Self-efficacy refers to a person’s belief in his or her ability to perform a certain behavior in order to achieve a specific outcome (1). The concept of self-efficacy explains many of the behaviors and disabilities observed in patients with chronic pain (2,3). Lack of belief in self-efficacy predicts pain and physical disability in patients with fibromyalgia (4), chronic back pain (5), and mixed chronic pain conditions (6).
Patients with chronic low back pain who reportedhigher self-efficacy had higher activity levels, lowerpsychological distress, lower pain severity,less pain behaviors, and worked longer hours (7). The lack of self-efficacyin managing pain is a significant predictor of the severity of disability and depression (8). Gains in self-efficacy that occurred during treatment were associated with improved functioning and lower self-reported pain at 6 mo (9). In general, increased self-efficacy appears to enhance and maintain the long-term effects of rehabilitation (10,11). Self-efficacy has a significant influence on the utilization rate of pain-coping strategies in chronic pain patients (2), and is an important factor in the control of pain and disability (12), adaptive psychological functioning (13), and treatment outcome (14).
While the relevance of assessing self-efficacy in patients with chronic pain has been widely accepted (11), there is no validated Chinese version of a measure of self-efficacy for use with chronic pain patients of Chinese background. The Pain Self-Efficacy Questionnaire (PSEQ) (Appendix 1; available at www.anesthesia-analgesia.org) has been translated into Chinese, but its validation for global use in a chronic pain population remains to be demonstrated. The psychometric properties of the PSEQ have been reported in Australian pain clinic studies (6,15). In Australia and the United Kingdom the PSEQ correlates with activities such as return to work despite pain (10,16), medication use in chronic pain patients (17), treatment-related changes in function (18,19), and is predictive of future pain behavior (6).
The present study aimed to validate a Chinese version of PSEQ (PSEQ-HK) in a heterogeneous population of Chinese patients with chronic nonmalignant pain.
After local Clinical Research Ethics Committee approval and written informed consent, 120 consecutive patients attending the pain management center at the Alice Ho Miu Lin Nethersole Hospital, with a history of chronic nonmalignant pain longer than 6 mo, were invited to participate in the study. Patients who were unable to comprehend or self-complete the questionnaire were excluded.
The PSEQ is a 10-item self-report inventory that assesses the strength and generality of a patient’s self-efficacy beliefs and his or her confidence to accomplish a range of activities despite chronic pain. Each item is scored on a 7-point Likert scale (ranging from 0 = “not at all confident” to 6 = “completely confident”), with a higher total score indicating stronger self-efficacy beliefs. The maximum possible score is 60.
A Chinese version of PSEQ, previously translated by a local clinical psychologist, was used with the consent of the author. The translated Chinese version of the PSEQ was independently reviewed and back-translated into English by an independent native Chinese, who was blinded to the original English version of the PSEQ. This translator is proficient in both English and Chinese, and has a Master’s degree in Linguistic Studies. The participation of a linguistic expert in the verification process ensured that the translated tool would be easily comprehensible by patients. The back-translated version was then compared with the original English version and any inconsistency in the Chinese version was corrected.
Content validity of each item was evaluated by subjecting the two versions to critical appraisal by four bilingual clinical psychologists for semantic equivalence, clarity, and grammatical accuracy. The PSEQ-HK would be revised should there be any disagreement and the content validity would then be re-examined.
For reliability and repeatability testing, 55 patients completed PSEQ-HK twice, first after outpatient clinic visits and then again 1 wk later. Patients were instructed to circle the appropriate answers without any assistance from others. To minimize the learning or memory effect, the order of the 10 items in PSEQ-HK was reshuffled for the repeat test.
Construct validity was determined by comparing the PSEQ-HK scores to the validated local Chinese version of various psychometric instruments to see whether similar relationships could be obtained as with the English version. All patients recruited into the study completed these questionnaires at the same time as the PSEQ-HK. As there was no alternative Chinese tool that measures self-efficacy, criterion validity in this respect was not performed.
