Background
Evaluation of patients with musculoskeletal disorders can rely not only on clinical examination and radiological imaging, but also on scores from self-reported outcome instruments. The information acquired from self-reported outcome instruments is useful only if there is evidence to support interpretation of the obtained scores[
1,
2]. Currently there is a lack of evidence to support the use of an instrument in the French language for individuals with musculoskeletal foot and ankle disorders.
In the foot and ankle literature there are many instruments that have been implemented in outcome related research with little or no evidence to support their use[
3,
4]. Over 49 instruments have been identified[
3] with 14 having some evidence to support their use[
4]. Over these 14 instruments, the 4 clinimetric qualities (content validity, construct validity, reliability, and responsiveness) were only reached by the Foot and Ankle Ability Measure, the Foot Function index, the Foot Health Status Questionnaire, the Lower Extremity Function Scale, and the Sports Ankle Rating System quality of life measure[
4]. Instruments that offer specific information for score interpretation may be more useful. This information can include score error associated with single measure, change in score that represents a minimal detectable change (MDC), and a change in score that represents minimal clinical important difference (MCID). The MDC quantifies the change in score value over time that is beyond measurement error. The MCID is a cut-off value over which changes discriminates between patients that have clinically improved from those that have not improved[
4,
5]. Even in instruments that have evidence to support their use, difficulties arise when they are applied to individuals who are not fluent in the language the instrument has been developed in. While there are many instruments available there is not a universally accepted instrument in the French language for individuals with musculoskeletal foot and ankle disorders.
The Foot and Ankle Ability Measure (FAAM), originally published in English[
6], has evidence for reliability, responsiveness, and validity as measure of physical function. This evidence was provided in individuals with a wide range of musculoskeletal disorders of the lower leg, ankle, and foot and therefore has broad application. To evaluate activities that require a higher level of ability the FAAM also contains a Sports subscale[
7]. Specific information for score interpretation, including MDC and MCID values are provided for the FAAM. Evidence to support its use of is also available for athletes with chronic ankle instability[
8] and patients with Diabetes Mellitus[
7]. Moreover, a cross cultural adaptation and validation has been completed to create German[
9] and Persian versions of the FAAM[
10].
Although there is increasing evidence to support the use of the FAAM, it has not been adapted and validated for French speaking individuals. In our clinical and research practice, most patients are French speaking, and not able to appropriately respond to items written in English. The aim of this study is to translate the FAAM into French, perform a cross-cultural adaptation, and provided evidence for validity and reliability for this French version of the FAAM (FAAM-F).
Results
The response proportion was 76% (105 responders). The average FAAM-F score was 74 (range 8-100 SD 22.1) and 44 (range 0-100 SD 31.0) for the ADL and Sports subscales, respectively (Table
1). A total of 7 (6.7%) and 5 (4.8%) patients scored 100 (best possible score) on the ADL and Sports subscales, respectively. No patient scored 0 (worst possible score) for the ADL subscale and 6 patients (5.7%) scored 0 for the Sports subscale. Twenty-four patients (22.9%) had scores above 95 for the ADL subscale and 7 (6.7%) for the Sports subscale. Percentage of patients having scores below 5 for ADL and Sports subscales were 0% and 11.4% (12 patients), respectively.
Cross cultural adaptation
The translation of the different items of the FAAM was carried out without any relevant difficulties. The back-translation of the T12 version in English was very similar to the original version. Only some typically American expressions were different as our back-translators were native from Great-Britain. During the phone interview with 40 patients for the pre-testing of the final French version, it appeared that the item "walking initially", translated first as "commencer à marcher", was understood in different ways by the patients. Therefore, this item was adapted into "faire les premiers pas (le matin au réveil/après une position assise prolongée)" in order to precise the question. Upon re-assessment with patients it was noted that this change improved item interpretation. No further difficulties in understanding the items were discovered during the pre-testing phase.
Evidence for Convergent and Divergent Validity
For the convergent validity, we found correlation coefficients of 0.85 between the ADL and Physical Functioning, and 0.81 between ADL and PCS. The Sports subscale had correlations of 0.72 with both Physical Function and PCS. The assessment of divergent validity resulted in an ADL-Mental Health correlation of 0.26 and an ADL-MCS correlation of 0.37. The Sports subscale had a correlation of 0.21 with Mental Health and of 0.29 with MCS (Table
2). The calculated t-values assessing for difference in the correlation coefficients between the ADL and Sports subscales to measures of physical and mental functioning were significant with P < 0.0005.
Table 2
Correlation coefficients between the FAAM-F (ADL and Sports subscales) and the SF-36 (physical and mental scales).
Physical Functioning | 0.85 (n = 104) | 0.72 (n = 102) | 0.84 (n = 151) | 0.78 (n = 130) | 0.61 (n = 83) | 0.63 (n = 83) |
Physical Component Summary | 0.81 n = 100) | 0.72 (n = 98) | 0.84 (n = 151) | 0.80 (n = 130) | 0.71 (n = 83) | 0.72 (n = 83) |
Mental Health | 0.26 (n = 102) | 0.21 (n = 100) | 0.18 (n = 151) | 0.11 (n = 130) | 0.32 (n = 83) | 0.22 (n = 83) |
Mental Component Summary | 0.37 (n = 100) | 0.29 (n = 98) | 0.05 (n = 151) | -0.02 (n = 130) | 0.29 (n = 83) | 0.17 (n = 83) |
Evidence of Internal Structure
The assessment of internal consistency found coefficient alpha to be 0.97 for both the ADL and Sports subscales. The score error associated with a single measure at 95% confidence was +/- 8 and +/- 10 points for the ADL and Sports subscale, respectively
Evidence for Score Stability
The assessment of test re-test reliability over a two day period found ICC values of 0.97 and 0.94 for the ADL and Sports subscales, respectively. The MDC at 95% confidence was +/- 7 and +/- 18 points for the ADL and Sports subscales, respectively
Discussion
The aims of this study were met as original English version of the FAAM was successfully translated and adapted into French to create the FAAM-F ADL and Sports subscales. Evidence for reliability and validity were offered to support the use of the FAAM-F as a self-report outcome instrument for individuals with a wide range of chronic foot and ankle disorders. Specifically, evidence for convergent validity, divergent validity, internal structure, and score stability were provided for the FAAM-F. Information for score interpretation with error associated with a single measure and MDC was also provided for the FAAM-F.
