European Journal of Pain 11 (2007) 153–163
www.EuropeanJournalPain.com
The pain self-efficacy questionnaire: Taking pain into account
Michael K. Nicholas
*
Pain Management and Research Centre, University of Sydney at Royal North Shore Hospital, St. Leonards, Sydney, NSW 2065, Australia
Received 19 July 2005; received in revised form 12 December 2005; accepted 14 December 2005
Available online 30 January 2006
Abstract
Self-efficacy beliefs in people with chronic pain have been assessed either by reference to confidence in ability to perform specific
tasks or to confidence in performing more generalised constructs like coping with pain. Both approaches reflect aspects of the original conceptualisations of self-efficacy and both have proved useful, but it is noteworthy that confidence in performing activities in
the context of pain is rarely addressed. An important element in the original formulations of self-efficacy referred to persistence in
the face of obstacles and aversive experiences. In this context, self-efficacy beliefs for people experiencing chronic pain might be
expected to incorporate not just the expectation that a person could perform a particular behaviour or task, but also their confidence
in being able to do it despite their pain. This aspect of the self-efficacy construct has been included in a measure for people with
chronic pain, the Pain Self-Efficacy Questionnaire (PSEQ). The accumulated evidence from a number of published studies and a
confirmatory analysis with a large cohort of heterogeneous chronic pain patients attending a pain management program provide
support for the PSEQ’s original psychometric properties developed with a sample of chronic low back pain patients. The importance
of taking the context of pain into account in the assessment of self-efficacy beliefs in pain populations and the ways in which this
measure can be used to improve the assessment of people experiencing chronic pain, before and after treatment, are examined.
Ó 2005 European Federation of Chapters of the International Association for the Study of Pain. Published by Elsevier Ltd. All
rights reserved.
Keywords: Chronic pain; Self-efficacy; Disability; Coping; Pain management
1. Introduction
Bandura (1977) proposed that ‘‘efficacy expectations
determine how much effort people will expend and
how long they will persist in the face of obstacles and
aversive experiences’’ (p. 194). In the study of pain, efficacy expectations (or ‘self-efficacy’ beliefs) have been
used to explain a range of behaviours and aspects of
pain experience (e.g., Jensen et al., 1991; Lackner and
Carosella, 1993; Anderson et al., 1995; Arnstein et al.,
1999; Pincus and Morley, 2002; Turk, 2002a; Altmaier
et al., 1993; Asghari and Nicholas, 2001; Rudy et al.,
2003; Keefe et al., 2004). In experimentally induced
*
Tel.: +612 9926 7318; fax: +612 9926 6548.
E-mail address: miken@med.usyd.edu.au.
pain, stronger confidence in ability to tolerate pain has
been found to predict actual tolerance, regardless of
whether cognitive or pharmacological agents were
employed to control pain (Bandura et al., 1987). In
patients with chronic osteoarthritic knee pain, those
with high self-efficacy for controlling arthritis pain have
been found to have higher pain thresholds and tolerance
for experimentally induced thermal pain than those
whose self-efficacy was low (Keefe et al., 1997). Similarly, in patients with chronic pain, confidence in ability
to perform specified activities has been correlated with
the subsequent performance of those activities (Council
et al., 1988).
Bandura (1989) went beyond the idea that self-efficacy beliefs related only to specific behaviours and
argued that people could hold efficacy beliefs about their
1090-3801/$32 Ó 2005 European Federation of Chapters of the International Association for the Study of Pain. Published by Elsevier Ltd. All rights
reserved.
doi:10.1016/j.ejpain.2005.12.008
154
M.K. Nicholas / European Journal of Pain 11 (2007) 153–163
ability to cope in the face of adversity. In this context he
referred to strong self-efficacy beliefs as a ‘‘resilient selfbelief system’’ whereby ‘‘people who believe they can
exercise control over potential threats do not conjure
up apprehensive cognitions and, hence, are not perturbed by them’’ (p. 419). In support of this argument
Bandura cited a path analysis study that demonstrated
links between coping efficacy beliefs, avoidance behaviours and anxiety (Ozer and Bandura, 1990). The idea
that self-efficacy beliefs may be related not only to specific behaviours, but also to broader constructs, like coping, has been reflected in many self-efficacy scales (e.g.,
Lackner and Carosella, 1993; Anderson et al., 1995;
Edwards et al., 2000). However, the relationship
between self-efficacy beliefs and coping may require further study in the light of the ongoing debate over the
construct of ‘coping’ in the context of pain (e.g., Eccleston, 1995; Turner et al., 2000). There is a good case for
distinguishing between the effect (coping) with the
actions involved in coping. To date, this distinction
has received relatively little attention in studies of coping
self-efficacy and pain.
