The Hausa Northwick Park Neck Pain Questionnaire: translation, cross-cultural adaptation and psychometric assessment in patients with non-specific neck pain

Abstract Purpose To translate and cross-culturally adapt the Northwick Park Neck Pain Questionnaire (NPQ) into Hausa and assess its psychometric properties. Materials and Methods The NPQ was translated and cross-culturally adapted into Hausa using recommended guidelines. A consecutive sample of 92 Hausa-speaking patients with non-specific neck pain recruited from three tertiary hospitals in north-western Nigeria, completed the questionnaire to assess factorial validity (using confirmatory factor analysis), convergent validity (by correlating the Hausa-NPQ with the Numerical Pain Rating Scale [NPRS]), and internal consistency (using Cronbach’s α). A subsample of 50 patients completed the questionnaire again 3 days after the first administration to assess relative reliability using intraclass correlation coefficients (ICC) and absolute reliability using standard error of measurement (SEM), smallest detectable change (SDC), and 95% limits of agreement (LOA). Results The factor analysis confirmed a single-factor structure with excellent internal consistency (α = 0.94). The questionnaire showed a strong positive correlation with the NPRS (rho = 0.68). The ICC was 0.86, with SEM and SDC of 6.32 and 17.5, respectively. The LOA was − 29.3 to + 37.1 with no evidence of proportional bias. Conclusions The Hausa-NPQ is a valid and reliable measure of disability due to neck pain. Implications for Rehabilitation This study describes the translation, cross-cultural adaptation, and psychometric assessment of the Hausa-NPQ in Hausa-speaking patients with non-specific neck pain. The questionnaire demonstrated adequate psychometric properties in terms of factorial and convergent validity, internal consistency, and test-retest reliability. The questionnaire will be useful in clinical and research settings to assess disability due to neck pain for screening purposes, evaluation of treatment effectiveness, as well as cross-cultural comparisons involving Hausa-speaking individuals with neck pain.


Introduction
Non-specific neck pain, defined as pain experienced in the neck without a specific underlying disease causing the pain [1], is one of the most common debilitating musculoskeletal conditions and leading causes of disability throughout the world [2,3].It is associated with an enormous socioeconomic burden, thus, considered to be a major public health problem in the general population [3,4].
The age-standardized point prevalence (males 4.7%; females 6.7%) estimated for sub-Saharan Africa is relatively higher than the global estimates (males 4.0%; females 5.8%) [3].In Nigeria, the prevalence of neck pain over 12 months was reported to be between 6.2% and 81.9% among adults [5-8], suggesting neck pain is a common problem and may constitutes a significant burden on Nigerian physiotherapy care facilities.
Given that neck pain is a common source of functional limitation, assessment of neck pain-related disability is crucial and depends largely on accurate clinical data.This would necessitate the use of standardized patient self-report outcome measures [9].Several instruments such as the Neck Disability Index (NDI) [10,11], the Neck Pain and Disability Scale (NPDS) [12], the Copenhagen Neck Functional Disability Scale (CDS) [13], and the Northwick Park Neck Pain Questionnaire (NPQ) [14] have been developed and validated for the assessment of disability due to neck pain disorders in English-speaking countries.The use of these standardized instruments, however, is limited in non-English-speaking countries/cultures.Fortunately, guidelines for cross-cultural adaptation if followed accordingly provide with the opportunity to adapt an instrument developed in English into another language for clinical and research purposes [9,15].
A couple of patient self-report questionnaires for assessing disability status were adapted and validated for use in the three major Nigerian languages (i.e., Hausa, Igbo and Yoruba).However, the majority of these questionnaires were developed for patients with low back pain [16][17][18][19][20], and there have been limited validated questionnaires for patients with neck pain [21] despite its significant burden.Adapting standardized self-report questionnaires for measuring disability due to neck pain into Nigerian languages, particularly Hausa (the largest indigenous spoken language in Nigeria and many other West African countries [22]) will enhance clinical use and research uptake, as well as prevent the exclusion of patients who cannot speak English.
The NPQ was devised by Leak et al. [14] in 1994 following adaptation of the Oswestry Disability Index (a commonly used questionnaire for assessing disability due to low back pain) [23] to overcome the difficulties in quantifying subjective pain and disability associated with neck pain disorders.The questionnaire has been so far translated and validated in seven different languages/cultures [24][25][26][27][28][29][30][31].The psychometric properties of the NPQ in terms of construct validity, internal consistency, and test-retest reliability reported in these studies were generally acceptable.Additionally, the questionnaire has been shown to be easily administered (requiring < 10 min to complete) and scored, thus supporting its clinical utility in screening, diagnosis and management of disability due to neck pain [25,29,31].
As no Hausa version of the NPQ exists at the time when this study was initiated, the purpose of this study was, therefore, to translate and cross-culturally adapt the NPQ into Hausa and assess its psychometric properties in Hausa-speaking patients with nonspecific neck pain.

