SPINE Volume 40, Number 9, pp E538-E544 ©2015, Wolters Kluwer Health, Inc. All rights reserved.
HEALTH SERVICES RESEARCH
Validation of the Simplified Chinese Version of the Functional Rating Index for Patients With Nonspecific Neck Pain in Mainland China Xianzhao Wei, MD,* Ximing Xu, MD,* Yongfei Zhao, MD,† Kai Chen, MD,* Fei Wang, MD,* Jianping Fan, MD,* Changwei Yang, MD,* Ziqiang Chen, MD,* Yushu Bai, MD,* Xiaodong Zhu, MD,* and Ming Li, MD*
Study Design. Validation of a self-report questionnaire, Functional Rating Index (FRI). Objective. To evaluate the psychometric properties of the simplified Chinese (SC) version of FRI in patients with nonspecific neck pain (NP). Summary of Background Data. FRI has been cross-culturally validated in few languages with excellent reliability, validity, and clinical utility when it is applied in patients with low back pain. Recently, it has been pointed out that FRI can be employed to assess patients with NP. However, FRI has not been validated in patients with NP in China. Methods. The cross-culturally adapted 10-item SC-FRI was completed by 122 patients with nonspecific NP, along with the Neck Disability Index, Neck Pain and Disability Scale, 36-Item Short Form Health Survey, and Pain Visual Analogue Scale. Psychometric evaluation included score distribution, internal consistency, testretest reliability, and construct validity. Results. SC-FRI attained a high completion rate (96.9%). Each item was scored with a normal distribution without any floor and ceiling effects. The internal consistency and test-retest reliability were good to excellent (Cronbach α coefficient = 0.86; intraclass r = 0.97). Construct validity was confirmed by a strong correlation with Neck Disability Index, Neck Pain and Disability Scale, and Visual Analogue Scale (r = 0.77, 0.78, and 0.86, respectively, all
From the *Department of Orthopaedic Surgery, the Affiliated Changhai Hospital of the Second Military Medical University, Shanghai, China; and †Department of Orthopaedic Surgery, Chinese PLA General Hospital, Beijing, People’s Republic of China. Acknowledgment date: August 9, 2014. First revision date: December 28, 2014. Second revision date: January 21, 2015. Acceptance date: January 21, 2015. The manuscript submitted does not contain information about medical device(s)/drug(s). No funds were received in support of this work. No relevant financial activities outside the submitted work. Address correspondence and reprint requests to Ming Li, MD, and Xiaodong Zhu, MD, Orthopedic Department of Changhai Hospital, Second Military Medical University, No. 168, Changhai Rd, Shanghai, People’s Republic of China, 200433; E-mail: [email protected]; [email protected] DOI: 10.1097/BRS.0000000000000806
E538
P < 0.0001), and with the physical functioning and bodily pain domains (r = −0.69 and −0.66, respectively, both P < 0.0001), and by a weak correlation with the vitality, role-emotional, and mental health domains of 36-Item Short Form Health Survey. Conclusion. SC-FRI showed satisfactory clinical utility, internal consistency, test-retest reliability, and construct validity in Chinesespeaking patients with nonspecific NP. Key words: Functional Rating Index, nonspecific neck pain, reliability, validity. Level of Evidence: 3 Spine 2015;40:E538–E544
N
onspecific neck pain (NP) was defined as simple NP without specific underlying diseases (e.g., herniated disc, rheumatic disease, and malignancy).1 Nonspecific NP can cause pain and functional impairment and has a profound impact on health-related quality of life. The prevalence of nonspecific NP varies in different studies. One study reported 1-year prevalence rate of 48.7% in China.2 During the past decade, increasing importance has been attributed to self-reported outcome measures using diseasespecific questionnaires for patients with musculoskeletal diseases. Several instruments have been developed to assess the pain perception and dysfunction in patients with NP, such as the Neck Disability Index (NDI),3 Neck Pain and Disability Scale (NPDS),4 and Functional Rating Index (FRI).5 Currently, physicians and surgeons choose a scale on the basis of clinical utility, response rate, and the availability of a suitable version. Separate instruments may have to be used for different symptoms, such as back pain and NP, although they may belong to the same clinical entity. A recent study challenged the use of the NDI as the first choice for outcome measures in patients with NP because of its poor content validity.6 To overcome the disadvantages and apply a uniform instrument to patients with back pain and/or NP, Feise and Mneke5 developed an English version of FRI, which contains 10 items stemmed from the item pool of the Oswestry Disability Index and NDI. They demonstrated the reliability and validity of FRI for the evaluation of pain and functional status in patients with NP and/or back pain.
www.spinejournal.com
May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E538
3/30/15 7:54 PM
HEALTH SERVICES RESEARCH The growth in multicenter international research projects has necessitated cross-cultural adaptation and validation of self-administered, disease-specific questionnaires in multiple languages.7,8 Cross-cultural adaptations of FRI into Turkish,9 Brazilian-Portuguese,10,11 and Persian12 have satisfactory psychometric properties when applied to patients with low back pain (LBP). Also, FRI has been cross-culturally adapted into Korean13 and Persian14 for the evaluation of patients with NP. FRI increases the clinical utility, decreases the answering and scoring time, and reduces the efforts for assessing the medical conditions for both patients and doctors. Therefore, FRI has become internationally recognized as a valid outcome measure instrument for patients with LBP or NP.15–17 Previous study demonstrated that a cross-culturally adapted simplified Chinese version of FRI (SC-FRI) was reliable and valid for the assessment of LBP.18 However, the psychometric properties of SC-FRI for the assessment of patients with NP remain unknown. Therefore, the aim of the present study was to validate SC-FRI in a cohort of native Mandarinspeaking patients with non-specific NP in mainland China.
MATERIALS AND METHODS Translation and Cross-cultural Adaptation According to guidelines for the process of cross-cultural adaptation of self-report measures,19,20 we translated and crossculturally adapted the English version of FRI into simplified Chinese. The detailed procedures were described in previous studies.18,21
Participants
Consecutive adult Chinese-speaking patients (>18 yr) with a diagnosis of nonspecific NP for more than 8 weeks who attended the outpatient sections of the Department of Orthopedics, Changhai Hospital of the Second Military Medical University between September 2012 and July 2013 were recruited into the study. Patients were excluded from the study if they reported a history of spine surgery, tumors, traumatic injuries, systemic rheumatologic disease, ankylosing spondylitis, and/or disc herniation demanding surgery, or were unable or unwilling to complete the questionnaires independently. At the time of recruitment, demographic and clinical data were obtained and a physical examination was performed. Radiological evaluations were taken into account whenever possible. FRI and other related Chinese adapted questionnaires were put in 1 copy and distributed to all the participants. Each patient was asked to fill out these 5 questionnaires. The 5 questionnaires are the following described questionnaires, namely, Functional Rating Index, Neck Disability Index, Neck Pain and Disability Scale, 36-Item Short Form Health Survey (SF-36), and Pain Visual Analogue Scale (VAS). The project was approved by the Human Research Ethics Committee of the Second Military Medical University, and each patient signed a written informed consent. Spine
Validation FRI in Chinese • Wei et al
Assessment of Psychometric Properties Functional Rating Index FRI is specifically used for the evaluation of pain perception and function of daily life. It contains 10 items, and each is scored from 0 to 4. The original score is converted to a 100 scale after standardization, which is performed by dividing the sum of individual scores of items answered by the patient by the possible maximum total score and then multiplying by 100 to produce the standardized score. The possible maximum total score refers to the maximum scores of all items that have been answered; for example, it is 40 (10 × 4) if all 10 items have been answered. Thus, the standardized score ranges from 0 (no pain or disability) to 100 (worst pain or disability).5 Neck Disability Index NDI is a modification of Oswestry Disability Index and specifically measures functional limitations due to NP. NDI consists of 10 items, which is scored from 0 (no activity limitations) to 5 (major activity limitations), which are summed to yield a total score. The simplified Chinese version of NDI was first developed in 2010 and showed acceptable reliability and validity in patients with NP.22 Neck Pain and Disability Scale NPDS is one of the most applied instruments for measuring physical and emotional functions in patients with NP. It consists of 20 items, with each item scored from 0 to 5. The total score ranges from 0 to 100, with a higher score indicating greater disability. A simplified Chinese NPDS was validated with satisfactory reliability and validity in our previous study of patients with NP.23 36-Item Short Form Health Survey 36-Item Short Form Health Survey (SF-36) is a health status assessment questionnaire that contains 8 domains: physical functioning (PF), role-physical (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), roleemotional (RE), and mental health (MH).24 Each of the raw subscale scores is transformed to a 100-point scale. SF-36 has already been translated into Chinese and is widely used in many studies.25 Pain Visual Analogue Scale VAS allows patients to rate their pain intensity on a 100-mm line anchored with 2 endpoints labeled “No Pain” (0) and “Worst Possible Pain” (100). It has been commonly used worldwide for its simplicity and good reliability, validity, and responsiveness.26
Acceptability To evaluate the acceptability of SC-FRI, patients were asked about any difficulties they encountered after completing the questionnaire booklet. The data were checked for missing or multiple responses, and the completeness www.spinejournal.com
E539
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E539
3/30/15 7:54 PM
HEALTH SERVICES RESEARCH of SC-FRI was calculated to evaluate the quality of the data. The time taken to complete SC-FRI was also recorded.