While generally supportive of the thesis that self-efficacy beliefs held by people in chronic pain are related to
their behaviour, the published findings on self-efficacy
beliefs also reveal considerable variability in both the
measures used to assess self-efficacy beliefs and in the
relationships with behaviours and other aspects of pain
experience. For example, some follow Bandura’s own
methods and use single-item scales that focus on particular behaviours or tasks (e.g., Dolce et al., 1986; Rudy
et al., 2003), while others have used more general measures with multiple items whose scores are summed to
measure more generalised constructs like ‘ability to cope
with pain’ (e.g., Lackner and Carosella, 1993; Anderson
et al., 1995; Edwards et al., 2000; McCracken and Eccleston, 2003). Some scales have included a number of subscales or subcategories of self-efficacy. For example,
Lackner et al. (1996) described a modification of Lorig
et al.’s (1989) arthritis self-efficacy scale that has three
self-efficacy subscales – confidence in performing a certain task, the ability to tolerate pain and to control pain.
In addition, some researchers have reported that selfefficacy beliefs are predictive of physical function (e.g.,
Lackner et al., 1996), while others have found that
improved physical functioning was not associated with
increased self-efficacy, at least initially (Altmaier et al.,
1993). Similarly, Dolce et al. (1986) found self-efficacy
beliefs were enhanced in some chronic pain patients
who increased their level of exercise, but not in all. Gibson and Strong (1996) reported that one measure of selfefficacy was a better predictor of assessed work capacity
than another. It is unclear if the variable findings are due
to the measures used to assess self-efficacy beliefs or to
problems with the actual construct of self-efficacy in
relation to pain.
Somewhat surprisingly, most published measures of
self-efficacy beliefs used in pain samples do not explicitly
ask the patient to take their pain into account when
describing their confidence in performing specific tasks
(e.g., Lorig et al., 1989; Jensen et al., 1991; Lackner
and Carosella, 1993; Anderson et al., 1995; Rudy et al.,
2003). It is arguable that it is one thing to feel able to perform some task and another thing to feel able to do it
while in pain. The significance of this point is that lasting
relief from chronic pain is rarely achieved by current
treatments (Turk, 2002b). Accordingly, most people with
chronic pain are confronted with the challenge of leading
as normal a life as possible despite their pain. By overlooking the context (i.e., the presence of pain) in which
a person is being asked to estimate their confidence in
performing an activity, it would also seem to make the
task somewhat artificial and create difficulties in the
interpretation of their responses. An analogous point
has been described in relation to patients with generalised social phobia where the context or social framework
of information on outcomes was found to influence
patients’ anxiety predictions (Alden et al., 2004).
One attempt to assess self-efficacy beliefs in the context of pain was described by Altmaier et al. (1993).
But in contrast to the common finding from other selfefficacy scales, Altmaier et al. did not find a relationship
between strengthened self-efficacy beliefs and increased
physical functioning. Whether this was due to the nature
of the scale itself or some aspect of taking pain into
account was not clear. Examination of the scale’s items
does reveal possible problems in its general utility. For
instance, many of the items refer to specific activities that
may not be relevant to all individuals or groups (e.g.,
shoveling snow; driving the car; raking leaves; working
on a house repair; riding a bicycle). It is possible that a
measure that does not refer to such specific activities
might have broader applicability and test the importance
of including the context of pain more effectively.
Nicholas (1989) described another self-efficacy scale
(the Pain Self-Efficacy Questionnaire; PSEQ) for people
in chronic pain that also asks the respondents to take
pain into account when rating their self-efficacy beliefs.
