Cross-cultural adaptation and psychometric validation of the Hausa version of Örebro Musculoskeletal Pain Screening Questionnaire in patients with non-specific low back pain

Objectives: Orebro Musculoskeletal Pain Screening Questionnaire (OMPSQ) is widely used in clinical practice and for research purpose to screen the risk of chronicity in patients with Non-specific low back pain (NSLBP). The questionnaire has been cross-culturally adapted into different languages, but to date, there has not been Hausa version of the questionnaire. This study is important as the Hausa language is widely spoken across sub-Saharan Africa. The study aims to cross-culturally translate the English version of the (OMPSQ) into Hausa language (OMPSQ-H) and to test its psychometric properties in Hausa patients with NSLBP. Methods: This observational study involved the use of forward-backwards translation method for the English version of OMPSQ. Thus, 124 male and female participants with subacute NSLBP were recruited using convenient sampling techniques. The psychometric properties statistically tested included reliability, internal-consistency, ceiling and floor effects, acceptability and construct validity. Results: The Hausa version of OMPSQ has demonstrated good reliability (ICC=0.82) and internal consistency (Cronbach’s alpha=0.72) with good acceptability as all questions were answered in 5 min. Responsiveness was adequate as OMPSQ-H retest scores demonstrated good correlation with the global rating of change scale scores (r=0.67, p=0.01). Construct validity was evaluated using principal component analysis and it reveals six components structure for the OMPSQ-H. Conclusions: The OMPSQ-H was successfully translated and cross-culturally adapted with no problem of comprehension. Moreover, it has shown adequate psychometric properties in terms of internal consistency, reliability, responsiveness and constructs validity. Consequently, the OMPSQ-H can be considered as a valid tool for identifying and screening both psychosocial risk factors and risk of chronicity of NSLBP in Hausa population.


Background
Low back pain (LBP) is arguably the most prevalent musculoskeletal condition found among both developed and developing nations [1,2]. Despite the high incidence and prevalence of LBP, little is known about its precise causes. As a clear pathoanatomic diagnosis cannot be UAA, TN and SSM conceptualized and designed the study protocol, UAA and BK collected and analyzed the data, SSM and TN provided critical review of the manuscript and they are the supervisors of UAA who is the correspondent author and a PhD student. Also, TN obtained funding for the study. identified in 85% of the patients [3], thus, LBP in these patients are labelled as non-specific low back pain (NSLBP). According to the Global Burden of Disease (GBD) 2010 study, LBP is currently the 6th highest burden on a list of 291 conditions and is the cause of more years lived with disability globally than any other disease [4].
The prevalence of LBP worldwide is estimated to be between 30 and 80 percent among the general population and has been found to increase with age [4]. Besides, a higher prevalence of LBP has been associated with lower socioeconomic status and lower education levels [2,5]. A recent systematic review and meta-analysis published in 2018 revealed that the lifetime, annual and point prevalence of LBP among African nations, was higher than the global LBP prevalence [6]. The majority of studies included in this meta-analysis were conducted in Nigeria, which is a lower-middle-income country. In Nigeria, the annual prevalence rate of LBP reported has been between 33 and 74%, mostly affecting workers [7].
Affecting just about anyone, of any gender, race or socioeconomic background [8], LBP has a substantial impact on the overall physical, economic and psychosocial well-being of an individual and the society [1,8]. The prognosis after an acute episode of LBP is less favourable than once thought [9,10], as 60-80% of the patients will experience recurrence or chronicity of this disabling condition [10]. Furthermore, chronic NSLBP disability results in a poor quality of life, decreased functional disability, activity limitations and participation restrictions.
Predictive factors resulting in disabling chronic NSLBP have been implicated in the transition from acute to chronic NSLBP [11]. The risk of developing chronic disability is associated with psychosocial factors (yellow flags) [10,11]. The psychosocial factors are important risk factors for longterm disability and guidelines for the management of LBP highlights the importance of screening for psychosocial factors [12,13]. Therefore, it's clinically important to assess and evaluate the psychosocial factors to screen patients at risk of future chronicity which will help make clinical decisions regarding prognosis and treatment plan.
This should be done with a valid and specific instrument [14,15]. World Health Organization International Classification of Functions and Diseases (WHO-ICF), recommends Örebro Musculoskeletal Pain Screening Questionnaire (OMPSQ) as an instrument for screening the risk of chronicity and measurement of psychosocial risk factors in patients with NSLBP [16]. The Örebro Musculoskeletal Pain Screening Questionnaire (OMPSQ), originally referred to as the Acute Low Back Pain Screening Questionnaire (ALBPSQ) was developed by Linton & Halldén [17] to assist primary care practitioners in identifying psychosocial "yellow flags" and patients at risk of future work disability due to persistent pain.
The Hausa Language is widely spoken in sub-Saharan Africa. It is the predominant language in the Northern part of Nigeria. It is the first language of about 50 million people and the second language of about 20 million people. Therefore, approximately 70 million people use the language as a medium of communication [18]. It has however been reported that a sizeable number of patients in Nigeria do not speak or write English [19].
Therefore, the purpose of this study was to translate the original OMPSQ into the Hausa Language and determine its psychometric properties in terms of validity, reliability and internal consistency of its use among Hausa speaking patients with NSLBP for its easy administration by clinicians and researchers.