The Roland–Morris Disability Questionnaire (RMDQ) is a 24-item self-completed questionnaire designed to assess the degree of functional limitation in patients with low back pain (20). The Chinese version (RMDQ-HK) was validated locally by the Working Group on Low Back Pain Outcomes Assessment of Hospital Authority in 1999, showing good test–retest reliability (correlation coefficient (ICC) 0.94), high internal consistency (Kuder–Richardson 20 coefficient 0.86) as well as satisfactory construct validity. In our study, the RMDQ-HK was modified such that the term “back” is replaced by “pain.” A similar modified version (6,21), previously used in a heterogeneous group of chronic pain patients, confirmed the excellent internal reliability, validity, and sensitivity of the modified questionnaire.
The Numeric Pain Rating Scale (NPRS) measures pain intensity (22). It consists of 11 points from 0 to 10, with 0 being “no pain” and 10 being “worst pain imaginable.” Its validity and sensitivity to treatment have been demonstrated (23).
The Hospital Anxiety and Depression Score is a simple 14-item scale originally developed by Zigmond and Snaith (24) to screen for anxiety and depression. Seven items are allocated to each measure of anxiety and depression. Each item is scored on a 4-point scale from 0 to 3. The Chinese version has good internal consistency, external validity as well as favorable sensitivity and specificity for screening purposes (25).
The Medical Outcome Study Short Form 36 (SF36) is a 36-item instrument that was designed to measure general health status and health-related quality of life (26). One item assesses perceived change in health status, while 35 items examine eight generic domains in both physical and mental health. Scores in each subscale range from 0 to 100, with higher scores indicating better health status. The Chinese version was shown to be reliable with good construct validity (27).
The Pain Catastrophizing Scale is a self-reported, 13-item questionnaire used to assess the catastrophizing cognitions of individuals by asking them to reflect on thoughts or feelings associated with past painful experiences (28). Each item is scored on a 5-point scale ranging from 0 (not at all) to 4 (always) and can be subdivided into three subscales: helplessness, magnification, and rumination. It has good temporal stability, internal consistency (coefficient α 0.87), and validity. A similar result was demonstrated with the Chinese version in a local group of chronic pain patients (29).
Self-efficacy was shown to influence chronic pain patients’ use of coping strategies (2), which in turn, may influence their compliance with treatment or use of medication. A patient’s history regarding the number and type of medications taken for the week before completing the PSEQ-HK was recorded. The medications were classified into 10 categorical groups: antidepressants, anticonvulsants, nonsteroidal antiinflammatory drugs, non-narcotic analgesics, narcotic analgesics, compound analgesics, steroids, muscle relaxants, sedatives-hypnotics, and Chinese herbal medicines.
All statistical analyses were conducted using the Statistical Package for Social Sciences version 11 for Windows.
Demographic data were analyzed using descriptive statistics. The associations between PSEQ-HK and demographic characteristics were evaluated using Pearson product–moment correlations.
Cronbach’s α was used to evaluate the internal consistency of the questionnaire. It measures how well the items in a scale with unidimensional construct are inter-related and ranges in value from 0 to 1, with higher value implying greater consistency. It is a commonly used reliability coefficient and can be used for both dichotomous and Likert scale questionnaire. In the presence of a redundant item (i.e., items asking the same question in a slightly different way), the α coefficient will be higher when that redundant item was deleted. Thus this test will also help to identify any redundant item, which can subsequently be excluded from the scale to improve its internal consistency.
Test–retest reliability was assessed using analysis of variance ICC (3,1) (30), which was shown by the following formula, whereby BMS is the between-person mean square, k is the number of rater, and EMS is the error mean square. In our analysis, ICC (3,1) was interpreted in terms of consistency as our questionnaire was a 7-point Likert scale.
A principal components method with varimax rotation was used to analyze the factorial structure ofthe PSEQ-HK. Pearson product–moment correlations were used to assess the relationship between PSEQ and other measures of psychometric properties. Because of multiple correlations used in our study, only correlations with r > 0.4 and P < 0.001 were considered as significant in our analysis.