Unlike the scale used by Altmaier et al. (1993), the activities referred to in the PSEQ are more general (e.g., paid/
unpaid work; social activities), although specific examples are given as a guide. This was intended to make
the measure applicable to a broad range of respondents.
The PSEQ has now been used in a number of different
clinical settings and in different countries (e.g., Nicholas
et al., 1992; Williams et al., 1993, 1996, 1999; Frost et al.,
1993; Estlander et al., 1994; Ralphs et al., 1994; Coughlan et al., 1995; Gibson and Strong, 1996; Watson et al.,
1997; Cohen et al., 2000; Asghari and Nicholas, 2001;
Ayer and Tyson, 2001; Strong et al., 2002; Adams and
Williams, 2003; Dehghani et al., 2004). To date, however, a full account of the PSEQ and its psychometric
M.K. Nicholas / European Journal of Pain 11 (2007) 153–163
properties has not been published. This paper is intended
to examine this body of work as it relates to the psychometric properties and uses of the PSEQ, as well as the
contribution it could make to exploring the application
of the construct of self-efficacy to understanding the
impact of persisting pain, especially in relation to other
constructs such as fear-avoidance and pain acceptance.
2. Method
2.1. Development of the scale and initial psychometric
evaluation study
2.1.1. Item development
Ten items were selected to reflect a wide variety of
classes of activities and tasks, with indicative examples,
commonly reported as problematic by patients with
chronic pain (see Appendix A for a list of items). Some
items were derived from existing scales (e.g., Pain Beliefs
Questionnaire: Gottlieb, 1984) and modified to fit the
requirements of the new scale. Other items were derived
from the author’s experience in working with chronic
pain patients. Most importantly, all items include mention of performing the activities despite their pain (e.g.,
‘‘I can do most of the household chores (e.g., tidying-up,
washing dishes), despite the pain’’; ‘‘I can gradually
increase my activity level, despite the pain’’). The items
were tested for comprehension by patients attending
the pain clinic and modified until all items were considered comprehensible. Respondents are asked to rate
how confident they are that they can do each of the 10
activities or tasks at present despite the pain they are
experiencing. Each item is rated by selecting a number
on a 7-point scale, where 0 equals ‘‘not at all confident’’
and 6 equals ‘‘completely confident’’. A total score is
calculated by summing the scores for each of the 10
items, yielding a maximum possible score of 60. Higher
scores reflect stronger self-efficacy beliefs.
Once an acceptable version was derived, more formal
assessment of its properties began when the PSEQ was
included in a battery of scales for use with a sample of
chronic low back pain patients selected for a randomised
trial of cognitive-behavioural pain management (see
Nicholas et al., 1992).
2.1.2. Original Westmead subjects
The original (Westmead) sample on which the PSEQ
was tested comprised a consecutive series of 103 patients
with chronic low back pain attending a pain management
program at Westmead Hospital in Sydney (Australia), a
large city teaching hospital. The 103 subjects consisted
of an initial series (of 18 patients) who completed only
the PSEQ, and a subsequent series of 85 patients (who
completed the PSEQ as part of a battery of measures –
see below). After multidisciplinary pain clinic assessment
155
all 103 subjects had been accepted into the program on the
basis that they met the criteria of a history of chronic (i.e.,
more than 6-months) low-back pain; were not considered
suitable for further invasive treatments; were aged
between 18 and 60 years; had no compensation claim
due for settlement within 12-months; were able to read
and speak English; and were willing to participate in a
research-based treatment program. The study was
approved by the hospital’s ethics committee and conducted from May 1986 to February 1988.
The mean age of the 103 subjects selected (50 men, 53
women) was 42.0 years (range = 18–60). The median
duration of low back pain was 36 months (interquartile
range: 24–120 months). Forty-four (42.3%) patients had
had one or more back surgeries. One hundred and one
(97%) had had some type of physiotherapy. Thirty-five
(33.7%) had had one or more nerve blocks. Most
patients reported having tried a range of medications
and 94 (90.4%) were taking one or more pain-related
medications at the pretreatment assessment. Seventeen
(16.3%) were employed and 71 (68.3%) were claiming
compensation insurance payments. Twenty-six (25%)
had had more than three years of secondary education.