Study design
A cross-sectional study aimed at translating OMPSQ into Hausa language and then testing its psychometric properties in patients with chronic NSLBP.

Participants
According to Terwee et al. [31] on the quality criteria for measurement properties of health status questionnaires, they recommend recruitment of ≥50 participants for construct validity, reliability, responsiveness and ceiling/floor effect analyses. This was also reported by a study on Cross-cultural adaptation, validity and reliability of the Hausa version of the Neck Disability Index Questionnaire [18]. For this study, 124 participants were initially recruited but only 120 completed the study, the participants comprised both male and female patients receiving treatment at Physiotherapy departments.
Only participants who met the inclusion criteria were considered for the study. The inclusion criteria included; participant who can read, write and speaks Hausa language; NSLBP as diagnosed by a physician and commonly associated with referred lower extremity pain; NSLBP ≤3 months; participants with age ≥18 years; pain in the lumbopelvic region; movement coordination impairments while performing community, recreational or occupational activities; low back pain not of specific origin diagnosed by a physician (Fracture, tumours, malignancy, ankyloses, Infections and pregnant women). Participants were excluded when they presented with previous spinal surgery; chronic low back pain with related generalized pain; spinal pathologies and diseases associated with cognitive impairment or widespread neurological symptoms.

Instruments
Orebro Musculoskeletal Pain Screening Questionnaire: The OMPSQ was developed by Linton and Boersma [20] to assist primary care practitioners in identifying psychosocial "yellow flags" and patients at risk of future work disability due to persistent pain. The questionnaire consists of 25 items, item 1-4 take sociodemographic information whilst item 5-25 formed the six components (factors) of the questionnaire (i.e. Pain, Coping Strategy, Distress, Work return expectancy, Fear-avoidance beliefs, Function) each item was scored from 0-10 [27]. The scored items were summed to provide a total score ranging from 0 to 210, a higher score indicates a higher disability. The maximum score is 210 points; a score of <105 points indicates a low disability, that between 105 and 130 points indicate a moderate disability and that >130 points indicates a high disability [20].

Global rating of change Scale (GROC):
The global rating of change scale (GROCS) was used to evaluate the participants' overall perceived effect of treatment. This outcome tool was used to evaluate overall patient satisfaction on health status after treatment and to determine the responsiveness of the OMPSQ-H by checking the relationship between OMPSQ changes and GROC scores of the participants.
The global rating of change (GROC) is a scale which ranges from −7 (a very great deal worse) to 0 (about the same) to +7 (a very great deal better). Incremental descriptors of worsening and improving were assigned values from −1 to −7 and from +1 to +7, respectively [33]. The following classifications have been proposed regarding the GROC scores: -0, 1, or −1 signifies no change -±2 to 3 signifies a minimal change -±4 to 5 signifies a moderate change -±6 to 7 signifies a large change in a patient's condition.
Patients with an average GROC rating ≥3 (somewhat better) were considered to have improved, while patients with an average rating of <3 to −3 were considered to have remained stable, while patients with an average rating of ≤−3 (somewhat worse) were considered to have worsened [34]. In this study, all the patients were classified into the four groups for data analysis.