Score Distribution The score distribution was evaluated by floor and ceiling effects and the response trend. A floor or ceiling effect exceeding 15% was considered significant, as described previously.27 The response trend was determined by the score distribution of an item. If an item has a mean value close to the extreme scores (the possible highest or lowest score), it would correlate poorly with other items of the same construct.28 Therefore, the standardized values of skewness and kurtosis (Z-skewness) were computed for each item. Any items with a skewness value of more than 1.96 were defined to have a response trend that deviates from a pattern of normal distribution.28,29
Internal Consistency The internal consistency was employed to explore the homogeneity of SC-FRI. Internal consistency was evaluated by calculating the Cronbach α coefficients for the overall SC-FRI when any one of the items was deleted. A Cronbach α coefficient exceeding 0.70 was considered acceptable, whereas more than 0.80 was regarded as good.30,31
Test-Retest Reliability The initial questionnaire was completed in an outpatient room at the hospital, and the second questionnaire was completed either in the outpatient room or at home and returned by mail. The time interval between the 2 questionnaires was 3 to 7 days. An intraclass correlation coefficient (ICC, 2-way random effects model) was calculated to quantify the reproducibility of the test and re-test. An ICC value between 0.60 and 0.80 was considered good reliability, and values higher than 0.80 were considered excellent reliability.32 In addition, within-subject variation and the limits of agreement were examined using the Bland and Altman plot.33
Statistical Analysis The number of patients enrolled in the study was in accordance with the quality criteria for health status questionnaires proposed by Terwee et al,27 who proposed that the sample size should be more than 100 for reliability and validity analysis and more than 50 for test-retest analysis. Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) version 18.0 (SPSS, Chicago, IL). Numerical data are expressed by mean values ± standard deviation. ICC values with 95% confidence intervals are reported. A P value of less than 0.05 was considered statistically significant.
RESULTS Participants A total of 122 patients were recruited and completed the study. Of these patients, 60 were randomly selected to participate in the test-retest analyses. The participants’ demographic and clinical characteristics are summarized in Table 1.
Acceptability Most patients answered SC-FRI quickly and smoothly with a mean time of 67.3±19.7 seconds, which is in accordance to our previous study.18 Patients reported that they did not encounter any problems or difficulties in completing SC-FRI. The completion rate of SC-FRI was 96.9% (1182/1220). Improperly answered items (i.e., items with missing or multiple responses) were found in 3.1% (38/1220) of the cases and were dealt according to the scoring rule of FRI.
Score Distribution SC-FRI scores were well distributed, without any floor and ceiling effects (Table 1). The response trend, as determined by Z-skewness, showed that all the items had a normal distribution. Also, none of the items had a poor item-total correlation, with an r less than 0.30 (Table 2). These findings indicated that all the items could be included in FRI.
Internal Consistency
Validity Content validity was assessed during the translation and cross-cultural adaptation stage by testing the prefinal SC-FRI with patients and incorporating expert opinion.18 Construct validity was described as the orderliness of findings. To be valid, the results had to correlate well with measures of the same construct (convergent validity) while correlating poorly with measures of different constructs (discriminant validity).34 Construct validity of SC-FRI was assessed by calculating the Pearson correlation coefficients of the SC-FRI scores versus scores from the NDI, NPDS, VAS, and the 8 SF-36 domains. Pearson correlation values of r = 0 to 0.25, 0.25 to 0.50, 0.50 to 0.75, and more than 0.75 were considered to indicate poor, fair, moderate to good, and very good to excellent correlation, respectively.35 E540
Validation FRI in Chinese • Wei et al
The Cronbach α coefficient for the overall SC-FRI was 0.86. The Cronbach α coefficients ranged from 0.83 to 0.86 when any one of the items was deleted, indicating good internal consistency (Table 2). The item-total score correlation was high, with Pearson coefficients ranging from 0.48 to 0.81, indicating that each item correlated to the overall SC-FRI (Table 2). We noted that the item-total correlation coefficients were relatively low for items 9 and 10 (0.48 and 0.48, respectively).
Test-Retest Reliability All 60 participants completed the test-retest analysis with an average time interval of 4.4 ± 1.2 days. The ICC value for the SC-FRI was 0.97 (0.95–0.98), indicating excellent test-retest reliability. The Bland-Altman plots showed that there were no significant differences between measures from the 2 test
www.spinejournal.com
May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E540
3/30/15 7:54 PM
HEALTH SERVICES RESEARCH
Validation FRI in Chinese • Wei et al
Validity
TABLE 1. Demographic and Clinical
Characteristics and the Score Distribution of Patients With Neck Pain (n = 122)
Variables
Values
Sex, n (%) Female
73 (59.3%)
Male
49 (40.7%)
Age in years, mean ± SD
50.1 ± 9.1
Pain duration in weeks, mean ± SD
18.3 ± 6.4
DISCUSSION
Occupation, n (%) Office worker
34 (27.9%)
Manual worker
28 (22.9%)
Housewife
31 (25.4%)
Retired
29 (23.8%)
Education, n (%) Elementary school
44 (36.1%)
Middle school
30 (24.6%)
High school
27 (22.1%)
University
21 (17.2%)
VAS in mm, mean ± SD
61.8 ± 20.8
SC-FRI in score, mean ± SD
41.8 ± 15.5
SC-NDI in score, mean ± SD
37.8±16.1
SC-NPDS in score, mean ± SD
42.1 ± 14.3
SC-FRI floor effects, %
0
SC-FRI ceiling effects, %
0
SF-36 PF in score, mean ± SD
55.0 ± 18.5
SF-36 RP in score, mean ± SD
36.7 ± 28.1
SF-36 BP in score, mean ± SD
38.2 ± 15.9
SF-36 GH in score, mean ± SD
76.7 ± 12.4
SF-36 VT in score, mean ± SD
62.3 ± 16.9
SF-36 SF in score, mean ± SD
64.0 ± 18.2
SF-36 RE in score, mean ± SD
43.7 ± 28.8
SF-36 MH in score, mean ± SD
75.8 ± 8.7
VAS indicates visual analogue scale; SC-FRI, simplified Chinese version of functional rating index; SC-NDI, simplified Chinese version of neck disability index; SC-NPDS, simplified Chinese version of neck pain and disability scale; SF-36, 36-Item Short Form Health Survey; PF, physical functioning; RP, role-physical; BP, bodily pain; GH, general health; VT, Vitality; SF, social functioning; RE, role-emotional; MH, mental health.