At the time of initial assessment all patients were continuing to seek treatment for their pain conditions.
2.1.3. Psychometric evaluation
The reliability and validity of the PSEQ were initially
examined using data gathered at the pre-treatment
assessments for the pain management program. Reliability was assessed by examination of internal consistency (Cronbach’s a coefficient) and stability over time
(test–retest analysis with Pearson correlations and analysis of change). Validity was assessed through analysis
of the PSEQ’s factor structure (principal components
factor analysis) and by examination of the PSEQ’s relationships with validated measures of constructs that
would be expected to have different types of relationship
with self-efficacy.
2.2. Confirmatory study
Data on the PSEQ were obtained from a sample of
1306 heterogeneous chronic pain patients at pre-admission to a 3-week cognitive behavioural pain management program at another tertiary referral pain centre
(RNSH) in Sydney. The sample represents a consecutive
series of patients admitted to the program over a period
of 8 years, from November 1994 to October 2002.
Admission criteria were similar to those of the Westmead program, with the exception that at RNSH there
were no age limits and the program was part of regular
treatment rather than a primary research project. In
order to complete the measures the subjects had to be
able to read English. In addition to completing the
PSEQ, these patients also completed a battery of other
156
M.K. Nicholas / European Journal of Pain 11 (2007) 153–163
2.2.1. RNSH sample characteristics
The 1306 heterogeneous chronic pain patients had a
mean age of 41 years (range: 13–88); male/female ratio
of 47/53%; their work status was: full-time or part-time
(20%), unemployed due to pain (62%); their median
duration of pain was 36-months (lower and upper quartiles: 23–72 months); their mean pain severity score (on
the 0–6 pain scale of the Multidimensional Pain Inventory: Kerns et al., 1985) was 4.4 (SD: 1.8; median: 4.3;
upper and lower quartiles: 5.0 and 3.7); their main pain
sites were: more than two main sites: 32%; low-back:
43%. Medication for pain was being used by 86.5%,
and the mean score on the modified Roland and Morris
(1983) Disability Questionnaire (Asghari and Nicholas,
2001) was 13.2 (SD: 5.2).
antidepressants, and sedatives/hypnotics, yielding a
scoring range of 0–5. Tendency to focus on somatic
symptoms was assessed by the Modified Somatic Perception Questionnaire (MSPQ) (Main, 1983). Range is 0–39,
with a higher score indicating greater somatic focussing.
Impact of pain on daily life was assessed by the Sickness Impact Profile-Self (SIP-S) (Bergner et al., 1981). In
this study, the SIP was interviewer administered and
rather than ‘health status’, the subjects were asked to
relate each item to the impact of their back pain (e.g.,
Turner, 1982). Only the total score was used and this
was expressed as a percentage, with a higher score indicated greater disability due to pain. A spouse/significant
other version of the SIP (SIP-O) was also used to provide their perspective on the impact of pain on the subject’s daily life (e.g., Turner, 1982). In the confirmatory
study, instead of the SIP, impact of pain on daily life
was assessed by the modified Roland and Morris Disability Questionnaire (see Asghari and Nicholas, 2001),
in which the word ‘pain’ was substituted for ‘back’.
2.3. Other measures (also administered at initial
assessment)
3. Results
self-report measures as part of their pre-admission
assessment. The Northern Sydney Area Health’s
Human Research Ethics Committee gave approval for
the use of these (de-identified) data for research.
These instruments were intended to capture the typical multidimensional features of chronic pain and
included: Average Weekly Pain – derived from the Pain
Rating Chart (Budzynski et al., 1973) on a 0–5 scale.
Pain qualities were measured by the McGill Pain Questionnaire (MPQ) (Melzack, 1975), but only the dimensions of Sensation (range: 0–42), Affect (0–14) and
Evaluation (0–5) were scored. Severity of depressive
symptoms was measured by the Beck Depression Inventory (BDI) (Beck et al., 1961). Severity of current anxiety
was assessed by the state version of the State-Trait Anxiety Scale (STAI) (Spielberger et al., 1970). Current use
of pain coping strategies was assessed by the Coping
Strategies Questionnaire (CSQ) (Rosenstiel and Keefe,
1983). This includes seven subscales (6-items each) and
two further items for rating control over pain and ability
to decrease pain. Each subscale is scored out of 36.