Translation process
The consent and approval of the original developer of OMPSQ were sought and obtained before the commencement of this study and the original English OMPSQ was used as the basis for this translation.
The translation of OMPSQ into the Hausa Language followed the guidelines for the process of translation of the self-reported questionnaire as recommended by Beaton et al. [35].
The Forward translation: Two independent bilingual translators from the department of Linguistics Bayero University Kano whose mother tongue was Hausa Language and were proficient in English with expertise in translation of questionnaires produced the translated version of OMPSQ into Hausa language (T 1 and T 2 ). The two translators had different backgrounds; one of the translators was a Physiotherapist and knew the concepts being examined in the questionnaire. The other translator was a linguist who did not have an idea or information of the concept to be quantified in the questionnaire.
The Synthesis of the Translations: The two translators were instructed to meet and synthesize the result of their translation to produce one common translation (T 1, 2 ), the process obligates both translators to produce a written report of the completed translation, highlighting challenging phrases and uncertainties as well as a summary of the rationale behind their choice of words.
The Back Translation: Another two independent translators who were blinded to the original version of the questionnaire and without medical background translated the synthesized Hausa version (T-12 ) back to English (source language).
The Expert Committee: For the targeted questionnaire to achieve a cross-cultural equivalence with the source version, an expert committee was established which involved two physiotherapists (with experience in cross-cultural adaptation studies), the translators (involved in both forward and backward translations) and two research experts in cross-cultural studies. The committee consolidated all the versions of the questionnaire and developed the pre-final version of the questionnaire.
Test of the pre-final version: The pre-final version of the OMPSQ-H was administered to 30 patients diagnosed with chronic NSLBP to assess face validity. The participants after completing the questionnaire were called for an interview to probe what they think was meant by each item of the questionnaire and their chosen response. In the meeting both the meaning of the items and responses were explored. This was to ensure that the adapted version still retained its equivalence as the original version and thus it was considered the final Hausa version of OMPSQ.

Study procedure
Initially, 132 participants indicated interest to participate in the study, but only 124 successfully met the study inclusion criteria, thus were recruited and administered the OMPSQ-H questionnaire and GROCS for baseline assessment. After two weeks, the participants were reassessed using the same outcome tools, those who were not available were contacted and reassessed through phone calls. However, four participants could not be contacted after two weeks and were excluded from the study. Therefore, only 120 participants completed the present study and thus their data reported.

Data analyses
Normality of data distribution was tested using the Kolmogorov-Smirnov test. Acceptability of the questionnaire was assessed by recording the time spent answering the questionnaire, once completed the patients were asked about whether they might have encountered any difficulty while filling the questionnaire and the examiners checked for any missing or multiple responses.
Sociodemographic variables were analyzed using descriptive statistics of mean, frequency, percentages and correlation matrix. The descriptive statistics were also used to determine the floor/ ceiling effects, which could be considered present when more than 15% of the patients scored either the lowest or highest possible scores. Reliability was investigated by testing internal consistency using Cronbach alpha coefficient, test-retest reliability (interclass correlation coefficient; ICC) and using correlation analysis, the testretest reliability measures reliability over time by administering the same questionnaire to the same subjects after a certain interval (2 weeks for this study). The internal consistency reflects the homogeneity of the questionnaire, which would be considered good if the value of Cronbach α was more than 0.70 [31]. For responsiveness, Pearson correlation coefficient was used to check the relationship between OMPSQ changes and Global Rating of Change (GROC) scores of the participants, α value set at 0.05. For responsiveness, Pearson correlation coefficient was used to check the relationship between OMPSQ changes and Global Rating of Change (GROC) scores of the participants, α value set at 0.05.
Construct Validity (factor analysis), the 21 items of the OMPSQ-H scale were subjected to Principal Component Analysis (PCA). PCA was used to elucidate the covariance structure of a set of variables. Before performing PCA, the suitability of the data for factor analysis was assessed. Inspection of the correlation matrix revealed the presence of many coefficients of 0.3 and above. The Kaiser-Meyer-Olkin value was 0.60; which was equivalent to the recommended value of ≥0.6, and Bartlett's test of sphericity reached statistical significance, supporting the factorability of the correlation matrix [36].