sessions for the individual patients in terms of the overall means of the 2 sessions (Figure 1). Thus, no systematic bias between the test and retest SC-FRI scores was found, indicating good test-retest agreement of the SC-FRI. Spine
Pearson correlation coefficients for SC-FRI and NDI, NPDS, VAS, and SF-36 are shown in Table 3. SC-FRI had very good to excellent correlations with NDI, NPDS, and VAS (r = 0.77, 0.78, and 0.86, respectively, all P < 0.0001). SC-FRI also had moderate to good correlations with the PF, BP, GH, and SF domains of the SF-36 (r = −0.69 and −0.66, = −0.51 and −0.60, respectively, all P < 0.0001). In contrast, SC-FRI correlated fairly with the RP (r = −0.47, P < 0.0001), VT (r = −0.31, P < 0.0001), and MH (r = −0.26, P = 0.004) domains and correlated poorly with the RE (r = −0.18, P = 0.050) domain of SF-36. Thus, SC-FRI had good construct validity for patients with NP. The present study demonstrated that SC-FRI was reliable and valid for assessing pain perception and functional status in patients with nonspecific NP. In addition, the psychometric properties were similar to other adapted versions such as Korean 13 and Persian 14 used in patients with NP (Table 4). These results were also similar to those obtained when the Chinese version was first cross-culturally adapted in an LBP population in China.18 Recent studies have reported that FRI is easy to administer and complete. A Korean study found that FRI had a much higher response rate than that of NDI and NPDS.13 In the present study, the completion rate was 96% and the administration time was 67.3 ± 19.7 seconds. SC-FRI exhibited no ceiling and floor effects when administrated to nonspecific patients with NP. Item response trend analysis of SC-FRI revealed that no item could be omitted because of poor item-total correlation or skewed score distribution. These results are in accordance with the original study5 and subsequent studies.9–14 A measure assessing a particular construct consists of items that are homogeneous. Cronbach α and item-total correlation coefficients were used to test the internal consistency of SC-FRI. As shown in Table 4, the Cronbach α coefficient of SC-FRI (0.86) was similar with that of the Korean and Persian versions (0.88 and 0.88, respectively) in patients with NP, as well as other studies in which SC-FRI was applied to assess patients with LBP.17 The Cronbach α coefficients of SC-FRI when an item was deleted were also high (0.83–0.86). Although the last 2 items (walking and standing) had lower item-total correlations, the coefficients were more than 0.30 and thus they could not be excluded from the construct. Test-retest reliability indicates whether the instrument can remain stable over time. ICC was calculated to explore testretest reliability of SC-FRI. The ICC value of SC-FRI (0.97) in the present study was similar to that in the Persian study (0.96) and greater than that in the Korean study (0.86).13 In studies assessing LBP, FRI has excellent internal consistency and test-retest reliability,17 suggesting that FRI is reliable across patients with NP and/or LBP. Various instruments have been used to evaluate the construct validity of different language versions of FRI, including the Oswestry Disability Index, SF-36, SF-12, and VAS in patients with LBP, and NDI, NPDS, global perceived effect, www.spinejournal.com
E541
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E541
3/30/15 7:54 PM
HEALTH SERVICES RESEARCH
Validation FRI in Chinese • Wei et al
TABLE 2. The Response Trend, Item-Total Score Pearson Correlations, and Cronbach α Coefficients
(When 1 Item Is Deleted From SC-FRI) Z-Skewness
Item-Total Score Correlation (r)
Cronbach α Coefficient (r)
−0.38
0.64
0.86
Sleeping
0.50
0.73
0.84
Personal care
0.11
0.77
0.84
Travel
0.63
0.81
0.83
Work
0.81
0.66
0.85
Recreation
0.70
0.79
0.84
Frequency of pain
0.46
0.72
0.85
−0.21
0.58
0.86
Walking
0.16
0.48
0.86
Standing
0.25
0.48
0.86
Item in SC-FRI Pain intensity
Lifting
SC-FRI indicates the simplified Chinese version of Functional Rating Index.
SF-36, and VAS in patients with NP. Here, the construct validity of SC-FRI in patients with nonspecific NP was evaluated by NDI and NPDS, in addition to SF-36 and VAS. SCFRI correlated strongly with the PF and BP domains of the SF-36, indicating that SC-FRI has good convergent validity. Meanwhile, there were low Pearson correlation coefficients for SC-FRI and the VT, RE, and MH domains of SF-36, demonstrating discriminant validity of SC-FRI. Our findings were in accordance with the original English study, which found that FRI had a high correlation with the SF-12 PF and a low correlation with the SF-12 MH.5 In the present study, there was a strong correlation between SC-FRI and
Figure 1. Bland-Altman plot for test-retest agreement of the Simplified Chinese version of Functional Rating Index in patients with neck pain (n = 60). There are no differences in the mean total scores measured at the 2 test sessions. The dashed line indicates the 95% (±1.96 standard deviation) limits of agreement.
E542
NDI and NPDS, which are comparable with the results of the Persian study.14 In addition, the Pearson correlation coefficients for SC-FRI and the BP and PF domains of SF-36 and VAS were greater than the correlation coefficients for NDI and NPDS and the BP and PF domains of SF-36 and VAS (Table 3). These findings indicate that SC-FRI has good construct validity, as evaluated by NDI, NPDS, SF-36, and VAS. With the globalization of clinical research, self-reported outcomes assessed by disease-specific instruments are attracting increasing attention from orthopedic surgeons in mainland China. FRI is a valid, reliable, and internationally used scale that is designed to subjectively assess patients with NP and/or LBP and dysfunction.9–14 As demonstrated in our previous study in patients with LBP, the present study further provides strong evidence that SC-FRI is similarly reliable and valid for evaluating pain and functional status in Chinese patients with nonspecific NP. There are a few limitations in present study. First, although patients enrolled in this study came from different areas of mainland China, the study cohort may not fully represent the Chinese population, and thus there may be a patient selection bias due to the limited sample size. Second, SC-FRI was intended primarily for the use in mainland China and may not be valid for Chinese speakers in Hong Kong, Taiwan, and other Chinese-speaking locations. Therefore, SC-FRI needs to be further validated in different areas, especially where simplified Chinese is the standard written form. Moreover, as there is no report on validation of FRI in patients with NP in geographical regions of Europe and North America, the value of this construct should be further determined in these regions. Third, the responsiveness of SC-FRI was not assessed in patients with nonspecific NP, as well as in patients with back pain. Thus, it is important to validate SC-FRI in a larger population in mainland China in order to probe the responsiveness of this instrument in patients with NP or LBP.
www.spinejournal.com
May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E542
3/30/15 7:54 PM
HEALTH SERVICES RESEARCH
Validation FRI in Chinese • Wei et al
TABLE 3. Construct Validity Evaluation of SC-FRI Against SF-36, NDI, NPDS, and VAS by Pearson
Correlations Analysis
SF-36 Subscales
SC-FRI
SC-NDI
SC-NPDS
PF
−0.69 (P < 0.0001)
−0.58 (P < 0.0001)
−0.58 (P < 0.0001)
RP
−0.47 (P < 0.0001)
−0.40 (P < 0.0001)
−0.40 (P < 0.0001)
BP
−0.66 (P < 0.0001)
−0.48 (P < 0.0001)
−0.66 (P < 0.0001)
GH
−0.51 (P < 0.0001)
−0.45 (P < 0.0001)
−0.39 (P < 0.0001)
VT
−0.31 (P < 0.0001)
−0.40 (P < 0.0001)
−0.34 (P < 0.0001)
SF
−0.60 (P < 0.0001)
−0.50 (P < 0.0001)
−0.47 (P < 0.0001)
RE
−0.18 (P = 0.050)
−0.079 (P = 0.386)
−0.075 (P = 0.414)
MH
−0.26 (P = 0.004)
−0.23 (P = 0.011)
−0.22 (P = 0.015)
SC-NDI
0.77 (P < 0.0001)
NA
0.67 (P < 0.0001)
SC-NPDS
0.78 (P < 0.0001)
0.68 (P < 0.0001)
NA
VAS
0.86 (P < 0.0001)
0.70 (P < 0.0001)
0.74 (P < 0.0001)
Numbers in the table are Pearson correlation coefficients (r), with P value within parentheses. Pearson correlation values of r = 0–0.25, 0.25–0.50, 0.50–0.75, and > 0.75 indicate poor, fair, moderate to good, and very good to excellent correlation, respectively. SF-36 indicates 36-Item Short Form Health Survey; SC-FRI, the simplified Chinese version of Functional Rating Index; SC-NDI, simplified Chinese version of neck disability index; SC-NPDS, simplified Chinese version of neck and pain disability scale; PF, physical functioning; RP, role-physical; BP, bodily pain; GH, general health; VT, vitality; SF, social functioning; RE, role-emotional; MH, mental health; VAS, visual analogue scale; NA, not applicable.