Beliefs about pain that could promote disability and psychological distress were assessed by the Pain Beliefs
Questionnaire (PBQ) (Gottlieb, 1984). Examples of
unhelpful statements include: ‘‘I have to lie down when
I am in pain’’ and ‘‘I think I am harming my body when
I have pain’’. For scoring purposes in this study seven
items that were judged by the author to reflect coping
or more adaptive beliefs were reverse-scored (e.g., ‘‘For
the most part, I live a normal life’’). Thus, a higher score
reflects more unhelpful beliefs (range was 0–129). Medication use was assessed by summing the number of classes of medication subjects reported taking for pain or
pain-related problems (e.g., sleep, depression) (Turner
et al., 1982). Classes included: narcotic analgesics, nonnarcotic analgesics, non-steroidal anti-inflammatories,
3.1. Initial psychometric evaluation with the Westmead
sample
As can be seen from the mean scores in Table 1, the
Westmead sample had moderately severe depressive and
anxiety symptoms (e.g., the mean depression score was
just above the BDI cut-off score, of 18, for depression
used by Kerns and Haythornthwaite (1988) in their
study of depression in chronic pain patients). Mean pain
levels were moderate (5.6/10), and interference in normal daily activities due to pain was quite marked (e.g.,
the means score on both versions of the SIP were substantially worse than those reported in the study on
chronic low back pain patients by Turner and Clancy
(1988)). The median score on the PSEQ was 26.5 (lower
and upper quartiles: 15 and 34).
3.1.1. Reliability
Reliability was established in a number of ways.
3.1.2. Internal consistency
The measure of internal consistency of items, Cronbach’s a coefficient, was calculated as 0.92. This value
is very high and indicates the instrument has excellent
internal consistency (Nunnally and Bernstein, 1984).
3.1.3. Test–retest reliability
This was tested on a different sample of a mixed
group (n = 145) of chronic pain patients treated at the
Royal North Shore Hospital in Sydney in 1995–96 (see
Asghari and Nicholas, 2001, for a description of the full
sample). For comparison purposes, this sample had a
M.K. Nicholas / European Journal of Pain 11 (2007) 153–163
3.2. Validity
Table 1
Westmead sample of chronic low back pain patients
Measure
Average pain (1–5 scale)
MPQ: Sensory
MPQ: Affective
MPQ: Evaluative
Modified Somatic Perception
Questionnaire (MSPQ)
State-Trait Anxiety Inventory (State)
Beck Depression Inventory (BDI)
Pain Beliefs Questionnaire (PBQ)
CSQ: Divert attention (0–36)
CSQ: Reinterpret (0–36)
CSQ: Catastrophise (0–36)
CSQ: Ignore (0–36)
CSQ: Pray/hope (0–36)
CSQ: Coping statements (0–36)
CSQ: Increase behaviour (0–36)
CSQ: Control pain (0–6)
CSQ: Decrease pain (0–6)
Sickness impact
Profile-Self (0–100.3)
Sickness impact
Profile-Other (0–100.3)
Medications used
(Total no. of types)
PSEQ (0–60)
N
Mean score
157
SD
85
70
70
70
70
2.8
18.8
5.9
3.2
12.8
0.77
6.55
3.29
1.46
7.58
70
85
70
85
85
85
85
85
85
85
85
85
85
51.1
19.2
57.5
15.6
7.0
16.9
13.8
17.8
20.5
17.4
2.6
2.0
33.3
13.55
9.71
20.69
8.94
6.82
8.39
6.90
9.52
6.81
6.58
1.22
1.19
12.27
71
26.1
13.72
85
1.9
1.02
103
25.8
12.4
Means and standard deviations of dependent variables at pretreatment
(not all subjects completed all measures, depending on stage of the
overall research program).
MPQ: McGill Pain Questionnaire; CSQ: Coping Strategies Questionnaire; PSEQ: Pain Self-Efficacy Questionnaire.
mean age of 50.3 years (SD: 12.8); mean pain duration
of 9.9 years (SD: 8.9); mean pain level (0–10 scale) of
6.12 (SD: 2.3); and mean depression severity (BDI)
was 14.9 (SD: 8.1).