Ethical considerations
Ethical approval was sought and obtained from the Research and Ethics Committees of Aminu Kano Teaching Hospital (AKTH/MAC/SUB/12A/P-3/VI/1700) and Hospitals Management Board (MOH/Off/797/T.I/104), Kano State, Nigeria.The procedure and purpose of the study were explained to each participant, thereafter, informed consent form was applied.

Study design
Cross-cultural translation, adaptation, and psychometric assessment of the Hausa version of the NPQ.

Northwick park neck pain questionnaire (NPQ)
The NPQ consists of nine sections focused on daily activities that may be affected by neck pain intensity, neck pain and sleeping, pins and needles or numbness in the arms at night, duration of symptoms, carrying, reading and watching television, working/ housework, social activities, and driving [14].Each section contains five statements scored from 0 to 4 as possible answers.To obtain the total score expressed in percentage, the scores obtained for each section (0-36) are added together and divided by the number of answered questions and then multiplied by 100.Higher scores indicate higher disability due to neck pain.Section ten evaluates the patient's assessment of changes in pain after follow-up and is not included in the NPQ final score [14].The original English version has been shown to have good repeatability, internal consistency and responsiveness among patients with neck pain [14,32].

Numerical pain rating scale (NPRS)
The Hausa version of the NPRS [33] was used to assess neck pain intensity.It is a horizontal bar or line of 0-100 mm anchored on the left with the phrase "No Pain" and on the right with the phrase "Worst Imaginable Pain."The respondents were asked to mark a single number that best reflect their current pain intensity.The scale has been shown to be valid, reliable and responsive [33].

Translation and cross-cultural adaptation
The translation and cross-cultural adaptation of the NPQ into Hausa language was performed in accordance with the guidelines established by Beaton et al. [9].The process (Figure 1) consists of the following five stages: 1. Forward translation: The English version of the NPQ was translated into Hausa by two independent language experts who are fluent in English and Hausa and whose first language is Hausa.One of the translators was provided with information on the conceptual basis of the scale, whereas the other translator was not provided with such information.
The translators produced two forward-translated versions (T 1 and T 2 ).The two forward translators and the lead author (AAA) sat together to review inconsistencies and differences in their forward translations, and a consensus translation (T 3 ) was produced.2. Backward translation: The consensus Hausa translation (T 3 ) was translated back into English by two independent translators who had no medical background and were blinded to the forward translation.This step produced two backwardtranslated versions (T 4 and T 5 ).The two backward translators and the lead author reviewed and summarize differences in T 4 and T 5 .3. Expert committee review: A group of experts was constituted who reviewed the forward and backward translations including all reports from the translators and agreed by consensus to produce a prefinal version of the Hausa-NPQ.4. Pilot testing: The prefinal Hausa-NPQ version was tested among 15 patients with non-specific neck pain recruited from the physiotherapy clinic of Aminu Kano Teaching Hospital, Kano State, Nigeria to assess face and content validity.The participants were asked to answer the questionnaire and thereafter they were asked about the clarity of the questionnaire items and whether they experienced any difficulty in completing the questionnaire.The lead author in collaboration with the expert panel reviewed and resolved the comments of the participants by consensus.5. Proofreading: The final Hausa-NPQ version was sent to an independent professional translator for proofreading to correct any text errors that may have been missed in the previous stages.The final version (Supplemental material 1) was then produced and ready for psychometric assessment.

Sample size estimation
The sample size was estimated based on the recommendation of the quality criteria for measurement properties of health status questionnaires, which suggest a minimum sample size of 50 participants would be adequate for test-retest reliability, construct validity and ceiling/floor effects analyses, as well as a minimum of 100 participants would be adequate for internal consistency [34].
Following this recommendation, we intended to recruit 152 participants, however, we were able to obtain 102 participants due to Covid-19 pandemic and industrial action of hospital workers during the study period.