TABLE 4. Summary of Evidence on Reliability and Validity of Various Versions of the Functional
Rating Index for Neck Pain
Internal Consistency Versions Lee et al
13
Ansari et al
14
Present study
Test-Retest Reliability
Construct Validity
N
Cronbach α Coefficient
N
ICC
Tested Construct
Pearson Correlation (r)
261
0.88
40
0.86
GPE
−0.46
100
0.88
50
0.96
NDI
0.72
NPDS
0.63
VAS
0.75
NDI
0.77
NPDS
0.78
VAS
0.86
SF-36 PF
−0.69
SF-36 RP
−0.47
SF-36 BP
−0.66
SF-36 GH
−0.51
SF-36 VT
−0.31
SF-36 SF
−0.60
SF-36 RE
−0.18
SF-36 MH
−0.26
122
0.86
60
0.97
ICC indicates intraclass correlation coefficient; GPE, global perceived effect; NDI, neck disability index; NPDS, neck and pain disability scale; VAS, visual analogue scale; PF, physical functioning; RP, role limitations due to physical health; BP, bodily pain; GH, general health perception; VT, vitality; SF, social functioning; RE, role limitations due to emotional problems; MH, mental health.
Spine
www.spinejournal.com
E543
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E543
3/30/15 7:54 PM
HEALTH SERVICES RESEARCH
Validation FRI in Chinese • Wei et al
CONCLUSION The cross-culturally adapted SC-FRI showed good clinical utility, internal consistency, test-retest reliability, and construct validity when administrated to Chinese patients with nonspecific NP. This work will promote application of SC-FRI to evaluate health status in patients with NP, especially those with both NP and LBP in mainland China.
➢ Key Points Functional Rating Index (FRI) is culturally adapted to simplified Chinese for measurement of nonspecific neck pain in mainland China. The simplified Chinese FRI is easy to understand and quick to complete, so it is practical for wide use in clinical practice. The simplified Chinese FRI shows satisfactory internal consistency, test-retest reliability, and construct validity in simplified Chinese-speaking patients with nonspecific neck pain in mainland China.
Acknowledgments Xianzhao Wei, MD, Ximing Xu, MD, Yongfei Zhao, MD, and Kai Chen, MD, contributed equally to this study.
References
1. Tsakitzidis G, Remmen R, Peremans L, et al. Non-specific neck pain: diagnosis and treatment. KCE Reports 119C Good Clinical Practice (GCP). Brussels, Belgium: Belgian Health Care Knowledge Centre (KCE); 2009. 2. Yue P, Liu F, Li L. Neck/shoulder pain and low back pain among school teachers in China, prevalence and risk factors. BMC Public Health 2012;12:789. 3. Vernon H, Mior S. The Neck Disability Index: a study of reliability and validity. J Manipulative Physiol Ther 1991;14:409–15. 4. Wheeler AH, Goolkasian P, Baird AC, et al. Development of the Neck Pain and Disability Scale. Item analysis, face, and criterionrelated validity. Spine (Phila Pa 1976) 1999;24:1290–4. 5. Feise RJ, Michael Menke J. Functional rating index: a new valid and reliable instrument to measure the magnitude of clinical change in spinal conditions. Spine (Phila Pa 1976) 2001;26:78–86; discussion 7. 6. Ailliet L, Knol DL, Rubinstein SM, et al. Definition of the construct to be measured is a prerequisite for the assessment of validity. The Neck Disability Index as an example. J Clin Epidemiol 2013;66:775–82; quiz 782.e1–2. 7. Bullinger M, Alonso J, Apolone G, et al. Translating health status questionnaires and evaluating their quality: the IQOLA Project approach. International Quality of Life Assessment. J Clin Epidemiol 1998;51:913–23. 8. Uysal-Bozkir Ö, Parlevliet JL, de Rooij SE. Insufficient crosscultural adaptations and psychometric properties for many translated health assessment scales: a systematic review. J Clin Epidemiol 2013;66:608–18. 9. Ceran F, Ozcan A. The relationship of the Functional Rating Index with disability, pain, and quality of life in patients with low back pain. Med Sci Monit 2006;12:CR435–9. 10. Costa LO, Maher CG, Latimer J, et al. Clinimetric testing of three self-report outcome measures for low back pain
E544
11.
12. 13. 14. 15. 16. 17. 18. 19. 20.
21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35.
patients in Brazil: which one is the best? Spine (Phila Pa 1976) 2008;33:2459–63. Costa LO, Maher CG, Latimer J, et al. Psychometric characteristics of the Brazilian-Portuguese versions of the Functional Rating Index and the Roland Morris Disability Questionnaire. Spine (Phila Pa 1976) 2007;32:1902–7. Ansari NN, Feise RJ, Naghdi S, et al. The functional rating index: reliability and validity of the Persian language version in patients with low back pain. Spine (Phila Pa 1976) 2011;36:E1573–7. Lee H, Nicholson LL, Adams RD, et al. Development and psychometric testing of Korean language versions of 4 neck pain and disability questionnaires. Spine (Phila Pa 1976) 2006;31:1841–5. Ansari NN, Feise RJ, Naghdi S, et al. The functional rating index: reliability and validity of the Persian language version in patients with neck pain. Spine (Phila Pa 1976) 2012;37:E844–8. Rebbeck T, Sindhusake D, Cameron ID, et al. A prospective cohort study of health outcomes following whiplash associated disorders in an Australian population. Inj Prev 2006;12:93–8. Rebbeck TJ, Refshauge KM, Maher CG, et al. Evaluation of the core outcome measure in whiplash. Spine (Phila Pa 1976) 2007;32:696–702. Feise RJ, Menke JM. Functional Rating Index: literature review. Med Sci Monit 2010;16:RA25–36. Wei X, Chen Z, Bai Y, et al. Validation of the simplified Chinese version of the functional rating index for patients with low back pain. Spine (Phila Pa 1976) 2012;37:1602–8. Guillemin F, Bombardier C, Beaton D. Cross-cultural adaptation of health-related quality of life measures: literature review and proposed guidelines. J Clin Epidemiol 1993;46:1417–32. Wild D, Grove A, Martin M, et al. Principles of good practice for the translation and cultural adaptation process for patient-reported outcomes (PRO) measures: report of the ISPOR task force for translation and cultural adaptation. Value Health 2005;8:94–104. Wei X, Yi H, Wu B, et al. A valid cross-culturally adapted simplified Chinese version of the Quebec Back Pain Disability Scale. J Clin Epidemiol 2012;65:1321–8. Wu S, Ma C, Mai M, et al. Translation and validation study of Chinese versions of the neck disability index and the neck pain and disability scale. Spine (Phila Pa 1976) 2010;35:1575–9. Chen Z, Zhao Y, Wang C, et al. An adapted Chinese version of Neck Pain and Disability Scale: validity and reliability. Spine (Phila Pa 1976) 2011;36:1322–7. Hays RD, Sherbourne CD, Mazel RM. The RAND 36-item health survey 1.0. Health Econ 1993;2:217–27. Li L, Wang HM, Shen Y. Chinese SF-36 Health Survey: translation, cultural adaptation, validation, and normalisation. J Epidemiol Community Health 2003;57:259–63. Kane RL, Bershadsky B, Rockwood T, et al. Visual Analog Scale pain reporting was standardized. J Clin Epidemiol 2005;58:618–23. Terwee CB, Bot SD, de Boer MR, et al. Quality criteria were proposed for measurement properties of health status questionnaires. J Clin Epidemiol 2007;60:34–42. Frich LH, Noergaard PM, Brorson S. Validation of the Danish version of Oxford Shoulder Score. Dan Med Bull 2011;58:A4335. Field A. Discovering Statistics Using SPSS. London: Sage Publications; 2009. Barten JA, Pisters MF, Huisman PA, et al. Measurement properties of patient-specific instruments measuring physical function. J Clin Epidemiol 2012;65:590–601. Fayers PM, Machin D. Quality of Life: Assessment, Analysis, and Interpretation. Chichester, England: John Wiley; 2000. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420–8. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10. Streiner DL, Norman GR. Health measurement scales. A Practical Guide to Their Development and Use. New York, NY: Oxford University Press; 2003. Dawson B, Trapp RG. Basic and Clinical Biostatistics. 3rd ed. New York, NY: Lange Medical Books; 2001.
