Study designs
Translation and cultural adaptation, test–retest measurements, and cross-sectional study of the psychometric properties of the CSQ were performed among Igbo populations with chronic low back pain living in rural and urban settings in Nigeria.
Ethical issues
Ethical approvals were obtained from King’s College London (Ref: BDM/13/14-99) and University of Nigeria Teaching Hospital (Ref: UNTH/CSA/329/Vol.5). Written permission was obtained from the original developers of the questionnaire. Informed consent was sought and obtained from all the participants involved in this study. Participants were attended to in their homes and workplaces and were not given remuneration for participating in the study.
Outcome measures
Coping Strategies Questionnaire (CSQ)
CSQ was developed to assess cognitive and behavioural coping strategies for dealing with pain [23]. Further validation of the CSQ following initial development by the original developers of the measure produced the 42-item version which was obtained by removing the eighth subscale ‘increasing pain behaviours’. The original authors found that the eight subscale had an unacceptable level of internal consistency and recommended the use of the 42-item version of the CSQ as the standard CSQ [23]. CSQ consists of seven subscales with six items each: (diverting attention [items 3, 9, 12, 26, 27, 38], reinterpreting pain sensation [items 1, 4, 10, 16, 29, 41], catastrophising [items 5, 11, 13, 25, 33, 37], ignoring pain sensations [items 17, 19, 21, 24, 30, 35], praying or hoping [items 14, 15, 18, 22, 28, 36], coping self-statements [items 6, 8, 20, 23, 31, 32] and increased behavioural activities [items 2, 7, 34, 39, 40, 42]). Each item has a numeric rating scale ranging from 0 (never do that) to 6 (always do that). Hence each subscale has a maximum score of 36 and a minimum score of 0. A higher score indicates greater use of a particular coping strategy. Additional two items assess overall effectiveness of pain control and ability to decrease pain. The internal consistencies of the subscales range between 0.71 and 0.85 [23].
Eleven-point box scale (BS-11)
The BS-11 is an eleven-point numeric scale for pain intensity [24]. It consists of eleven numbers (0 through 10) surrounded by boxes. Zero represents ‘no pain’ and 10 represents ‘pain as bad as you can imagine’ or ‘worst pain imaginable’. It is easy to comprehend and administer, with high test–retest reliability in both literate and illiterate patients with rheumatoid arthritis (ICC = 0.96 and 0.95, respectively). BS-11 has high correlations (0.86–0.95) with the visual analogue scale (VAS) in patients with rheumatic and other chronic pain conditions; and a reduction of 2 points is clinically significant [24].
Igbo Roland Morris Disability Questionnaire (Igbo-RMDQ)
The RMDQ is simple to administer, easily understood, and is most suitable for population-based studies [25]. The Igbo-RMDQ [26], adapted from the original version [27], is a 24-item back specific self-report measure. Each item has possible scores of 0 or 1. A total maximum score of 24 signifies the highest disability and 0 denotes no disability. The Igbo-RMDQ has good face and content validity, construct validity (moderately high correlations [r > 0.6] with performance-based disability and pain intensity), internal consistency (α = 0.84), test–retest reliability (intraclass correlation coefficient = 0.80) [26], and responsiveness (2–3-point change from baseline is considered clinically important) [25].
Igbo World Health Organisation Disability Assessment Schedule (Igbo-WHODAS 2.0)
The Igbo-WHODAS 2.0 is a 36-item interviewer administered questionnaire that assesses six domains of disability. These include cognition (understanding and communicating), mobility (getting around), self-care (taking care of oneself), getting along with people (good relationship with people), life activities (maintaining an individual’s household or work/school activities) and participation (participating in society and the impact of the health problem on them and their family). Difficulties encountered are measured within the last 30 days. The measure has good face and content validity, construct validity, internal consistency, test–retest reliability and responsiveness. The complex scoring method considers multiple levels of difficulty for each item. It involves summing recoded item scores in each domain, summing all six domain scores, and converting the total score into a value that range from 0 (no disability) to 100 (maximum disability) [28].
Igbo fear avoidance beliefs questionnaire (Igbo-FABQ)
The Igbo-FABQ is a sixteen-item back pain-specific self-report tool that measures the level to which pain is believed to be caused or aggravated by general physical activity (FABQ-PA) and work-related activities (FABQ-W) [29]. The two subscale scores give a total score of 66. Greater scores reflect more fear avoidance beliefs [30]. The physical activity subscale (FABQ-PA) has five items, each with a score ranging from 0 (completely disagree) to 6 (completely agree). Item 1 is a distractor and is not scored. The maximum score for FABQ-PA is 24 and the minimum is 0, with higher scores indicating stronger fear avoidance beliefs related to physical activity. FABQ-W has 11 items, each with a score ranging from 0 (completely disagree) to 6 (completely agree). Items 8, 13, 14, 16 are distractors, and do not contribute to total score. The maximum score for FABQ-W is 42 and minimum score is 0 with higher scores indicating stronger fear avoidance beliefs related to work activities. Igbo-FABQ has good face and content validity, construct validity, internal consistency, test–retest reliability and responsiveness [29]. A change of 13 from baseline is clinically significant [31].