Sample Size for Exploratory Factor Analysis
Nicholas, in his paper presented at the annual conference of the British Psychological Society in 1989, showed that PSEQ conformed to a simple one-factor model in an Australian population with chronic nonmalignant pain. Thus a minimum sample size of 100 would be needed (31,32).
After critical appraisal, we adopted the Chinese PSEQ-HK questionnaire without any further modification (Appendix 2; available at www.anesthesia-analgesia.org).
One-hundred-twenty consecutive, chronic, nonmalignant pain patients were recruited during the period between July 2003 and March 2004. The mean age of our patients was 40 (sd, 9.6) yr, while the median duration of their pain was 34 (range, 6–454) mo. Patients’ demographic characteristics are summarized in Table 1. There was no association between PSEQ-HK with demographic characteristics, especially age, duration of chronic pain, and education level.
The median and mean PSEQ-HK scores were both 28.5 (range 0–58; sd 13.3). The scores for all instruments are summarized in Table 2.
The overall ICC value was 0.75. Individual ICC values (Table 3) ranged from 0.51 (item 7) to 0.71 (items 3 and 5). Cronbach’s coefficient α of PSEQ-HK was 0.93. The highest value for Cronbach’s α if an item was deleted was 0.93, indicating there was no redundant item in PSEQ-HK. In addition, all item-total correlations were relatively high, ranging from 0.622 (item 7) to 0.807 (item 6), with P values < 0.001. The minimal factor loading was 0.69 (item 7).
Factor analysis using principal components obtained a one-factor solution using eigen-value-greater-than-one criterion (Kaiser criterion), which accounted for 61% of the total variance (Table 4). Since only one factor is extracted, the solution cannot be rotated.
The Pearson product–moment correlation was used to assess the relationship between PSEQ-HK and other instruments. The results are summarized in Table 5. Despite having no correlation between PSEQ-HK and NPRS, there were significant correlations between SF36-BP and PSEQ-HK (r = 0.402) as well as between SF36-BP and NPRS (r = −0.462).
This study describes the translation and validation of PSEQ-HK. Factor analysis demonstrated relatively high values for communalities (>0.6) for at least four items in the PSEQ-HK. This supported that the validation study had adequate case recruitment (31). The scores of all SF36 domains were lower than the norm of the Hong Kong population (27) but similar to those observed in a Danish chronic nonmalignant pain population (33), suggesting that our population was a mild to moderately disabled sample.
PSEQ-HK is not a measure of pain or physical disability as it has an insignificant correlation with NPRS or measures of physical disabilities (RMDQ, SF36-RP, and SF36-GH). However, PSEQ-HK was significantly correlated with SF36-BP (r = 0.402). This may be explained by the SF36-BP scale being composed of two items: one assessing pain severity and the other, the degree to which pain was interfering with normal activities that likely reflected SF36-BP’s relationship with the PSEQ-HK. This provided support for the discriminative validity of PSEQ-HK. The expected positive relationships found between PSEQ-HK and measures of psychological well-being (SF36-VT, SF36-SF, SF36-MH, and SF36-RE), and corresponding negative relationships with measures of psychological distress (Hospital Anxiety and Depression Score scales and Pain Catastrophizing Scale subscales), provide additional support for the concurrent validity of PSEQ-HK. These results are consistent with those described by Levin et al. (7), who reported that individuals with higher self-efficacy experienced lower psychological distress and were more persevering in tasks, despite repeated failures.
Our study did not show any association between PSEQ-HK and medications used. In this study, the categorical approach was used given the unreliability of a dosage report (34). To reflect the wide variety of medications available, 10 categories of medication were chosen instead of a maximum of five categories used in earlier studies (18,34). As a result, our sample size may have been inadequate to produce any meaningful correlation (P = 0.295). Alternatively, this discrepant finding may reflect cultural differences between Chinese and Caucasian population, whereby the use of medications may be viewed differently. For example, Chinese patients usually expect to receive some medications or physical treatment at the end of the medical consultation. The possibility of cultural differences between different populations in approaches to taking medication is beyond the scope of this study, but clearly it merits further investigation.