www.spinejournal.com
May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E544
3/30/15 7:54 PM
Copyright of Spine is the property of Lippincott Williams & Wilkins and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
HEALTH SERVICES RESEARCH
Validation of the Simplified Chinese Version of the Functional Rating Index for Patients With Nonspecific Neck Pain in Mainland China Xianzhao Wei, MD,* Ximing Xu, MD,* Yongfei Zhao, MD,† Kai Chen, MD,* Fei Wang, MD,* Jianping Fan, MD,* Changwei Yang, MD,* Ziqiang Chen, MD,* Yushu Bai, MD,* Xiaodong Zhu, MD,* and Ming Li, MD*
Study Design. Validation of a self-report questionnaire, Functional Rating Index (FRI). Objective. To evaluate the psychometric properties of the simplified Chinese (SC) version of FRI in patients with nonspecific neck pain (NP). Summary of Background Data. FRI has been cross-culturally validated in few languages with excellent reliability, validity, and clinical utility when it is applied in patients with low back pain. Recently, it has been pointed out that FRI can be employed to assess patients with NP. However, FRI has not been validated in patients with NP in China. Methods. The cross-culturally adapted 10-item SC-FRI was completed by 122 patients with nonspecific NP, along with the Neck Disability Index, Neck Pain and Disability Scale, 36-Item Short Form Health Survey, and Pain Visual Analogue Scale. Psychometric evaluation included score distribution, internal consistency, testretest reliability, and construct validity. Results. SC-FRI attained a high completion rate (96.9%). Each item was scored with a normal distribution without any floor and ceiling effects. The internal consistency and test-retest reliability were good to excellent (Cronbach α coefficient = 0.86; intraclass r = 0.97). Construct validity was confirmed by a strong correlation with Neck Disability Index, Neck Pain and Disability Scale, and Visual Analogue Scale (r = 0.77, 0.78, and 0.86, respectively, all
From the *Department of Orthopaedic Surgery, the Affiliated Changhai Hospital of the Second Military Medical University, Shanghai, China; and †Department of Orthopaedic Surgery, Chinese PLA General Hospital, Beijing, People’s Republic of China. Acknowledgment date: August 9, 2014. First revision date: December 28, 2014. Second revision date: January 21, 2015. Acceptance date: January 21, 2015. The manuscript submitted does not contain information about medical device(s)/drug(s). No funds were received in support of this work. No relevant financial activities outside the submitted work. Address correspondence and reprint requests to Ming Li, MD, and Xiaodong Zhu, MD, Orthopedic Department of Changhai Hospital, Second Military Medical University, No. 168, Changhai Rd, Shanghai, People’s Republic of China, 200433; E-mail: [email protected]; [email protected] DOI: 10.1097/BRS.0000000000000806
E538
P < 0.0001), and with the physical functioning and bodily pain domains (r = −0.69 and −0.66, respectively, both P < 0.0001), and by a weak correlation with the vitality, role-emotional, and mental health domains of 36-Item Short Form Health Survey. Conclusion. SC-FRI showed satisfactory clinical utility, internal consistency, test-retest reliability, and construct validity in Chinesespeaking patients with nonspecific NP. Key words: Functional Rating Index, nonspecific neck pain, reliability, validity. Level of Evidence: 3 Spine 2015;40:E538–E544
N
onspecific neck pain (NP) was defined as simple NP without specific underlying diseases (e.g., herniated disc, rheumatic disease, and malignancy).1 Nonspecific NP can cause pain and functional impairment and has a profound impact on health-related quality of life. The prevalence of nonspecific NP varies in different studies. One study reported 1-year prevalence rate of 48.7% in China.2 During the past decade, increasing importance has been attributed to self-reported outcome measures using diseasespecific questionnaires for patients with musculoskeletal diseases. Several instruments have been developed to assess the pain perception and dysfunction in patients with NP, such as the Neck Disability Index (NDI),3 Neck Pain and Disability Scale (NPDS),4 and Functional Rating Index (FRI).5 Currently, physicians and surgeons choose a scale on the basis of clinical utility, response rate, and the availability of a suitable version. Separate instruments may have to be used for different symptoms, such as back pain and NP, although they may belong to the same clinical entity. A recent study challenged the use of the NDI as the first choice for outcome measures in patients with NP because of its poor content validity.6 To overcome the disadvantages and apply a uniform instrument to patients with back pain and/or NP, Feise and Mneke5 developed an English version of FRI, which contains 10 items stemmed from the item pool of the Oswestry Disability Index and NDI. They demonstrated the reliability and validity of FRI for the evaluation of pain and functional status in patients with NP and/or back pain.
www.spinejournal.com
May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E538
3/30/15 7:54 PM
HEALTH SERVICES RESEARCH The growth in multicenter international research projects has necessitated cross-cultural adaptation and validation of self-administered, disease-specific questionnaires in multiple languages.7,8 Cross-cultural adaptations of FRI into Turkish,9 Brazilian-Portuguese,10,11 and Persian12 have satisfactory psychometric properties when applied to patients with low back pain (LBP). Also, FRI has been cross-culturally adapted into Korean13 and Persian14 for the evaluation of patients with NP. FRI increases the clinical utility, decreases the answering and scoring time, and reduces the efforts for assessing the medical conditions for both patients and doctors. Therefore, FRI has become internationally recognized as a valid outcome measure instrument for patients with LBP or NP.15–17 Previous study demonstrated that a cross-culturally adapted simplified Chinese version of FRI (SC-FRI) was reliable and valid for the assessment of LBP.18 However, the psychometric properties of SC-FRI for the assessment of patients with NP remain unknown. Therefore, the aim of the present study was to validate SC-FRI in a cohort of native Mandarinspeaking patients with non-specific NP in mainland China.
MATERIALS AND METHODS Translation and Cross-cultural Adaptation According to guidelines for the process of cross-cultural adaptation of self-report measures,19,20 we translated and crossculturally adapted the English version of FRI into simplified Chinese. The detailed procedures were described in previous studies.18,21
Participants
Consecutive adult Chinese-speaking patients (>18 yr) with a diagnosis of nonspecific NP for more than 8 weeks who attended the outpatient sections of the Department of Orthopedics, Changhai Hospital of the Second Military Medical University between September 2012 and July 2013 were recruited into the study. Patients were excluded from the study if they reported a history of spine surgery, tumors, traumatic injuries, systemic rheumatologic disease, ankylosing spondylitis, and/or disc herniation demanding surgery, or were unable or unwilling to complete the questionnaires independently. At the time of recruitment, demographic and clinical data were obtained and a physical examination was performed. Radiological evaluations were taken into account whenever possible. FRI and other related Chinese adapted questionnaires were put in 1 copy and distributed to all the participants. Each patient was asked to fill out these 5 questionnaires. The 5 questionnaires are the following described questionnaires, namely, Functional Rating Index, Neck Disability Index, Neck Pain and Disability Scale, 36-Item Short Form Health Survey (SF-36), and Pain Visual Analogue Scale (VAS). The project was approved by the Human Research Ethics Committee of the Second Military Medical University, and each patient signed a written informed consent. Spine
Validation FRI in Chinese • Wei et al
Assessment of Psychometric Properties Functional Rating Index FRI is specifically used for the evaluation of pain perception and function of daily life. It contains 10 items, and each is scored from 0 to 4. The original score is converted to a 100 scale after standardization, which is performed by dividing the sum of individual scores of items answered by the patient by the possible maximum total score and then multiplying by 100 to produce the standardized score. The possible maximum total score refers to the maximum scores of all items that have been answered; for example, it is 40 (10 × 4) if all 10 items have been answered. Thus, the standardized score ranges from 0 (no pain or disability) to 100 (worst pain or disability).5 Neck Disability Index NDI is a modification of Oswestry Disability Index and specifically measures functional limitations due to NP. NDI consists of 10 items, which is scored from 0 (no activity limitations) to 5 (major activity limitations), which are summed to yield a total score. The simplified Chinese version of NDI was first developed in 2010 and showed acceptable reliability and validity in patients with NP.22 Neck Pain and Disability Scale NPDS is one of the most applied instruments for measuring physical and emotional functions in patients with NP. It consists of 20 items, with each item scored from 0 to 5. The total score ranges from 0 to 100, with a higher score indicating greater disability. A simplified Chinese NPDS was validated with satisfactory reliability and validity in our previous study of patients with NP.23 36-Item Short Form Health Survey 36-Item Short Form Health Survey (SF-36) is a health status assessment questionnaire that contains 8 domains: physical functioning (PF), role-physical (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), roleemotional (RE), and mental health (MH).24 Each of the raw subscale scores is transformed to a 100-point scale. SF-36 has already been translated into Chinese and is widely used in many studies.25 Pain Visual Analogue Scale VAS allows patients to rate their pain intensity on a 100-mm line anchored with 2 endpoints labeled “No Pain” (0) and “Worst Possible Pain” (100). It has been commonly used worldwide for its simplicity and good reliability, validity, and responsiveness.26
Acceptability To evaluate the acceptability of SC-FRI, patients were asked about any difficulties they encountered after completing the questionnaire booklet. The data were checked for missing or multiple responses, and the completeness www.spinejournal.com
E539
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E539
3/30/15 7:54 PM
HEALTH SERVICES RESEARCH of SC-FRI was calculated to evaluate the quality of the data. The time taken to complete SC-FRI was also recorded.