Igbo Brief Illness Perceptions Questionnaire (Igbo-BIPQ)
The Igbo-BIPQ is a self-report measure of cognitive and emotional illness perceptions [32] adapted from the original English version [33] with eight items (consequences, timeline, personal control, treatment control, identity, illness concern, coherence and emotional representation), each of which assesses one dimension of illness perceptions. There is an incremental ten-point scale in each item, anchored at 0 and 10 depicting minimal and maximal level of the assessed dimension. The eight items may be combined as one total score, or each item may be assessed separately to give eight dimensions of illness perceptions [33]. Eighty and 0 are the maximum and minimum total scores. A higher score signals a more threatening view of an illness [33]. The ninth item is open and is the causal item. Igbo-BIPQ has good face and content validity, construct validity, internal consistency, test–retest reliability and responsiveness [32].
Cross-cultural adaptation
Participants
A clinical musculoskeletal physiotherapist (bilingual in English and Igbo, native Igbo speaker) who had been practicing in Nigeria for 18 years; and three non-clinical translators (two native English speakers and one native Igbo speaker; all bilingual in English and Igbo) were involved in the cross-cultural adaptation. Two of the non-clinical translators (one native English speaker and one native Igbo speaker) were linguistic experts. Two English experts (health psychologist and academic physiotherapist) in the United Kingdom, and two Igbo experts (clinical psychologist and clinical physiotherapist) in Nigeria made up the expert review committee.
Twelve adults recruited conveniently from a rural population in Enugu State pre-tested/piloted the adapted measure. They were invited to participate in this study via telephone, but data were collected face-to-face following informed consent. Only participants whose CLBP were non-specific (not due to malignancy, spinal fracture, infection, inflammation, or cauda equina syndrome) were recruited.
Procedure
The questionnaire was translated and culturally adapted following evidence-based guidelines for a period of one month [34, 35] (Fig. 1).
One bilingual clinical musculoskeletal physiotherapist and one bilingual non-clinical professional translator (native Igbo speakers, bilingual in Igbo and English) forward translated the original CSQ from English to Igbo. Item definitions were provided for the clinical translator to promote understanding of the construct being assessed. Item definitions were not provided for the non-clinical translator to ensure that the translation replicated the lay language used in Igbo culture. T1 and T2 versions of the questionnaire were produced respectively.
A synthesis of T1 and T2 was then performed following discussions between the two forward translators. This was mediated by the first author who is bilingual in English and Igbo to produce a T-12 version. Inconsistencies in the translations were noted by the lead author.
The two back translators back-translated the T-12 CSQ version from Igbo to English. They were both native English speakers from non-clinical backgrounds and blinded to the original measure. One of these back translators was an Igbo/English linguistic expert/professional translator. Hence, BT1 and BT2 back-translated English versions were produced. Back translation validated the translation process ensuring that the translated (T-12) CSQ version was reflecting the construct in the original CSQ.
T1, T2, T-12, BT1 and BT2 versions were subsequently discussed by the expert review committee together with the four translators to produce the pre-final Igbo-CSQ. This committee aimed to achieve semantic, idiomatic, experiential and conceptual equivalence [34]. The expert committee explored Igbo and English words to assess if they meant the same thing, if any item had multiple meanings, and if there were any grammatical difficulties in the translations. The committee helped to formulate alternative Igbo idioms, where English versions were not applicable in the population. The committee also ensured that questionnaire items were experienced similarly in English and Igbo cultures. The committee determined that the words in the items, instructions, and response options had similar conceptual meanings in Igbo and English cultures. They ensured that the Igbo words used were simple and basic.
The pre-final Igbo CSQ was field tested in rural Nigeria among the twelve rural adults. The lead author interviewer-administered the pre-final CSQ using the ‘think-aloud’ cognitive interviewing procedure to assess comprehensibility, acceptability of items and cultural equivalence. Each item was read out by the lead author. Participants were asked to verbalise their thoughts as they tried to answer each question. They were asked if they understood each item, what they understood from each item, the meaning of their chosen response, and if they found any item offensive or irrelevant. They were encouraged to keep verbalising their thoughts while their responses were recorded by the first author. Items that were offensive, irrelevant, or unclear were reviewed by the external review committee together with the translators. This was aimed at ensuring that equivalence was maintained in Nigeria to confirm face and content validity. Technical equivalence was assured via the use of interviewer-administration with all participants.
