backward translation approach”, which remains the most common and is considered the preferred approach by many researchers [17–19]. The method and guidelines provided in the “Guidelines for the process of cross-cultural adaptation of self-report measures” proposed by Beaton et al. [14], that were specifically adapted for the Delphi DOLBaPP definitions and questionnaires, were used. The following steps were followed:

Step 1: Initially, three translators who had French for mother tongue independently translated the Delphi DOLBaPP optimal questionnaire from English into French. The three translators had various experiences in translation. Two of them were student translators, who received formal supervision by professional translators. The third translator was a research professional without formal training in translation but with professional experience in health research and in translation of health-related scientific material and questionnaires.

Step 2: The study coordinator (SP) compared the individual French translations and prepared a synthesis version; the discrepancies identified between the different translations were discussed with the first author of the definitions (CED) [10]. At the end of this exercise, new proposals for the translation of the problematic elements were prepared and submitted to the three translators. Their comments enriched the discussions and led to a consensual French version. In case of disagreement between the translators, the final decision was taken by the main author (CED) of the DOLBaPP on the proposal of the coordinator of the study (SP).

Step 3: Two other translators, whose mother tongue was English, translated this French version back into English. They proceeded independently and without reference to the original English version of the questionnaire. The two backward translators were professional translators. The first one was a freelancer and the second one was working for a translation firm; English was their native tongue.

Step 4: It was at this stage that an expert committee, made up of the main authors of the original English version of the DOLBaPP (CED, KD, PRC) [10], musculoskeletal epidemiology specialists, and the coordinator of the cultural adaptation in French (SP), was formed to resolve all problematic translation issues in a consensual manner. In case of disagreement between the translators, the final decision was taken by CED on the proposal of SP. A pre-final version of the questionnaire in French was produced at the end of this step.

Step 5: In order to pre-test the French version of the questionnaire, it was self-administered to adults who were recruited among participants in the third data collection phase of the PROspective Quebec Study on Work and Health [20], a 24-year cohort study conducted among 9,188 white collar workers aged 18 years or older in the Quebec City area (“the cohort study”). In this step, according to the method proposed by Beaton et al. [14], the adaptation should be applied to a sample of at least 30 to 40 participants. In the current study, since access was provided to a large group of the cohort study, a larger convenience sample was selected. During five weeks, the first two participants of the cohort study who came to the Centre de recherche du CHU de Quebec-Université Laval for their follow-up meeting in the morning, and the first two who came in the afternoon were systematically selected for our study. They were asked to complete the Delphi DOLBaPP questionnaire and an evaluation form which concerned i) the time required to complete the questionnaire, ii) the wording of the questions, iii) the illustration used in the questionnaire, iv) the structure of the questionnaire and v) the length of the questionnaire. A research professional informed the participants about this additional task and explained it to them. If any selected morning or afternoon participant was over 70 years of age or had cognitive impairment (whether self-reported or according to his or her companion), he or she was replaced by the person appearing next in the appointment list of the cohort study. The participants were allowed to take the time they needed to complete the questionnaire and the evaluation form. The age criterion was applied so as not to increase the burden on the older participants, who generally needed more time to complete the full set of questionnaires of the cohort study. After they had completed the questionnaire and the evaluation form, the participants’ comments on certain issues, words or phrases were also collected in the evaluation form and documented.

Data were analysed using SAS software version 9.4. The distribution of participants’ characteristics and their responses to the questionnaire on LBP and to the evaluation form were described.

Qualitative analyses of the answers to the three open-ended questions in the evaluation form were carried out; a thematic content analysis was conducted on the basis of these answers [21]. The qualitative data focused on i) the assessment of the wording of the questions, ii) the conditions for completing the questionnaire, and iii) comments or suggestions for improving the questionnaire. Responses were codified and categories were constructed. For each category identified, its distribution was analyzed among the respondents.

Results

The result of the French translation of the questionnaire is shown in Table 1. The third column illustrates, in bold type, the changes that were made to the original English version of the questionnaire as a result of the standardized adaptation process. In questions one, three, five, six and seven, the words “in the past four weeks” were translated into French as “au cours des quatre dernières semaines”, and then back-translated into English as “in the last four weeks”. When asked about their choice, the English-speaking translators indicated that there was a slight difference in the meaning of “past” and “last” in English, and that the latter term was considered to correspond to the concept of “the latest, most recent”, which they felt better matched the time reference specified in the question. The expert committee therefore decided to modify the original English version of the questionnaire by replacing “past” with “last”. Taking into account the remarks of the French-speaking translators, who found that the words “report” and “feverish illness” (exclusions mentioned in question 1) were somewhat difficult to understand and translate into French, a proposal was made to reformulate “Please do not report pain from feverish illness or menstruation” as “Please do not take into account pain caused by menstruation or by an illness accompanied by fever”. The two English-speaking translators also recommended the use of “lower” rather than “low” in the expression “pain in your lower back” (questions one and five), as “lower” sounds more idiomatic in English. These two questions of the original English version were therefore amended accordingly. In question four, the English translators found that the expression “go below” was more familiar than “spread below” to describe a pain that goes down the length of the leg, and that it would be easier for English-speaking respondents to understand. In addition, they also felt that “go below” was more consistent with the previous question (“Have you had pain that goes down the leg?”).