Score Distribution The score distribution was evaluated by floor and ceiling effects and the response trend. A floor or ceiling effect exceeding 15% was considered significant, as described previously.27 The response trend was determined by the score distribution of an item. If an item has a mean value close to the extreme scores (the possible highest or lowest score), it would correlate poorly with other items of the same construct.28 Therefore, the standardized values of skewness and kurtosis (Z-skewness) were computed for each item. Any items with a skewness value of more than 1.96 were defined to have a response trend that deviates from a pattern of normal distribution.28,29
Internal Consistency The internal consistency was employed to explore the homogeneity of SC-FRI. Internal consistency was evaluated by calculating the Cronbach α coefficients for the overall SC-FRI when any one of the items was deleted. A Cronbach α coefficient exceeding 0.70 was considered acceptable, whereas more than 0.80 was regarded as good.30,31
Test-Retest Reliability The initial questionnaire was completed in an outpatient room at the hospital, and the second questionnaire was completed either in the outpatient room or at home and returned by mail. The time interval between the 2 questionnaires was 3 to 7 days. An intraclass correlation coefficient (ICC, 2-way random effects model) was calculated to quantify the reproducibility of the test and re-test. An ICC value between 0.60 and 0.80 was considered good reliability, and values higher than 0.80 were considered excellent reliability.32 In addition, within-subject variation and the limits of agreement were examined using the Bland and Altman plot.33
Statistical Analysis The number of patients enrolled in the study was in accordance with the quality criteria for health status questionnaires proposed by Terwee et al,27 who proposed that the sample size should be more than 100 for reliability and validity analysis and more than 50 for test-retest analysis. Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) version 18.0 (SPSS, Chicago, IL). Numerical data are expressed by mean values ± standard deviation. ICC values with 95% confidence intervals are reported. A P value of less than 0.05 was considered statistically significant.
RESULTS Participants A total of 122 patients were recruited and completed the study. Of these patients, 60 were randomly selected to participate in the test-retest analyses. The participants’ demographic and clinical characteristics are summarized in Table 1.
Acceptability Most patients answered SC-FRI quickly and smoothly with a mean time of 67.3±19.7 seconds, which is in accordance to our previous study.18 Patients reported that they did not encounter any problems or difficulties in completing SC-FRI. The completion rate of SC-FRI was 96.9% (1182/1220). Improperly answered items (i.e., items with missing or multiple responses) were found in 3.1% (38/1220) of the cases and were dealt according to the scoring rule of FRI.
Score Distribution SC-FRI scores were well distributed, without any floor and ceiling effects (Table 1). The response trend, as determined by Z-skewness, showed that all the items had a normal distribution. Also, none of the items had a poor item-total correlation, with an r less than 0.30 (Table 2). These findings indicated that all the items could be included in FRI.
Internal Consistency
Validity Content validity was assessed during the translation and cross-cultural adaptation stage by testing the prefinal SC-FRI with patients and incorporating expert opinion.18 Construct validity was described as the orderliness of findings. To be valid, the results had to correlate well with measures of the same construct (convergent validity) while correlating poorly with measures of different constructs (discriminant validity).34 Construct validity of SC-FRI was assessed by calculating the Pearson correlation coefficients of the SC-FRI scores versus scores from the NDI, NPDS, VAS, and the 8 SF-36 domains. Pearson correlation values of r = 0 to 0.25, 0.25 to 0.50, 0.50 to 0.75, and more than 0.75 were considered to indicate poor, fair, moderate to good, and very good to excellent correlation, respectively.35 E540
Validation FRI in Chinese • Wei et al
The Cronbach α coefficient for the overall SC-FRI was 0.86. The Cronbach α coefficients ranged from 0.83 to 0.86 when any one of the items was deleted, indicating good internal consistency (Table 2). The item-total score correlation was high, with Pearson coefficients ranging from 0.48 to 0.81, indicating that each item correlated to the overall SC-FRI (Table 2). We noted that the item-total correlation coefficients were relatively low for items 9 and 10 (0.48 and 0.48, respectively).
Test-Retest Reliability All 60 participants completed the test-retest analysis with an average time interval of 4.4 ± 1.2 days. The ICC value for the SC-FRI was 0.97 (0.95–0.98), indicating excellent test-retest reliability. The Bland-Altman plots showed that there were no significant differences between measures from the 2 test
www.spinejournal.com
May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E540
3/30/15 7:54 PM
HEALTH SERVICES RESEARCH
Validation FRI in Chinese • Wei et al
Validity
TABLE 1. Demographic and Clinical
Characteristics and the Score Distribution of Patients With Neck Pain (n = 122)
Variables
Values
Sex, n (%) Female
73 (59.3%)
Male
49 (40.7%)
Age in years, mean ± SD
50.1 ± 9.1
Pain duration in weeks, mean ± SD
18.3 ± 6.4
DISCUSSION
Occupation, n (%) Office worker
34 (27.9%)
Manual worker
28 (22.9%)
Housewife
31 (25.4%)
Retired
29 (23.8%)
Education, n (%) Elementary school
44 (36.1%)
Middle school
30 (24.6%)
High school
27 (22.1%)
University
21 (17.2%)
VAS in mm, mean ± SD
61.8 ± 20.8
SC-FRI in score, mean ± SD
41.8 ± 15.5
SC-NDI in score, mean ± SD
37.8±16.1
SC-NPDS in score, mean ± SD
42.1 ± 14.3
SC-FRI floor effects, %
0
SC-FRI ceiling effects, %
0
SF-36 PF in score, mean ± SD
55.0 ± 18.5
SF-36 RP in score, mean ± SD
36.7 ± 28.1
SF-36 BP in score, mean ± SD
38.2 ± 15.9
SF-36 GH in score, mean ± SD
76.7 ± 12.4
SF-36 VT in score, mean ± SD
62.3 ± 16.9
SF-36 SF in score, mean ± SD
64.0 ± 18.2
SF-36 RE in score, mean ± SD
43.7 ± 28.8
SF-36 MH in score, mean ± SD
75.8 ± 8.7
VAS indicates visual analogue scale; SC-FRI, simplified Chinese version of functional rating index; SC-NDI, simplified Chinese version of neck disability index; SC-NPDS, simplified Chinese version of neck pain and disability scale; SF-36, 36-Item Short Form Health Survey; PF, physical functioning; RP, role-physical; BP, bodily pain; GH, general health; VT, Vitality; SF, social functioning; RE, role-emotional; MH, mental health.
