This is the first study to culturally adapt the BBY-Y and determined its psychometric properties. The study provided the translation, cultural adaptation and psychometric properties of the Yoruba version of the BBQ among patients with chronic LBP, following the Guillemin criteria. The BBQ had good psychometric properties, comparable to other versions of the BBQ. The BBQ-Y showed excellent test-retest reliability (0.89). The findings of this study conform to the recommendation of an ICC of 0.75 or more, regarded as excellent in the literature [25]. The ICC coefficient of the BBQ-Y is higher than that reported in previous studies including the Modern Standard Arabic (0.80) [15], traditional Chinese (0.85) [29], original English (0.87) [10], the Simplified Chinese (0.88) [16] and the Arabic (0.88) [17] versions. Alongside the ICC, we explored the limits of agreement between test and retest data. ICC is commonly used to show proportion of the variability in the new method due to the ‘normal’ variability between individuals. One shortcoming of ICC is that it might indicate strong correlation between two measurements with minimal agreement. Bland-Altman accounts for this shortcoming by revealing both systematic and random errors during a test-retest analysis [30]. The results of this study on the limits of agreement indicates that the small measurement error was equally spread across the whole scale range.
The SEM and MDC of 2.3 and 6.4 obtained in this study suggest good clinical utility of the BBQ-Y. Changes in scores of two points and more on the BBQ have been reported to be clinically significant [31, 32]. However, a minimally clinically important difference (MCID) that takes into cognisance patients’ self-reported improvement and SEM has not been calculated for the BBQ in previous studies. Between-known groups mean is another alternative when an MCID is not obtainable [33]. A between-known groups mean difference of 20 points on BBQ scores among individuals off work because of LBP and those still working has been reported [10]. As such, in the absence of MCID, between-known groups mean difference could be used to interpret the observed limits of agreement. Therefore, the observed limits of agreement of the BBQ-Y are acceptable because they are smaller than the differences between “known groups”. Furthermore, MDC can serve as an indicator of true change in an individual. For instance, a difference in scores which exceeds that of the MDC upon two successive measurements can be viewed as a true change [34]. Thus, clinicians can use the MDC and SEM values of the BBQ-Y obtained in the current study to interpret the efficacy of interventions targeted at mal-adaptive back pain beliefs.
The BBQ-Y showed good internal consistency, as shown by the Cronbach’s coefficient of 0.7. Although slightly lower, the Cronbach’s coefficient of the BBQ-Y was comparable to those found in other translations [10, 15, 35, 36]. While our study focussed on persons with chronic LBP, others studies included patients with and without LBP [10, 16, 35, 36]. The results of this study further showed that removing items 1 (i.e. There is no real treatment for back trouble) and 13 (i.e. Back trouble must be rested) did not significantly change the homogeneity of the items within the scale. This is similar to the findings by Maki and colleagues [15] where BBQ still kept higher internal consistencies without including items 1 and 13. However, the authors in this present study, do canvas for the retention of the two items, as all nine items still contributed to the overall excellent internal consistency of the BBQ-Y.
Factor analysis of the BBQ-Y revealed a three-factor structure with 6 of the scoring items loading on one factor. The other 3 scoring items alongside the distractor items loaded on the remaining two factors. This is contrasting to the two-factor structure original version where all the scoring items loaded on one factor and the distractor items loading on the other factor. Most of the studies that conducted EFA on the BBQ-Y reported a three-factor structure [17, 35, 36]. Item 1 did not load along with the scoring items in the present study as well as the previous studies [17, 35, 36] that conducted an EFA on the BBQ. This item has been argued to have ambiguous meaning or measuring a different construct as the other items included in the final score of the BBQ [17, 35]. It is therefore likely that this item has a different connotation to the Yoruba speaking patient with LBP compared to the original validation population. While all other scoring items loaded into the same factor in the previous translations, items 12 and 13 did not for the BBQ-Y. Deleting Items 1 and 13, however, from the BBQ-Y will not significantly reduce the internal consistency of the instrument. However, to keep the original structure of the BBQ, items 1 and 13 may be retained. On the other hand, deleting item 12 would reduce the internal consistency of the BBQ-Y. It is therefore important that the item be retained in the total BBQ-Y score. The distractor item “Surgery is the most effective way to treat back trouble” loaded with the scoring items. Distractor items, including item 7, are often omitted from the final score of the BBQ irrespective of the factor loading, to reduce the time and cost of data collection [35]. Considering the two factor structure, the loading was less interpretable as both distractor and scoring items loaded in the two factors.
Chronic LBP beliefs and its interpretation are influenced by cultural upbringing, sociocultural environment, previous pain experiences and health literacy [37]. For instance, beliefs about LBP may be under-reported in the African settings due to cultural or religious reasons. Although, patient beliefs and other psychosocial factors influences LBP in comparable ways both in the developed and developing countries [38], culture can mediate how these beliefs and psychosocial factors are expressed. Thus, the difference in the factor structures of the different versions of the BBQ may be adduced to the differences in culture.
Based on difficulty and quality rating, the Yoruba version of the BBQ had a high rate of data completion with good quality data in the study population. The BBQ-Y achieved a response rate of 100% and did not have any ceiling or floor effect. We did not report non-responses to questionnaire items as such event was rare owing to the method of administration of the questionnaires, as the researchers hand-delivered the questionnaires and patiently waited to collect them. This method is still the preferred approach in many sub-Saharan African countries as the postal and electronic methods often fail due to technical and infrastructural challenges. However, we consider not having the details of non-response a shortcoming. The above findings suggest that the Yoruba version of the BBQ is an acceptable outcome measure for assessing back pain beliefs among the Yoruba-Speaking LBP patients. There was no significant structural alteration made to the BBQ-Y, other than required cultural-adaptation.
In summary, the moderate to excellent psychometric properties of the BBQ-Y lend credence to its usability and applicability in the clinic setting among patients with chronic LBP. The new tool may promote assessment of beliefs about chronic pain and also inform interventions to improve health outcomes of Yoruba speaking patients with chronic LBP. Anecdotal and empirical reports indicate that African patients with chronic pain have psychosocial problems, in this case, unhelpful beliefs, that have strong cultural undercurrents which in turn affect their health seeking behaviour [39, 40]. Early identification of patients with unhelpful beliefs about LBP is of utmost importance. This will help to focus interventions on those maladaptive beliefs rather than approaching the LBP problem from a biomedical perspective. Interventions incorporating the bio-psychosocial framework are recommended as the most effective approach in managing LBP [41]. Unlike the biomedical approach, bio-psychosocial interventions not only treat impaired anatomical structures but may also address psychological, socioeconomic, ecological and cultural factors that may impact on the onset and persistence of LBP.; hence, the importance of such translation as the BBQ-Y.
Limitations and methodological problems
Limitations of the study include the cross-sectional nature of the study, which did not allow for the examination of other aspects of validity, including sensitivity and responsiveness. Second, the authors conducted the study within the clinical settings, thus may not totally represent individuals with LBP in Nigeria. Also, we did not investigate the construct validity of the BBQ-Y with other measures commonly used in the literature, including the fear-avoidance beliefs questionnaire and the Tampa scale of kinesiophobia. Furthermore, we did not conduct confirmatory factor analysis in this study, as a large sample size will be required for it.
Implications for future research
Future research on the BBQ-Y using larger and more diverse sample is required. In addition, further psychometric testing on convergent validity and factor structure are recommended for future studies.