Discussion

The use of reliable and valid measures in research is essential to provide evidence-based health care [22]. The standardized adaptation process used in the study reported here made it possible to adapt the Delphi DOLBaPP questionnaire specifically for francophones, while preserving the content and structure of the original version. The fact that the process included a backward translation step ensured that the meaning of the items was maintained between the English and French versions. Moreover, the standardized adaptation process of the Delphi DOLBaPP questionnaire into French resulted in minor changes following the backward translation. While these minor changes did not alter the meaning of the questions in the original version, the proposed rewording of certain words or phrases not only improved the translation and understanding of the French questionnaire, but also contributed to improving the original instrument. This is the case, for example, for “Please do not take into account pain caused by menstruation or by an illness accompanied by fever”, which was included in the English original version after the back-translation step because of comments made by the French translators about the difficulty to understand and translate this sentence.

The results of this study are in line with the German adaptation of the DOLBaPP questionnaire, which also resulted in only minor modifications [16]. Some problems of conceptual equivalency were common to both the French and German translations, namely about the expression “having pain in your lower back”. However, the reactions of participants were not always the same in both languages. During the French translation process, it was suggested that “go below the knee” (“descendre au bas du genou”) replace “spread below the knee” in the original version, as this term was found to be the best match for the French translation. In the same version, the German translators suggested using the word “radiate” instead of “spread” or “go” for the same expression, as this word best represented the original construct when translated into German [16]. The question about duration of low back pain was found difficult to understand by some participants, both French and German. However, the French translators proposed a minor change to this question, with the exception of a modification to the last response choice “three years and more”, which was replaced by “three years or more” (“trois ans ou plus”), while German translators proposed a change to make the question easier to understand: replacing the phrase “how long was it since you had” with “when was the last time that you had” [16]. This kind of exercise (cultural adaptation between two languages) usually presents these types of translation difficulties encountered in previous studies [23–26].

The overall assessment of the Delphi DOLBaPP questionnaire by the participants was rather positive. Overall, more than 62% of the participants did not find any poorly worded or confusing questions. Many had an excellent appreciation of the diagram illustrating the lumbar region included in the questionnaire. While Leonhardt et al. [16] found that 95% of the respondents took less than five minutes to complete the questionnaire, in our study this proportion was 89.0%. The difference between these two results could perhaps be explained by the fact that the participants in Leonhardt et al.’s study were younger than those in our sample. In addition, some participants who had a reserved or negative assessment of certain questions, such as the question on the frequency of low back pain, may have had a problem of understanding the whole questionnaire. These participants wanted to integrate filters into the questionnaire to facilitate their understanding.

The prevalence estimates of LBP in the German and French versions are very similar. In both studies, they reached about 50% of the sample. The fact that these two prevalence estimates are comparable may be due to the use of the standardised definitions of LBP in the two studies. Other studies that have assessed the prevalence of LBP in white-collar workers using other definitions showed lower or slightly higher prevalence than the one that was found in the current study [27, 28], hence it is important to continue to promote the Delphi DOLBaPP definitions, as well as their cultural adaptation in other languages, and to work on their continuous improvement.

The choice of asking for three translations of the definitions from English to French instead of two represents an advantage in that it enabled to achieve a better consensus when necessary. Proceeding with a backward translation approach and the large number of participants who pre-tested the questionnaire in this study are strengths for the cultural adaptation method that was retained for this study.

For this pre-test, a representative sample was not chosen because it is not a recommendation for this kind of exercise. However, a possible limitation of this study is the fact that the participants were younger than the general population of LBP patients, which may have led to an overestimation of the understanding of the questionnaire among people aged 70 and over. In addition, white-collar workers are less at risk of LBP than blue-collar workers, and retired white-collar workers are not necessarily representative of the general population of LBP patients.

This study ended up in the successful production of a valuable French translation of the Delphi DOLBaPP questionnaire adapted for the Quebec cultural context. This statement is based on the results of the pre-test with the sample. This French version of the questionnaire can now be used, with eventual minor adjustments if judged necessary, in other French-speaking countries such as France, Belgium, Switzerland, Luxembourg and French-speaking African countries. Populations from these countries will then be able to contribute to the standardization efforts that this questionnaire allows in order to improve the quality of LBP studies, as well as to share the benefits that derive from its use. Future cross-cultural adaptation of the Delphi DOLBaPP in other languages is needed to improve a better use of these standardized definitions.

The cross-cultural adaptation procedure used in this study is based on the Guide for the intercultural adaptation of the standardized definitions of Delphi low back pain, which itself takes up on the method proposed by Beaton et al. [14]. This specific guide, which is available upon request, proposes the application of a standardized translation method and provides all the necessary information for adapting the Delphi DOLBaPP definitions from English into another language.