sessions for the individual patients in terms of the overall means of the 2 sessions (Figure 1). Thus, no systematic bias between the test and retest SC-FRI scores was found, indicating good test-retest agreement of the SC-FRI. Spine
Pearson correlation coefficients for SC-FRI and NDI, NPDS, VAS, and SF-36 are shown in Table 3. SC-FRI had very good to excellent correlations with NDI, NPDS, and VAS (r = 0.77, 0.78, and 0.86, respectively, all P < 0.0001). SC-FRI also had moderate to good correlations with the PF, BP, GH, and SF domains of the SF-36 (r = −0.69 and −0.66, = −0.51 and −0.60, respectively, all P < 0.0001). In contrast, SC-FRI correlated fairly with the RP (r = −0.47, P < 0.0001), VT (r = −0.31, P < 0.0001), and MH (r = −0.26, P = 0.004) domains and correlated poorly with the RE (r = −0.18, P = 0.050) domain of SF-36. Thus, SC-FRI had good construct validity for patients with NP. The present study demonstrated that SC-FRI was reliable and valid for assessing pain perception and functional status in patients with nonspecific NP. In addition, the psychometric properties were similar to other adapted versions such as Korean 13 and Persian 14 used in patients with NP (Table 4). These results were also similar to those obtained when the Chinese version was first cross-culturally adapted in an LBP population in China.18 Recent studies have reported that FRI is easy to administer and complete. A Korean study found that FRI had a much higher response rate than that of NDI and NPDS.13 In the present study, the completion rate was 96% and the administration time was 67.3 ± 19.7 seconds. SC-FRI exhibited no ceiling and floor effects when administrated to nonspecific patients with NP. Item response trend analysis of SC-FRI revealed that no item could be omitted because of poor item-total correlation or skewed score distribution. These results are in accordance with the original study5 and subsequent studies.9–14 A measure assessing a particular construct consists of items that are homogeneous. Cronbach α and item-total correlation coefficients were used to test the internal consistency of SC-FRI. As shown in Table 4, the Cronbach α coefficient of SC-FRI (0.86) was similar with that of the Korean and Persian versions (0.88 and 0.88, respectively) in patients with NP, as well as other studies in which SC-FRI was applied to assess patients with LBP.17 The Cronbach α coefficients of SC-FRI when an item was deleted were also high (0.83–0.86). Although the last 2 items (walking and standing) had lower item-total correlations, the coefficients were more than 0.30 and thus they could not be excluded from the construct. Test-retest reliability indicates whether the instrument can remain stable over time. ICC was calculated to explore testretest reliability of SC-FRI. The ICC value of SC-FRI (0.97) in the present study was similar to that in the Persian study (0.96) and greater than that in the Korean study (0.86).13 In studies assessing LBP, FRI has excellent internal consistency and test-retest reliability,17 suggesting that FRI is reliable across patients with NP and/or LBP. Various instruments have been used to evaluate the construct validity of different language versions of FRI, including the Oswestry Disability Index, SF-36, SF-12, and VAS in patients with LBP, and NDI, NPDS, global perceived effect, www.spinejournal.com
E541
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E541
3/30/15 7:54 PM
HEALTH SERVICES RESEARCH
Validation FRI in Chinese • Wei et al
TABLE 2. The Response Trend, Item-Total Score Pearson Correlations, and Cronbach α Coefficients
(When 1 Item Is Deleted From SC-FRI) Z-Skewness
Item-Total Score Correlation (r)
Cronbach α Coefficient (r)
−0.38
0.64
0.86
Sleeping
0.50
0.73
0.84
Personal care
0.11
0.77
0.84
Travel
0.63
0.81
0.83
Work
0.81
0.66
0.85
Recreation
0.70
0.79
0.84
Frequency of pain
0.46
0.72
0.85
−0.21
0.58
0.86
Walking
0.16
0.48
0.86
Standing
0.25
0.48
0.86
Item in SC-FRI Pain intensity
Lifting
SC-FRI indicates the simplified Chinese version of Functional Rating Index.
SF-36, and VAS in patients with NP. Here, the construct validity of SC-FRI in patients with nonspecific NP was evaluated by NDI and NPDS, in addition to SF-36 and VAS. SCFRI correlated strongly with the PF and BP domains of the SF-36, indicating that SC-FRI has good convergent validity. Meanwhile, there were low Pearson correlation coefficients for SC-FRI and the VT, RE, and MH domains of SF-36, demonstrating discriminant validity of SC-FRI. Our findings were in accordance with the original English study, which found that FRI had a high correlation with the SF-12 PF and a low correlation with the SF-12 MH.5 In the present study, there was a strong correlation between SC-FRI and
Figure 1. Bland-Altman plot for test-retest agreement of the Simplified Chinese version of Functional Rating Index in patients with neck pain (n = 60). There are no differences in the mean total scores measured at the 2 test sessions. The dashed line indicates the 95% (±1.96 standard deviation) limits of agreement.
E542
NDI and NPDS, which are comparable with the results of the Persian study.14 In addition, the Pearson correlation coefficients for SC-FRI and the BP and PF domains of SF-36 and VAS were greater than the correlation coefficients for NDI and NPDS and the BP and PF domains of SF-36 and VAS (Table 3). These findings indicate that SC-FRI has good construct validity, as evaluated by NDI, NPDS, SF-36, and VAS. With the globalization of clinical research, self-reported outcomes assessed by disease-specific instruments are attracting increasing attention from orthopedic surgeons in mainland China. FRI is a valid, reliable, and internationally used scale that is designed to subjectively assess patients with NP and/or LBP and dysfunction.9–14 As demonstrated in our previous study in patients with LBP, the present study further provides strong evidence that SC-FRI is similarly reliable and valid for evaluating pain and functional status in Chinese patients with nonspecific NP. There are a few limitations in present study. First, although patients enrolled in this study came from different areas of mainland China, the study cohort may not fully represent the Chinese population, and thus there may be a patient selection bias due to the limited sample size. Second, SC-FRI was intended primarily for the use in mainland China and may not be valid for Chinese speakers in Hong Kong, Taiwan, and other Chinese-speaking locations. Therefore, SC-FRI needs to be further validated in different areas, especially where simplified Chinese is the standard written form. Moreover, as there is no report on validation of FRI in patients with NP in geographical regions of Europe and North America, the value of this construct should be further determined in these regions. Third, the responsiveness of SC-FRI was not assessed in patients with nonspecific NP, as well as in patients with back pain. Thus, it is important to validate SC-FRI in a larger population in mainland China in order to probe the responsiveness of this instrument in patients with NP or LBP.
www.spinejournal.com
May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E542
3/30/15 7:54 PM
HEALTH SERVICES RESEARCH
Validation FRI in Chinese • Wei et al
TABLE 3. Construct Validity Evaluation of SC-FRI Against SF-36, NDI, NPDS, and VAS by Pearson
Correlations Analysis
SF-36 Subscales
SC-FRI
SC-NDI
SC-NPDS
PF
−0.69 (P < 0.0001)
−0.58 (P < 0.0001)
−0.58 (P < 0.0001)
RP
−0.47 (P < 0.0001)
−0.40 (P < 0.0001)
−0.40 (P < 0.0001)
BP
−0.66 (P < 0.0001)
−0.48 (P < 0.0001)
−0.66 (P < 0.0001)
GH
−0.51 (P < 0.0001)
−0.45 (P < 0.0001)
−0.39 (P < 0.0001)
VT
−0.31 (P < 0.0001)
−0.40 (P < 0.0001)
−0.34 (P < 0.0001)
SF
−0.60 (P < 0.0001)
−0.50 (P < 0.0001)
−0.47 (P < 0.0001)
RE
−0.18 (P = 0.050)
−0.079 (P = 0.386)
−0.075 (P = 0.414)
MH
−0.26 (P = 0.004)
−0.23 (P = 0.011)
−0.22 (P = 0.015)
SC-NDI
0.77 (P < 0.0001)
NA
0.67 (P < 0.0001)
SC-NPDS
0.78 (P < 0.0001)
0.68 (P < 0.0001)
NA
VAS
0.86 (P < 0.0001)
0.70 (P < 0.0001)
0.74 (P < 0.0001)
Numbers in the table are Pearson correlation coefficients (r), with P value within parentheses. Pearson correlation values of r = 0–0.25, 0.25–0.50, 0.50–0.75, and > 0.75 indicate poor, fair, moderate to good, and very good to excellent correlation, respectively. SF-36 indicates 36-Item Short Form Health Survey; SC-FRI, the simplified Chinese version of Functional Rating Index; SC-NDI, simplified Chinese version of neck disability index; SC-NPDS, simplified Chinese version of neck and pain disability scale; PF, physical functioning; RP, role-physical; BP, bodily pain; GH, general health; VT, vitality; SF, social functioning; RE, role-emotional; MH, mental health; VAS, visual analogue scale; NA, not applicable.