Normality test
A Kolmogorov-Smirnov test indicated a value of p>0.06; this showed that the data was normally distributed.

Sociodemographic characteristics
The age of the participants ranged from 22-70 years, with a mean age of 37.26 ± 11.3. Eighty-eight of the participants were male while 32 were female, i.e. 73.3 and 26.7% respectively (Table 1).

Floor/ceiling effects
Based on the interpretation of original OMPSQ by Linton and Boersma [20] the scores are computed as a single screening tool with a maximum score of 210 and a minimum of 0. For this study, the participants with maximum score were 6 (5%) while those with minimum score were 5 (4.2%). Given that ˂15% of the participants scored the maximum and minimum scores, therefore, the (OMPSQ-H) had no ceiling or floor effects.

Acceptability
All the OMPSQ-H questions were well accepted, the average time taken to complete filling in the questionnaire was 5 min. Additionally, no missing responses or multiple answers were found and there were no problems with comprehension as regards the questions in respect to literal and contextual meaning.

Reliability
The internal consistency Cronbach α of the OMPSQ-H was 0.718. The test-retest reliability (interclass correlation coefficient) was 0.82 (95% CI; 0.65-0.90) and the correlation between the two different administrations of the instrument-first administration and after two weeks-was determined using the Spearman's correlation. There was a significant correlation between the first and the second administration (r=0.84, p<0.05) of the OMPSQ-H and with a minimum detectable change score of 28.3 and standard error of measurement 24.7 (Table 2). This implies that the OMPSQ-H has shown good reliability and internal consistency.

Responsiveness
The post GROC scores and OMPSQ-H changes scores were correlated using the Pearson's moment correlation; this revealed a significant relationship (r=0.67, p=0.01).

Construct validity (factor analysis)
The principal components analysis used revealed the presence of six components with eigenvalues exceeding 1, this explained 23.80, 15.95, 9.11, 7.80, 7.10 and 6.50% of the variance respectively. A scree plot revealed a clear break after the sixth component. Using Cattell's scree test, it was decided to retain all the six components for further investigation. This was further supported by the results of a parallel analysis, which showed all the six components exceeded the eigenvalues corresponding to criterion values for a randomly generated data matrix of the same size. The six components solutions explained a total of 70.23% of the cumulative variance explained in Table 3.
To aid the interpretation of these six components, an Oblimin rotation was performed because varimix rotation does not allow factors to be inter-correlated, while by contrast, the oblimin rotation allows factors to be intercorrelated (28). The rotated solution revealed the presence of simple structures with both components showing some strong loadings >0.3, and all variables loading substantially on all the components. The rotation process simplifies the structure and the interpretation of how each question saturates the components in the OMPSQ-H questionnaire. Normally, the saturation level of a question should be close to one, so that it saturates on a single component. However, it has been noted that some questions have moderate saturation and thus saturate on more than one component.  (Table 4). Commonalities also gave information on the size of the variance in each item; values less than 0.4 indicated a misfit while values above 0.4 showed that the items were well fitted with one another (Table 4).