Participants and settings
The study population purposely included consecutive individuals with non-specific neck pain attending physiotherapy outpatient clinics of Aminu Kano Teaching Hospital (AKTH), Murtala Muhammad Specialist Hospital (MMSH), and Muhammad Abdullahi Wase Teaching Hospital (MAWTH), all in Kano State, Northwestern Nigeria.The participants were included in the study if they were: (1) males or females between the age of 18 to 65 years, (2) suffering from non-specific neck pain of at least � 12 weeks duration.Non-specific neck pain was defined as pain experienced in the neck and was associated with significant emotional distress such as anxiety, anger, frustration, or depressed mood and/or interference in activities of daily life and participation in social roles [35]), and (3) able to speak Hausa.Participants were excluded from the study if they exhibited: (1) symptoms of specific disorders of the cervical spine, such as fracture, tumor, infection, and spinal stenosis, (2) obvious spinal deformity, (3) evidence of nerve root compression such as myotomal weakness, (4) spasmodic torticollis, (5) fibromyalgia, (6) systemic inflammatory conditions such as rheumatic disease and ankylosing spondylitis, and (7) psychiatric illness.Eligibility of the participants was ensured through history taking and physical examination performed by a physiotherapist.

Assessments
The participants' demographic information such as age, gender, marital status, and occupational status was obtained via interviews.The participants were requested to complete the Hausa-NPQ and Hausa-NPRS.After three days, the Hausa-NPQ was re-administered among 50 consecutive patients to assess testretest reliability.

Statistical analyses
All statistical analyses were carried out using IBM SPSS version 23.0 (IBM Corp, Armonk, NY) with the level of significance set at 0.05.The following statistical methods were used in assessing the psychometric properties of the Hausa-NPQ.

Content validity
Content validity of a measure refers to the degree to which the measure is relevant and representative of the construct it is designed to measure [36].Two panels were constituted which consisted of nine physiotherapists (content experts) and five patients (lay experts) with non-specific neck pain of a duration � 12 weeks.Both panels were literate in Hausa.To arrive at content validity, the content validity index (CVI) was calculated.The panel members were asked to rate each item of the questionnaire in terms of its relevancy to the construct underlying the study on a four-point ordinal scale.Thereafter, the number of the panelist that judged the item as relevant (rating 3 or 4) was divided by the total number of the panelist, this represents the CVI in item level, I-CVI, and this expresses the proportion of agreement on the relevancy of each item, which ranges between 0 and 1, with value > 0.79 being considered relevant [37].

Ceiling and floor effects and response trend
Ceiling or floor effects occur if more than 15% of respondents scored the maximum or minimum possible score [34].Potential ceiling or floor effects of the Hausa-NPQ were assessed by calculating the percentage of respondents showing the maximum or minimum possible score in all 9 items of the questionnaire.
Response trend was assessed by calculating skewness.Items with a skewness >1.96 indicate a response trend that deviate from a normal distribution [34].

Factor structure and internal consistency
Exploratory factor analysis (EFA) with a direct oblimin extraction was conducted to examine whether the NPQ items form a unidimensional scale.Item 9 was removed from the EFA due to missing values.Suitability of the EFA was assessed by inspecting inter-item correlations of the 8 items of the Hausa-NPQ and calculating the Bartlett test of sphericity and the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy.A significant KMO value of > 0.6 was considered an acceptable measure of sampling adequacy [38].A Scree plot was applied to visualize the number of factors extracted based on eigenvalues >1.The internal consistency was then assessed by calculating Cronbach's alpha (a) with values 0.70-0.79,0.80-0.89,and � 0.90 being considered as adequate, good, and excellent, respectively [39].To confirm the underlying factor structure from the EFA, confirmatory factor analysis (CFA) was conducted using maximum likelihood estimates.Modification indices were observed for item's redundancy and correlation of error terms to improve model fit.The model fit was assessed using the ratio of chi-square to degrees of freedom (v2/df), comparative fit index (CFI), Tucker-Lewis index (TLI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR).The following cut-off criteria were considered acceptable for a model fit: v2/df of � 2.0, CFI of � 0.95, TLI � 0.90, RMSEA � 0.06, and SRMR � 0.08 [40,41].

Convergent validity
To assess convergent construct validity, the Hausa-NPQ was correlated with the Hausa-NPRS.We hypothesized that the Hausa-NPQ would demonstrate a moderate positive correlation with the Hausa-NPRS [33].Spearman's correlation coefficients (rho) were used as the data had a skewed distribution.The strength of the correlation was interpreted as strong (rho ¼ > 0.60), moderate (rho ¼ 0.30-0.60),and weak/low (rho ¼ < 0.30) [42].