TABLE 4. Summary of Evidence on Reliability and Validity of Various Versions of the Functional
Rating Index for Neck Pain
Internal Consistency Versions Lee et al
13
Ansari et al
14
Present study
Test-Retest Reliability
Construct Validity
N
Cronbach α Coefficient
N
ICC
Tested Construct
Pearson Correlation (r)
261
0.88
40
0.86
GPE
−0.46
100
0.88
50
0.96
NDI
0.72
NPDS
0.63
VAS
0.75
NDI
0.77
NPDS
0.78
VAS
0.86
SF-36 PF
−0.69
SF-36 RP
−0.47
SF-36 BP
−0.66
SF-36 GH
−0.51
SF-36 VT
−0.31
SF-36 SF
−0.60
SF-36 RE
−0.18
SF-36 MH
−0.26
122
0.86
60
0.97
ICC indicates intraclass correlation coefficient; GPE, global perceived effect; NDI, neck disability index; NPDS, neck and pain disability scale; VAS, visual analogue scale; PF, physical functioning; RP, role limitations due to physical health; BP, bodily pain; GH, general health perception; VT, vitality; SF, social functioning; RE, role limitations due to emotional problems; MH, mental health.
Spine
www.spinejournal.com
E543
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E543
3/30/15 7:54 PM
HEALTH SERVICES RESEARCH
Validation FRI in Chinese • Wei et al
CONCLUSION The cross-culturally adapted SC-FRI showed good clinical utility, internal consistency, test-retest reliability, and construct validity when administrated to Chinese patients with nonspecific NP. This work will promote application of SC-FRI to evaluate health status in patients with NP, especially those with both NP and LBP in mainland China.
➢ Key Points Functional Rating Index (FRI) is culturally adapted to simplified Chinese for measurement of nonspecific neck pain in mainland China. The simplified Chinese FRI is easy to understand and quick to complete, so it is practical for wide use in clinical practice. The simplified Chinese FRI shows satisfactory internal consistency, test-retest reliability, and construct validity in simplified Chinese-speaking patients with nonspecific neck pain in mainland China.
Acknowledgments Xianzhao Wei, MD, Ximing Xu, MD, Yongfei Zhao, MD, and Kai Chen, MD, contributed equally to this study.
References
1. Tsakitzidis G, Remmen R, Peremans L, et al. Non-specific neck pain: diagnosis and treatment. KCE Reports 119C Good Clinical Practice (GCP). Brussels, Belgium: Belgian Health Care Knowledge Centre (KCE); 2009. 2. Yue P, Liu F, Li L. Neck/shoulder pain and low back pain among school teachers in China, prevalence and risk factors. BMC Public Health 2012;12:789. 3. Vernon H, Mior S. The Neck Disability Index: a study of reliability and validity. J Manipulative Physiol Ther 1991;14:409–15. 4. Wheeler AH, Goolkasian P, Baird AC, et al. Development of the Neck Pain and Disability Scale. Item analysis, face, and criterionrelated validity. Spine (Phila Pa 1976) 1999;24:1290–4. 5. Feise RJ, Michael Menke J. Functional rating index: a new valid and reliable instrument to measure the magnitude of clinical change in spinal conditions. Spine (Phila Pa 1976) 2001;26:78–86; discussion 7. 6. Ailliet L, Knol DL, Rubinstein SM, et al. Definition of the construct to be measured is a prerequisite for the assessment of validity. The Neck Disability Index as an example. J Clin Epidemiol 2013;66:775–82; quiz 782.e1–2. 7. Bullinger M, Alonso J, Apolone G, et al. Translating health status questionnaires and evaluating their quality: the IQOLA Project approach. International Quality of Life Assessment. J Clin Epidemiol 1998;51:913–23. 8. Uysal-Bozkir Ö, Parlevliet JL, de Rooij SE. Insufficient crosscultural adaptations and psychometric properties for many translated health assessment scales: a systematic review. J Clin Epidemiol 2013;66:608–18. 9. Ceran F, Ozcan A. The relationship of the Functional Rating Index with disability, pain, and quality of life in patients with low back pain. Med Sci Monit 2006;12:CR435–9. 10. Costa LO, Maher CG, Latimer J, et al. Clinimetric testing of three self-report outcome measures for low back pain
E544
11.
12. 13. 14. 15. 16. 17. 18. 19. 20.
21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35.
patients in Brazil: which one is the best? Spine (Phila Pa 1976) 2008;33:2459–63. Costa LO, Maher CG, Latimer J, et al. Psychometric characteristics of the Brazilian-Portuguese versions of the Functional Rating Index and the Roland Morris Disability Questionnaire. Spine (Phila Pa 1976) 2007;32:1902–7. Ansari NN, Feise RJ, Naghdi S, et al. The functional rating index: reliability and validity of the Persian language version in patients with low back pain. Spine (Phila Pa 1976) 2011;36:E1573–7. Lee H, Nicholson LL, Adams RD, et al. Development and psychometric testing of Korean language versions of 4 neck pain and disability questionnaires. Spine (Phila Pa 1976) 2006;31:1841–5. Ansari NN, Feise RJ, Naghdi S, et al. The functional rating index: reliability and validity of the Persian language version in patients with neck pain. Spine (Phila Pa 1976) 2012;37:E844–8. Rebbeck T, Sindhusake D, Cameron ID, et al. A prospective cohort study of health outcomes following whiplash associated disorders in an Australian population. Inj Prev 2006;12:93–8. Rebbeck TJ, Refshauge KM, Maher CG, et al. Evaluation of the core outcome measure in whiplash. Spine (Phila Pa 1976) 2007;32:696–702. Feise RJ, Menke JM. Functional Rating Index: literature review. Med Sci Monit 2010;16:RA25–36. Wei X, Chen Z, Bai Y, et al. Validation of the simplified Chinese version of the functional rating index for patients with low back pain. Spine (Phila Pa 1976) 2012;37:1602–8. Guillemin F, Bombardier C, Beaton D. Cross-cultural adaptation of health-related quality of life measures: literature review and proposed guidelines. J Clin Epidemiol 1993;46:1417–32. Wild D, Grove A, Martin M, et al. Principles of good practice for the translation and cultural adaptation process for patient-reported outcomes (PRO) measures: report of the ISPOR task force for translation and cultural adaptation. Value Health 2005;8:94–104. Wei X, Yi H, Wu B, et al. A valid cross-culturally adapted simplified Chinese version of the Quebec Back Pain Disability Scale. J Clin Epidemiol 2012;65:1321–8. Wu S, Ma C, Mai M, et al. Translation and validation study of Chinese versions of the neck disability index and the neck pain and disability scale. Spine (Phila Pa 1976) 2010;35:1575–9. Chen Z, Zhao Y, Wang C, et al. An adapted Chinese version of Neck Pain and Disability Scale: validity and reliability. Spine (Phila Pa 1976) 2011;36:1322–7. Hays RD, Sherbourne CD, Mazel RM. The RAND 36-item health survey 1.0. Health Econ 1993;2:217–27. Li L, Wang HM, Shen Y. Chinese SF-36 Health Survey: translation, cultural adaptation, validation, and normalisation. J Epidemiol Community Health 2003;57:259–63. Kane RL, Bershadsky B, Rockwood T, et al. Visual Analog Scale pain reporting was standardized. J Clin Epidemiol 2005;58:618–23. Terwee CB, Bot SD, de Boer MR, et al. Quality criteria were proposed for measurement properties of health status questionnaires. J Clin Epidemiol 2007;60:34–42. Frich LH, Noergaard PM, Brorson S. Validation of the Danish version of Oxford Shoulder Score. Dan Med Bull 2011;58:A4335. Field A. Discovering Statistics Using SPSS. London: Sage Publications; 2009. Barten JA, Pisters MF, Huisman PA, et al. Measurement properties of patient-specific instruments measuring physical function. J Clin Epidemiol 2012;65:590–601. Fayers PM, Machin D. Quality of Life: Assessment, Analysis, and Interpretation. Chichester, England: John Wiley; 2000. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420–8. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10. Streiner DL, Norman GR. Health measurement scales. A Practical Guide to Their Development and Use. New York, NY: Oxford University Press; 2003. Dawson B, Trapp RG. Basic and Clinical Biostatistics. 3rd ed. New York, NY: Lange Medical Books; 2001.
www.spinejournal.com
May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141038.indd E544
3/30/15 7:54 PM
Copyright of Spine is the property of Lippincott Williams & Wilkins and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