There are several limitations to this study. First, despite the likely importance of cultural considerations, we did not evaluate the PSEQ-HK from a cultural perspective. Even so, the findings obtained with the PSEQ-HK are consistent with those reported in other countries and would seem to suggest that there are commonalities among chronic pain patients from different cultures. The findings reported here have also shown that the PSEQ-HK is not only acceptable to our heterogeneous Chinese patients with chronic pain in Hong Kong, but is also applicable clinically. Second, this study was based on one clinic, and it is unknown how representative the patients attending this clinic were of Hong Kong chronic pain patients in general. Nonetheless, it is worth noting that the population served by our health region is about 1.2 million, roughly 17% of the Hong Kong population. Third, the degree to which the PSEQ-HK is sensitive to treatment-related changes was not assessed. Given the similar findings between the present study and those of others reported here with the PSEQ, and other self-efficacy measures, it might be expected that the PSEQ-HK will be sensitive to treatment changes, as already reported for the PSEQ (19). However, this still remains to be demonstrated. Finally, the present study did not assess the threshold PSEQ-HK scores which may provide a degree of prediction for future outcomes such as return to work or dropouts from treatment, as has been done with the PSEQ in other countries (10,16). This is important, and may have significant clinical impact considering the current climate of limited public health resources in Hong Kong and elsewhere.
In summary, our preliminary results have demonstrated that the PSEQ-HK is a reliable clinical assessment tool with satisfactory construct validity for self-efficacy in our heterogeneous Chinese population with chronic nonmalignant pain.
We express our appreciation to Professor Michael Nicholas for his advice and permission to translate his PSEQ questionnaire; Mr. Raymond Chung, Dr. Jacqueline Yap, and Mr. Edward Choi for their assistances in statistical analysis as well as Dr. Sammy Cheng, for allowing us to use the Chinese translation of PSEQ in this validation study. We are also grateful to Ms. Wendy Fung (pain nurse) and Ms. Winnie Samy (research nurse) for their help in data collection, as well as the entire staff at the Pain Management Center who assisted us with this project.
1. Bandura A. Self-efficacy: toward a unifying theory of behavioral change. Psychol Rev 1977;84:191–215.
2. Jensen MP, Turner JA, Romano JM. Self-efficacy and outcome expectancies: relationship to chronic pain coping strategies and adjustment. Pain 1991;44:263–9.
3. Turk DC. Biopsychosocial perspective on chronic pain. In: Gatchel RJ, Turk DC, eds. Psychological approaches to pain management. New York: Guilford Press, 1996:3–22.
4. Buckelew SP, Parker JC, Keefe FJ, et al. Self-efficacy and pain behavior among subjects with fibromyalgia. Pain 1994;59:377–84.
5. Lackner JM, Carosella A, Feuerstein M. Pain expectancies, pain, and functional self-efficacy expectancies as determinants of disability in patients with chronic low back disorders. J Consult Clin Psychol 1996;64:212–20.
6. Asghari A, Nicholas MK. Pain self-efficacy beliefs and pain behaviour. A prospective study. Pain 2001;94:85–100.
7. Levin JB, Lofland KR, Cassisi JE, et al. The relationship between self-efficacy and disability in chronic low back pain patients. Int J Rehabil Health 1996;2:19–28.
8. Arnstein P, Caudill M, Mandle CL, et al. Self efficacy as a mediator of the relationship between pain intensity, disability and depression in chronic pain patients. Pain 1999;80:483–91.
9. Altmaier EM, Russell DW, Kao CF, et al. Role of self-efficacy in rehabilitation outcome among chronic low back pain patients. J Couns Psychol 1993;40:335–9.