Relative reliability
Test-retest reliability as relative reliability of the Hausa-NPQ was assessed by calculating the intraclass correlation coefficient (ICC) for agreement using a two-way random effects model (which assumes that measurement errors could arise from either raters or participants).ICC values � 0.70 were considered adequate testretest reliability [34].

Absolute reliability
Absolute reliability of the Hausa-NPQ was assessed by calculating the standard error of measurement (SEM), smallest detectable change (SDC) at 95% confidence interval, and 95% limits of agreement (LOA).The SEM indicates the spread of measurement errors when estimating true scores from observed scores, whereas the SDC indicates the minimum amount of change considered a true change above the measurement error [43].We calculated the SEM using the formula, SEM ¼ SD change x �(1 -ICC) [44] where SD change is the standard deviation of the mean change score of the Hausa-NPQ.Subsequently, the SDC was calculated using the formula, SDC ¼ 1.96 x �2 x SEM [44].LOA was assessed with Bland-Altman plots to visualize measurement error by plotting the mean scores of test-retest against the difference between testretest scores.Evidence of proportional bias was verified using simple linear regression analysis to explore the relationship of the mean of test and retest scores with the difference of test and retest scores.A non-significant (p > 0.05), unstandardized beta coefficient (b) values close to zero suggest no evidence of proportional bias.

Translation and cross-cultural adaptation
The Hausa-NPQ translation had no major translation problems as the translators ensured that standard Hausa dialects were used to attain semantic, idiomatic, experiential and conceptual equivalence between the Hausa version and the original English version.
For item 8 (social activities), the phrase "social gatherings/events, such as weddings or parties" was supplemented for clarity in the Hausa culture.All the questionnaire items were comprehensive during the pilot testing.

Psychometric assessment
One hundred two patients (66 males and 36 females) were recruited for the study between February 2017 and December 2021.The mean age of the participants was 47.0 ± 14.0 years.Twenty nine patients were recruited from AKTH, 46 from MMSH, and 27 from MAWTH.Of the total patients recruited, 50 returned the completed questionnaire for the test-retest assessment.
Table 1 shows the demographic and clinical characteristics of the study population.

Content validity
The calculated I-CVIs ranged from 0.85 to 0.92 indicating that all items were appropriate and the content validity of the questionnaire was acceptable among the raters.The questionnaire was, therefore, suitable for psychometric testing.

General aspects, ceiling and floor effects
Of the 102 participants completing the Hausa-NPQ, 10 participants did not complete the questionnaire and were excluded from the subsequent analyses.Out of the 92 participants that completed the questionnaire, 49 (53.3%) did not answer item 9, which is related to driving as the item is not applicable to them (Table 2).It took less than 10 min (mean ¼ 6.49 min) for the participants to complete the Hausa-NPQ.No floor effects were detected in the questionnaire total score and the individual items.However, except for item 9, ceiling effects were detected in all the questionnaire items (Table 2).Three participants (3.3%) achieved the highest total score (97.2%), whereas 7 participants (7.6%) had the lowest score (0.00%).The response trend for each item does not deviate from normal distribution as none of the items showed skewness > 1.96 (Table 2).

Factorial validity and internal consistency
Initial analysis revealed a KMO value of 0.912 and a significant Bartlett test of sphericity (v 2 ¼ 567.17, df ¼ 28, p < 0.001) suggesting the appropriateness of the data for factor analysis.The 8 items of the Hausa-NPQ loaded on 1 factor, explaining 69.5% of the total variance.The factor loadings ranged from 0.71 (item 5 "carrying") to 0.90 (Item 1 "neck pain intensity") (Table 2).As the Scree plot of eigenvalues, the percentage of variance, and the number of eigenvalues > 1 indicated a 1-factor solution, no rotation could be executed.The Cronbach's a was 0.94 and the itemtotal correlation ranged from 0.80 (item 3 "Pins and needles or numbness in arms at night") to 0.92 (item 2 "neck pain and sleeping") (Table 2).The CFA confirmed the Hausa-NPQ as a unidimensional scale as indicated by the acceptable model fit indices (Table 3) after allowing 3 error terms to covary (e1-e2, e3-e5, and e3-e8) (Figure 2).

Convergent validity
The total score of the Hausa-NPQ showed a strong positive and significant correlation with the Hausa-NPRS (rho ¼ 0.68, p < 0.001) as hypothesized.