Discussion
The translation and adaption process was successfully done following the recommendation from the work of Beaton et al. [35], however, the few problems encountered with wording were simply overcome by expert consensus. The OMPSQ-H can be self-administered and proved easy to administer in clinical practice during this study. The OMPSQ-H had a good score on psychometric properties and thus can reasonably be confirmed that it can be used successfully to screen patients with psychosocial factors who were at risk of developing long-term disabilities associated with NSLBP.
To this authors' knowledge, this is the first crosscultural adaptation study on OMPSQ into the Hausa language, and the aim was to determine its psychometric properties. There were no difficulties encountered in translating the questionnaire, however question 4 from the original English version which says "Were you born in Australia" was replaced with "are of native Hausa tribe from Nigeria". Due to cultural difference, there were only a few missing values, but the overall contextual equivalence was retained in all the questions as the participants demonstrated no problem in completing and understanding the questions given in the questionnaire.
Furthermore, in comparison with previous studies on translation and measurement of psychometric properties  of OMPSQ, this study showed that the OMPSQ-H has proven to be a reliable and valid tool in identifying patients with NSLBP that were at risk of developing chronicity of NSLBP. These results were overall comparable with the original English version of OMPSQ [21,37] and other OMPSQ cross-cultural studies into different languages e.g. German [22], Chinese [24], Chinese-Mandarin [30], Persian [23], Turkish [25] Brazilian-Portuguese [14], Spanish [27], French [26], Dutch [28] and Norwegian [29] ( Table 5). The determination of OMPSQ-H factor structure using PCA revealed six clear factors which replicate the theoretical conceptualization of the original English questionnaire [11]. The PCA was selected based on the study of Westman et al. [11] the results of this analysis showed that all the components have substantial adequate loading on all the items of the questionnaire and thus the items were further analyzed using Oblimin rotation to aid further interpretation of the six components. All items had adequate factor loading and in agreement with the analysis of their respective commonalities which revealed all items were well fitted with one another (Table 5) and that, the common features of the extracted factors were in agreement with the various domains of the ALBPSQ [17]. Therefore, it was not necessary to eliminate any of the items from the OMPSQ-H version. The percentage of the variance was a little higher than the original English version [17] but closely approximate with both ALBPSQ and Modified OMSQ version [17,38]. Additionally, the percentage of variance was comparable to the Spanish and Dutch version of OMPSQ [27,28], however, Grotle et al. [29] initially reported three components factor on the Norwegian OMSPQ, but Gabel et al. [38] stated that the reanalyzed data of the work reveals six components factor.
Patient satisfaction with the therapeutic intervention is very important in the management of patients with chronic NSLBP due to the implication of the psychosocial factors associated with chronic disabling LBP [39]. Thus, in this study, GROCS was used to determine if OMPSQ-Hausa was able to detect changes over some time. Consequently, a strong correlation was found between OMPSQ-H and GROCS which suggests that OMPSQ-H was able to detect changes over some time and thus it can be applied in the clinic for the assessment of Hausa speaking patients with NSLBP.
Some of the limitations of this study include; firstly, conducting the retest was not easy in this population, therefore, the fact that the administration methods were different from the test to the retest is a limitation of the, however, that was the only option available for the authors. Secondly, Dawson et al. [40] recommended 2-3 days for testretest assessment to ensure that changes in the patient's condition were minimal [40], however, both Linton et al. [37]  The following recommendations were that this study did not evaluate the predictability of OMPSQ to screen chronicity over a period of time; therefore, we recommend further study to determine the predictive validity of the questionnaire. Furthermore, this study did not involve the short-form of OMPSQ, however, we recognized its clinical importance with regards to ease of administration.
Consequently, we strongly recommend further study to be conducted on the short-form OMPSQ to determine its psychometric properties and thus provide clinicians and researchers with options for the assessment of the risk of chronicity in patients with NSLBP.

Conclusion
The OMPSQ-H was successfully translated, it was well accepted by the participants as all questions were adequately responded to with no problem of comprehension and there was no missing responses or multiple answers. Therefore, the OMPSQ-H has demonstrated in this study a good acceptability, reliability, internal consistency, responsiveness and constructs validity with no ceiling or floor effects. Thus, it has an ideal factor structure and constructs agreement with the original OMPSQ.
This shows that the OMPSQ-H can be considered as a validated tool for identifying and screening both psychosocial risk factors and risk of chronicity of symptoms of NSLBP in Hausa population of Nigeria and sub- Saharan Africa. In practical terms, this result indicates evidence that treatment targeting these psychosocial factors itemized in the questionnaire will significantly improve the treatment outcomes on disabilities associated with NSLBP.