The test–retest period was from initial assessment to 3months later. During this time all patients received some
form of ongoing treatment (mainly medication), but
reported no change in mean disability or pain. The test–
retest correlation (r) from baseline to 3-months was
0.73; (p < 0.001). The mean scores for the two occasions
were 26.7 (SD: 12.5) and 26.9 (SD: 12.6), respectively
(i.e., no significant change). Interestingly, similar findings
were reported by Williams et al. (1996) with a waiting-list
control group of mixed chronic pain patients (n = 31)
tested 12-weeks apart. In that study, where patients
(mean age: 51.1 years, SD: 10.7; mean pain duration:
7.2 years, SD: 6.6; mean BDI: 16.6, SD: 6.5; mean pain
severity (0–100): 67.9, SD: 22.3) continued with whatever
treatments their doctors had prescribed, the mean PSEQ
score at baseline was 26.3 (SD: 10.8) and after 12 weeks it
was 26.7 (SD: 6.2), again no significant change was found
(and no change in pain or disability either).
These findings would suggest that the PSEQ has a
high degree of reliability, both internally and across a
period of at least 3-months under conditions of no
change in either pain or disability.
Validity was assessed through analysis of the PSEQ’s
factor structure and by examination of the PSEQ’s relationships with validated measures.
3.2.1. Factorial structure
A principal components factor analysis with orthogonal rotation was conducted on the full (Westmead) sample in order to investigate the factorial structure of the
scale. A one-factor solution resulted from the imposition
of the eigen-value-greater-than-one criterion, accounting
for 58.6% of the total variance. All items had factor loadings above 0.64 (item 7) and these are reported in Table 2.
Corrected item–total correlations (which excluded the
relevant item from the total for each correlation) are
also presented in Table 2. These correlations varied from
0.67 (item 7) to 0.84 (items 9 and 10). Although item 7
(to do with coping with pain without medication) had
the lowest factor loading (and lowest item–total correlation), the loading of 0.64 was still moderately high and
the content of the item does have strong construct validity. For example, Ralphs et al. (1994) demonstrated that
this item was significantly correlated with mean morphine dose in chronic pain patients attending an inpatient program. As a result, item 7 was retained.
3.2.2. Correlational studies
Although there is no ‘gold standard’ measure of selfefficacy against which the PSEQ could be compared,
self-efficacy theory would predict a strong relationship
between the PSEQ and measures of activity. Given the
generalised nature of the PSEQ score, it would be
expected that PSEQ scores would correlate strongly
(but negatively) with the total score on the SIP. Similarly, it would be expected that PSEQ scores would be
correlated with other pain-related activities, such as
use of pain-related medication and pain-coping strategies, as measured by the CSQ. In the case of medication
usage, it would be expected that the PSEQ scores would
correlate negatively with higher medication usage (as
Table 2
Corrected item–total correlations, item means and standard deviations, and factor loadings for PSEQ (N = 103, Westmead chronic low
back pain sample)
Item no.
Item–total correlation
Mean
SD
Factor loading
1
2
3
4
5
6
7
8
9
10
0.70
0.72
0.71
0.83
0.74
0.79
0.67
0.79
0.84
0.84
2.7
3.3
2.8
3.0
3.0
2.3
1.9
2.2
2.3
2.5
1.5
1.6
1.7
1.5
1.7
1.6
1.9
1.7
1.6
1.6
0.703
0.722
0.703
0.834
0.735
0.791
0.643
0.795
0.852
0.848
158
M.K. Nicholas / European Journal of Pain 11 (2007) 153–163
their use could be considered to reflect lack of confidence in ability to function due to pain). In the case of
the strategies sampled by the CSQ, a positive correlation
would be expected with those strategies thought to
reflect active approaches (e.g., coping-self statements;
and increase behaviour) and perceived ability to control
pain, but negative correlations would be expected with
those strategies thought to reflect more passive strategies
(e.g., catastrophising and praying/hoping) (e.g., Brown
and Nicassio, 1987).