Psychometric testing
This took place following the completion of the cross-cultural adaptation phase and lasted for another one month.
Sample size
Sample for reliability testing
A minimum sample size of 27 will detect an intra-class correlation coefficient of 0.9 at a 95% confidence interval [28]. A convenience sample of 50 people with non-specific CLBP, between the ages of 18 and 69 years, were recruited from rural and urban communities in Enugu State, Nigeria. This sample was used for the estimation of internal consistency (Cronbach’s alpha), test–retest reliability (intra-class correlation coefficient—ICC and Bland–Altman plot), and minimal detectable change.
Sample for construct validity investigation
A sample of 194 would give an 80% power to detect a very small correlation coefficient of 0.2 at α level of 0.05. Evaluation of construct validity was done as part of a different study aimed at determining the biopsychosocial factors associated with chronic low back pain disability in rural Nigeria [36]. A representative sample of 200 adults, aged 18 to 69 years, with non-specific CLBP were recruited from rural communities in Enugu State, Nigeria using multistage cluster sampling [36]. A total of ten rural communities from ten rural local government areas were randomly selected. Ten community health workers (CHWs) were recruited and trained to collect data from 20 participants randomly selected from each community, resulting in a total of 200 participants. Informed consent was obtained prior to data collection.
Procedure
A training manual was developed, tested and used for training the CHWs for interviewer-administration of the questionnaires. The CHWs were trained on strategies to prevent bias to participants’ responses, and ensure that all questionnaire items were completed. Fidelity checks during data collection ensured that data collection was per protocol. CHWs firstly screened participants using screening questions and a body chart to ascertain that pain was non-specific and in the lower back. Igbo versions of CSQ, BS-11, RMDQ, WHODAS 2.0, FABQ and BIPQ were interviewer-administered with Likert scales presented to participants as ‘flash cards’ as each corresponding item was read out by each CHW.
The Igbo-CSQ was completed at baseline, and repeated 7 to 10 days after, for test–retest reliability investigation amongst the convenience sample of 50 rural and urban participants. The same CHW collected data from each participant on the two occasions.
Igbo versions of CSQ, BS-11, RMDQ, WHODAS 2.0, FABQ and BIPQ were completed at one time-point in a cross-sectional design among the random sample of 200 rural dwellers.
Recruiting different samples enabled a wider applicability of the questionnaire in rural and urban Nigeria, and across literacy levels.
Data analyses
Data analyses were completed with IBM SPSS version 22 and JASP version 0.14.1. Visual and statistical methods were used to assess data normality.
Reliability
Intra-class correlation coefficient (ICC) was used to assess test–retest reliability and evaluated how consistently the Igbo-BIPQ measured illness perceptions over time. A two-way random effects model (with the assumption that measurement errors could arise from either raters or subjects), using an absolute agreement definition between test–retest scores was utilised. Good, very good and excellent ICCs were defined as 0.7, 0.8 and 0.9 respectively [37]. The extent to which all the questionnaire items measure the same construct was investigated using internal consistency (Cronbach’s alpha), and was graded as strong (0.7–1.0), moderate (0.3–0.6) and low/weak (0–0.2) [38]. Visual assessment of the agreement between test–retest measurements were done by plotting mean Igbo-CSQ scores against difference in total Igbo-CSQ scores using Bland–Altman plots. This accounted for the weakness of ICC, which might indicate strong correlations between two measurements with little or no agreement [39].
Standard error of measurement (SEM) and minimal detectable change (MDC) also contributed to reliability investigations. MDC is the smallest change detected by a measure that truly denotes a noticeable change that is not from measurement error. MDC should be sufficiently small to detect minimal clinically important difference [40]. MDC was calculated using the standard error of measurement (SEM) (based on the distribution method), and the reliability of the questionnaire [40]. SEM was estimated using standard deviation (SD) of the sample and the test–retest reliability (R) of the Igbo-CSQ using Eq. (1) below [40]:
$${\text{SEM}} = {\text{SD}}\surd \left( {1 - {\text{R}}} \right)$$
(1)
Equation (1) Standard Error of Measurement.
MDC was estimated with Eq. 2 below:
$${\text{MDC}} = {1}.{96}*\surd {2}*{\text{SEM}}$$
(2)
Equation (2) Minimal Detectable Change where 1.96 = 95% confidence interval of no change; √2 = two measurements [40].