10. Adams JH, Williams AC. What affects return to work for graduates of a pain management program with chronic upper limb pain? J Occup Rehabil 2003;13:91–106.
11. Keefe FJ, Rumble ME, Scipio CD, et al. Psychological aspects of persistent pain: current state of the science. J Pain 2004;5:195–211.
12. Lorig K, Chastain RL, Ung E, et al. Development and evaluation of a scale to measure perceived self-efficacy in people with arthritis. Arthritis Rheum 1989;32:37–44.
13. Spinhoven P, Ter Kuile MM, Linssen AC, Gazenam B. Pain coping strategies in a Dutch population of chronic low back pain patients. Pain 1989;37:77–83.
14. O’Leary A, Shoor S, Lorig K, Holman HR. A cognitive-behavioral treatment for rheumatoid arthritis. Health Psychol 1988;7:527–44.
15. Gibson L, Strong J. The reliability and validity of a measure of perceived functional capacity for work in chronic back pain. J Occup Rehabil 1996;6:159–75.
16. Cohen M, Nicholas MK, Blanch A. Medical assessment and management of work-related low back or neck-arm pain: more questions than answers. J Occup Health Safety Aust NZ 2000;16:307–17.
17. Ralphs JA, Williams AC, Richardson PH, et al. Opiate reduction in chronic pain patients: a comparison of patient-controlled reduction and staff controlled cocktail methods. Pain 1994;56:279–88.
18. Nicholas MK, Wilson PH, Goyen J. Comparison of cognitive behavioral group treatment and an alternative non-psychological treatment for chronic low back pain. Pain 1992;48:339–47.
19. Williams AC, Richardson PH, Nicholas MK, et al. Inpatient vs. outpatient pain management: results of a randomised controlled trial. Pain 1996;66:13–22.
20. Roland M, Morris R. A study of the natural history of back pain. I. Development of a reliable and sensitive measures of disability in low-back pain. Spine 1983;8:141–4.
21. Jensen MP, Strom SE, Turner JA, Romano JM. Validity of the sickness impact profile Roland scale as a measure of dysfunction in chronic pain patients. Pain 1992;50:157–62.
22. Jensen MP, Karoly P, Braver S. The measurement of clinical pain intensity: a comparison of six methods. Pain 1986;27:117–26.
23. Jensen MP, Karoly P. Self-report scales and procedure for assessing pain in adults. In: Turk DC, Melzack R, eds. Handbook for pain assessment. New York: Guilford Press, 1992:193–213.
24. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983;67:361–70.
25. Leung CM, Ho S, Kan CS, et al. Evaluation of the Chinese version of the hospital anxiety and depression scale. A cross-cultural perspective. Int J Psychosom 1993;40:29–34.
26. Ware JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992;30:473–83.
27. Lam CL, Gandek B, Ren XS, Chan MS. Tests of scaling assumptions and construct validity of the Chinese (HK) version of the SF-36 Health Survey. J Clin Epidemiol 1998;51:1139–47.
28. Sullivan MJL, Bisop SC, Pivik J. The pain catastrophizing scale: development and validation. Psychol Assess 1995;7:524–32.
29. Yap JC, Chen PP, Gin T, et al. Validation of the Chinese pain catastrophizing scale (HK-PCS) in chronic pain patients. Qual Life Res 2004;13:1529.
30. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420–8.
31. MacCallum RC, Widaman KF, Zhang S, Hong S. Sample size in factor analysis. Psychol Methods 1999;4:84–99.
32. Bearden WD, Sharma S, Teel JE. Sample size effects on chi-square and other statistics used in evaluating causal models. J Mark Res 1982;19:425–30.
33. Becker N, Bondegaard Thomsen A, et al. Pain epidemiology and health related quality of life in chronic non-malignant pain patients referred to a Danish multidisciplinary pain centre. Pain 1997;73:393–400.
34. Turner JA, Calsyn DA, Fordyce WE, Ready LB. Drug utilization patterns in chronic pain patients. Pain 1982;12:357–63.