Relative reliability
As shown in Table 4, the relative reliability of the Hausa-NPQ using ICC was high (0.86; 95% CI ¼ 0.75-0.92).The item-by-item ICCs were high (� 0.73) except for item 4 which showed a low correlation coefficient (0.29).The SEM and SDC were 6.32 and 17.5, respectively.The mean difference (4.31) for the repeated measurements was not statistically significant (p > 0.05).
The Bland-Altman method showed a mean difference between test and retest of 3.91, with 95% LOA of À 29.3 to þ 37.1 (Figure 3).The unstandardized coefficient calculated was close to zero and statistically non-significant (b ¼ 0.15, p ¼ 0.148).

Discussion
Neck pain, especially in its chronic state is one of the leading causes of disability in Nigeria but there have been limited validated self-report measures in Nigerian indigenous languages to enable the assessment of disability due to neck pain.To the best of our knowledge, this is the first study to describe the translation and cross-cultural adaptation of the NPQ into Hausa and evaluate its psychometric properties in patients with non-specific neck pain.The findings suggested that the Hausa-NPQ had acceptable validity and reliability for clinical and research use among Hausaspeaking individuals with non-specific neck pain.
Consistent with other adaptations [9, 24,25], the Hausa-NPQ was easy to translate, cross-culturally adapt, and comprehend as there were no major issues encountered during the adaptation process.The questionnaire is closely related to the original English version as most of the terms were literally equivalent.It had a good content validity evidenced by the higher I-CVIs values (range: 0.85-0.92),which lies within the range of values (0.79-1.00) considered acceptable [37].The result that the questionnaire had a good response trend evidenced by items' skewness less than 1.96 further supports its content validity.Additionally, our questionnaire proved to be acceptable as most patients found it easy to complete requiring less than 10 min similar to previous adaptations [25,29,31].
Ceiling effect (highest score) was neither detected for all the items nor for the total score.However, a moderate floor effect (lowest score) was detected for all the items except item 9 (driving) and the total score.The presence of floor effect in most items of the questionnaire may implies its inability to discriminate among those with lower levels of disability due to neck pain.The fact that more than half (53.3%) of the respondents left item 9 unanswered indicates that driving does not apply to the majority of them.Higher missing data rates greater than 50% in item 9 were also reported in other adapted versions of the NPQ [24,25].In contrast to our study, the Brazilian Portuguese and French adaptations [26,30] reported lower missing data rates, with neither ceiling nor floor effects being detected.Similar to previous validation studies [24,26], the respondents in this study had low levels of disability due to neck pain (mean ¼ 24.3%) at the beginning of the study.This might however be explained by the fact that the majority of them were already receiving physiotherapy at the time this study was initiated.In contrast, higher levels of disability were reported in other previous validation studies [25,29,31].
It is worth noting that factor analysis was not conducted for the original English version of the NPQ in the earlier study [14].However, the potential underlying structure of the NPQ was examined in subsequent studies [26,30,45].Pickering et al. [45] reported a two-factor model and proposed a model of "dysfunction related to general activities" (items 2, 3, 5-9) as the first factor and a model of "neck pain" (items 1 and 4) as the second factor.Although Wlodyka-Demaille et al. [26] found a twofactor model for the French NPQ, the factors, however, could not be categorized.Moreover, the two-factor model obtained by both Pickering et al. [41] and Wlodyka-Demaille et al. [26] could not be confirmed in the Brazilian Portuguese validation of the NPQ by Almeida et al. [30]; however, a one-factor model was confirmed following extensive modifications (i.e., exclusion of items 1, 3, 4 and 5), which led to a short version consisting of only five items (i.e., items 2, 6, 7, 8 and 9) [30].In the present study, we confirmed the Hausa-NPQ as a unidirectional scale after removal of item 9 due to many missing values, and allowing three error terms to covary.The homogeneity of the 8-item Hausa-NPQ is confirmed by the excellent internal consistency as indicated by the Cronbach a value (0.94), which is higher than the range of 0.76-0.88[24,[27][28][29][30][31] reported by previous validation studies (Table 5).Additionally, the item-total correlations (range: 0.85-0.92)were greater than 0.40 indicating positive significant discrimination [46].
As hypothesized, a strong and significant correlation (rho ¼ 0.68) between the Hausa-NPQ and the NPRS was observed, thus establishing a good convergent construct validity.The strong correlation coefficient obtained in our study is in harmony with those obtained by previous similar studies using the NPRS [31] or Visual Analogue Scale (VAS) for pain [24,27,29].In contrast, lower correlation coefficients with the NPRS measured during rest and activity were obtained by the Brazilian Portuguese validation study [30] (Table 5).