Validity
Construct validity is the degree to which an outcome tool measures the construct it was intended to measure [41]. The domain of construct validity assessed was convergent validity using Spearman correlation coefficient (non-parametric data), and was rated as weak (0–0.2), moderate (0.3–0.6), or strong (0.7–1.0). Convergent validity assesses whether two tools that measure constructs that are assumed to be theoretically related, are related indeed. There are no Igbo pain coping tools. Hence, relationships between pain coping strategies and self-reported numeric pain intensity (BS-11), self-reported back pain specific disability (Igbo-RMDQ), self-reported generic disability (Igbo-WHODAS), self-reported fear avoidance beliefs (Igbo-FABQ), and self-reported illness perceptions (Igbo-BIPQ) reported in the literature were used for validity assessment adopting hypotheses set a priori. Regarding the relationships between the CSQ subscales, pain intensity and disability, catastrophising subscale is expected to have at least a moderate correlation with pain intensity measured with the BS-11 and disability measured with the Igbo-RMDQ and Igbo-WHODAS as suggested in people with CLBP [42,43,44]. Diverting attention, reinterpreting pain sensations, praying or hoping, and increased behavioural activities subscales are expected to have low to moderate correlations with pain intensity measured with the BS-11, and disability measured with the Igbo-RMDQ and Igbo-WHODAS [4,5,6, 9,10,11,12, 45]. Ignoring pain sensations and coping self-statements subscales are not expected to be significantly correlated with pain intensity measured with the BS-11, and disability measured with the Igbo-RMDQ and Igbo-WHODAS [5, 45]. Pain control and pain decrease are expected to be negatively correlated with pain intensity measured with the BS-11, and disability measured with the Igbo-RMDQ and Igbo-WHODAS [6].
Regarding the relationships between the coping strategies and fear avoidance beliefs, negative coping strategies and catastrophising (which may or may not be regarded as a coping strategy) is expected to be positively associated with fear avoidance beliefs [46, 47]. Negative coping strategies include passive coping which are often classified to include praying or hoping, coping self-statements, diverting attention, ignoring pain sensations [3, 5, 6, 9,10,11,12,13] However, the definitions of active, passive, positive, or negative coping strategies appear to differ in different contexts due to cultural differences in coping with pain [16, 48].
Regarding known relationships between coping strategies and illness perceptions, passive coping strategies are known to be stimulated by maladaptive illness perceptions [49]. They are therefore expected to be positively correlated.
Exploratory factor analysis (EFA) was used to determine the number of factors influencing the Igbo-CSQ, that is, its dimensionality [50]. EFA was applied according to Kaiser Meyer Olkin (KMO) and the Bartlett’s test with a minimum eigenvalue for retention set at ⩾1.0 (Kaiser’s rule) [51]. Parallel analysis was included as an additional method for determining the number of factors to be retained in the Igbo-CSQ to compensate for the weakness of the Kaiser criterion which can overestimate or underestimate the number of factors to be retained. In contrast, the parallel analyses shows fewer fluctuations in its accuracy and is more robust [52, 53]. For both methods of determining dimensionality (Parallel analysis and Kaiser criterion), promax (oblique) rotation, which assumes that factors can be related, was done, and factor loadings less than 0.3 were suppressed as recommended; and extraction was done using principal axis factoring as the data had a non-normal distribution [50, 54]. Empirical guidelines are useful, but they are not always correct, and the true number of factors is unknown in reality. Therefore several methods for estimating the number of factors should be utilised and the meanings of findings investigated [52, 53]. The accuracy of empirical guidelines is more likely to be compromised when factors are highly correlated, factor loadings are low, the number of factors is large, and the sample size is small; hence multiple criteria, including relevant theory and previous research, should be used to determine the number of factors to retain [54, 55]. Statistics experts recommend selecting from among a set of competing theoretical explanations the model that best balances the desirable characteristics of parsimony and fit to observed data in terms of interpretability and conceptual sense [54, 55]. Therefore, the number of factors of the adapted measure and their underlying associations were investigated and compared with the original CSQ. Factor pattern coefficients were used for the factor loadings. Scree plot was used for the visual exploration of the retained and excluded factors as recommended [54, 55]. The number of factors and the underlying relationships between Igbo-CSQ items were then compared with the factor structures of the original CSQ to enhance an understanding of population characteristics.
Floor and ceiling effects
When a significant number of participants score the maximum or the least score on a measure, ceiling or floor effect occurs. This implies that the two extremes of the scale are not sufficiently differentiated. For this study, 15% or above was regarded as floor or ceiling effect (Lim et al. 2015). This was estimated for each of the seven subscales found in the original CSQ.