There is a lack of well designed, validated, questionnaire in French for assessing foot and ankle disorders [
3]. Moreover, there is increasing evidence that self-questionnaires are useful for clinicians as well as for investigators in order to facilitate the patients' self evaluation. The FAAM is region specific measure, capable of being sensitive to changes in physical function. It is easy for patient to complete and uncomplicated for clinicians score, making it easy to add to the evaluation process. A systematic review concluded the FAAM and its predecessor were the most appropriate evaluative instruments to quantify functional disabilities in individuals with chronic ankle instability[
17]. The FAAM has undergone advanced psychometric testing including the use of item response theory (IRT)[
18]. The evidence to support the use of the FAAM as an outcome instrument continues to grow not only for the English version but for German and Persian versions as well. Therefore, it was felt the FAAM would be the most appropriate instrument to translate and adapt for French speaking individuals. Following the guidelines of the AAOS, the French adaptation of this questionnaire was performed without encountering any difficulties as only one of the 29 items had to be slightly modified for easier understanding (item related to "walking initially"). Moreover, the high response rate and the good feed-back during the phone interview demonstrated that the FAAM-F was well accepted.
The relationship between the FAAM-F and concurrent measures of physical function were significantly different than the associations between the FAAM-F and concurrent measures of mental health. This provides evidence that the FAAM-F is a measure of physical function as opposed to mental function. The coefficients of correlation between FAAM-F subscales and the SF-36 physical scales were high (0.85 and 0.72.). Moreover, they are very similar to the values found for the original version[
6] (0.84 and 0.78; Table
2). The correlations between FAAM subscales and the SF-36 mental scales for divergent validity were low, as hypothesized and were similar to those for the original version with a good overlap of their 95% confidence intervals. The possible exception was noted for the ADL to MSC, 0.37 and 0.05 respectively for the French and in English versions (Table
2). This could be explained by some cultural difference regarding pain or type of foot disorders between the two studies. Our patients were recruited in tertiary centers and suffered for chronic disorders (months up to years), which can obviously affect mental function and self-perception. This explanation is also supported by the comparison of our results with the ones described in the publication with patients presenting with chronic foot complaints related to diabetes mellitus[
7]. However, the correlation between ADL to MSC was low in our study, supporting our hypotheses for divergent validity.
The calculated MDC, for FAAM-F ADL and Sports subscales, differ from the values reported for the English version of the FAAM (ADL MDC 5,7 and 12,3; Sports MDC 7 and 18 for French and English versions, respectively). These discrepencies could be explained either by differences in patients population or by the time frame over which the MDC was calculated. The original study used a 4-week time period while the current study used a 2-day time period.
According to the literature, the CA value for a good internal consistency should be over 0.80 (over 0.90 for clinical applications)[
19]. Despite the adaptation of one of the 21 items during cross-cultural-adaptation the internal consistency of the FAAM-F remained excellent and widely above the limit of acceptance with a CA value of 0.97 for ADL subscale. The Sports subscale as well showed an excellent internal consistency (0.97). These two values were almost identical to the values provided for the English version (0.98). Values noted for the error associated with a single measure, 8 and 10 points for the ADL and Sports subscales were also very similar to the English version, 7 and 10 points, respectively. The test re-test reliability of the FAAM-F was found to be excellent. The ICC values of this study were superior to those reported for the original version. This finding might be expected as the time interval in the original study was over a 4 week period compared to 2 days in the current study. The findings of this study help to confirm that the French translation is close to the original version[
6].
When looking at the range of obtained scores, there were FAAM-F ADL and Sports scores with extreme high and low values. This could have potential problem with floor and ceiling effects if more than 15-20% of patients have scores in these values[
2,
5]. Twenty-four patients (22.9%) had scores above 95 for the ADL subscale and 7 (6.7%) for the Sports subscale. Therefore, a possible ceiling effect for the ADL subscale could be potentially problematic if used without the Sports subscale in individuals functioning at a high level. However, the ADL and Sports subscale were developed to be used together and complement one another. Therefore ceiling effects with the ADL subscale would be offset by the use of the Sports subscale which would collect information for those functioning at a high level of ability.
The limitations of this study must be recognized. Only patients with chronic foot disorders have been included in the present study. Patients with acute conditions may present differently. Furthermore, the responsiveness of the FAAM-F needs to be assessed with values for MCID offered to assist with score interpretation. It should be remembered that evidence to support the use of an instrument is an ongoing process with information needed under various study conditions. Furthermore, the enrolled patients were not randomly drawn from a population with chronic food and ankle disorders but were a convenience sample of available hospital patients.
Acknowledgements
The authors thank Mike Short, Trevor Mc Intosh, Christine Favre and Aude Anvari for the translation work, and also Viviane Dufour for her logistical help.
The French version of the FAAM is available at francois.luthi@crr-suva.ch
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
BS, CX, DO, LF and BP have designed the study. BS, CX and LF have gathered the data. BS, DO, BP, MRL and LF have analyzed the data. BS and MRL have written the initial draft. BS, CX, DO, BP, MRL and LF have ensured the accuracy of the data and analysis.
All authors read and approved the final manuscript.