Regarding the relative reliability of the Hausa-NPQ, the ICC obtained was adequate (0.86, 3-day retest) and comparable to the Turkish version (ICC ¼ 0.85, 1 to 3-day retest) [24] and higher than the range of 0.63 À 0.84 (3 to 10-day retest) [25][26][27] obtained by other language versions.However, ICC values (range: 0.89 À 0.95, 1 to 10-day retest) [28][29][30][31] greater than the value obtained in our study were reported in the literature (Table 5), which could be ascribed to population variation and the different test-retest intervals used among studies.The low item-by-item correlation (0.29) exhibited by Item 4 (duration of symptoms) indicates that the item may not be measuring the concept of the questionnaire (i.e., neck pain-related disability).This is rather not surprising considering that "duration of symptoms" does not actually indicate a functional or activity limitation.
Although the higher ICC obtained for the Hausa-NPQ suggests excellent reproducibility, however, it might be misleading since it does not account for measurement error.Thus, absolute reliability in terms of SEM and SDC were calculated to determine the precision and sensitivity of the questionnaire, respectively.Of note, only the Brazilian Portuguese validation study [30] reported the SEM and SDC for the NPQ (Table 5).In the present study, the SEM (6.32) and SDC (17.5) were higher than those calculated for the Brazilian Portuguese NPQ [30].Although larger scores of the SEM indicate greater variability and suggest more measurement error, whereas smaller scores indicate lesser variability and suggest high precision, the relatively higher SEM values obtained for the Hausa-NPQ may still guide clinicians when interpreting changes in scores that can be considered reliable or true change after intervention.For example, considering our participants' NPQ baseline  total score of 23.8, we can be 95% confident that the true score lies between 17.4 points and 30.1 points for the given SEM of 6.32.As for our SDC (17.5), it can be interpreted that a change equal to or above 17.5 points can be considered to be a true change beyond measurement error.Furthermore, our SEM is smaller than its Bland-Altman agreement limits (-29.3 to þ 37.1).However, agreement limits calculated for the French version (-8.21 to þ 4.67) [26] were lower than those calculated in our study.Given that the Bland-Altman plot only defines the intervals of agreements, and does not indicate if the agreement is sufficient or suitable [47], it can be inferred that the Hausa-NPQ had a good agreement with minimal proportional bias as revealed by the trend of the mean values and the differences.Moreover, evidence of minimal or no proportional bias was confirmed by a non-significant unstandardized coefficient value close to zero (b ¼ 0.15) for the relationship between the mean and difference of the repeated measurements.This study is without some limitations.First, the use of convenience sampling may not allow the generalization of findings.Secondly, although our test-retest interval was short (3 days) for considerable change to occur, participants' stability or status was not evaluated using the global rating of change scale when assessing the test-retest reliability.Thirdly, we did not examine convergent validity with other neck pain and disability outcome measures, such as the NDI and NPDS.Other aspects of construct validity, for example, divergent and known-group validity were not assessed in the current study.Lastly, responsiveness (sensitivity to change) which is an essential measurement property was not assessed.Nonetheless, the validity and reliability of a scale cannot be fully established by any single study, thus additional studies are needed to establish the validity and responsiveness of the Hausa-NPQ to further support its use in the Hausa culture.

Conclusions
The findings of this study suggest that the Hausa-NPQ was successfully adapted with adequate measurement properties regarding factorial and convergent validity, internal consistency and test-retest reliability.The questionnaire will be valuable in clinical and research settings to assess disability due to neck pain for screening purposes, evaluation of treatment effectiveness, as well as cross-cultural comparisons involving Hausa-speaking individuals with non-specific neck pain.

Figure 2 .
Figure 2. Factor structure of the Hausa-NPQ one-factor model.NPQ: Northwick Park Neck Pain Questionnaire.

Figure 3 .
Figure 3. Bland-Altman plot for test-retest agreement of Hausa-NPQ.Mean scores of test-retest (x-axis) plotted against the difference between test-retest scores (y-axis).NPQ: Northwick Park Neck Pain Questionnaire.

Table 1 .
Demographic and clinical characteristics of the participants.

Table 2 .
General characteristics of the Hausa Northwick Park Neck Pain Questionnaire (n ¼ 92).

Table 3 .
Confirmatory factor analyses of the Hausa Northwick Park Neck Pain Questionnaire models (n ¼ 92).

Table 4 .
Relative and absolute reliability of the Hausa Northwick Park Neck Pain Questionnaire (n ¼ 50).