It would also be expected that the PSEQ would be
correlated strongly with the PBQ. While the PBQ is a
more broadly based instrument than the PSEQ, it incorporates elements of self-efficacy beliefs. But given the
PBQ’s emphasis on non-adaptive beliefs a negative correlation with the PSEQ would be expected. Given the
inclusion of ‘despite pain’ in each item of the PSEQ, correlations between the PSEQ and pain (average pain ratings and MPQ subscales) and other somatic perceptions
(MSPQ) would be expected to be negative.
Pearson product–moment correlations between the
PSEQ and the other assessment measures were examined (see Table 3). Due to the large number of intercorrelations strict significance criteria were imposed.
Accordingly, only correlations of r > 0.40 and
p < 0.001 were considered significant.
As expected, significant negative correlations were
obtained between the PSEQ and total number of medications used, impact of pain on daily life (SIP-Self- and
Sig-other-rated), mood (BDI, STAI), and unhelpful coping strategies and beliefs (catastrophising subscale of the
CSQ, and PBQ).
Also as expected, significant positive correlations were
obtained between the PSEQ and active coping strategies
measured (ignore pain, coping self-statements, increase
behaviour and control pain subscales of the CSQ). In
contrast, no significant correlations were found between
the PSEQ and the measures of pain and somatic focussing (average pain ratings, MPQ subscales, or MSPQ),
but all were in the negative direction, as expected.
3.3. Confirmatory study with RNSH sample (n = 1306)
The mean PSEQ score was 23 (SD: 12.7) and the
median was 22 (lower and upper quartile scores: 14
and 33). As with the Westmead sample, the scores were
normally distributed, but the mean and median scores
were slightly lower than those in the original sample.
The mean PSEQ scores from the RNSH sample are
close to those reported by Williams et al. (1993)
(d = 24.1; SD: 11.4) with another heterogeneous chronic
pain sample at pre-admisison to a 4-week inpatient pain
management program in London.
Internal reliability (Cronbach a correlation) was calculated at 0.93. This finding accords closely with that
of the Westmead sample and figures cited in two pub-
Table 3
Correlations between PSEQ and other measures at Pre-treatment
(Westmead sample, all with chronic low back pain, n = 103)
Measure
Average pain
McGill Pain Questionnaire
Sensation
Affect
Evaluative
MSPQ
STAI-S
BDI
PBQ
CSQ
Divert attention
Reinterpret
Catastrophis
Ignore
Pray/hope
Coping statements
Increase behaviour
Control pain
Decrease pain
SIP-Self
SIP-Other
Total number of medications used
Correlation with PSEQ
0.17
NS
0.31
0.36
0.30
0.36
0.49
0.59
0.74
NS
NS
NS
NS
p < 0.001
p < 0.001
p < 0.001
0.13
0.24
0.55
0.46
0.10
0.48
0.45
0.56
0.25
0.60
0.48
0.45
NS
NS
p < 0.001
p < 0.001
NS
p < 0.001
p < 0.001
p < 0.001
NS
p < 0.001
p < 0.001
p < 0.001
Due to the large number of intercorrelations, strict criteria were set for
significance: r > 0.4 and p < 0.001.
MSPQ: Modified Somatic Perception Questionnaire; STAI-S: State
version of State-Trait Anxiety Inventory; BDI: Beck Depression
Inventory; PBQ: Pain Beliefs Questionnaire; CSQ: Coping Strategies
Questionnaire; SIP-Self: Self-report version of Sickness Impact Profile;
SIP-Other: Significant-other report version of Sickness Impact Profile;
PSEQ: Pain Self-Efficacy Questionnaire.
lished studies, one with chronic low back pain clients
where the PSEQ’s Cronbach’s a was assessed as 0.94
(Gibson and Strong, 1996). The other study (Asghari
and Nicholas, 2001), with a heterogeneous group of
chronic pain patients attending a pain clinic for initial
assessment, yielded a Cronbach a of 0.92.
Consistent with the high Cronbach a value, examination of (corrected) item–total correlations (which
excluded the relevant item from the total for each correlation) revealed these were also mostly high and ranged
between 0.5 (for item 7) and 0.8. Apart from item 7,
dealing with confidence in coping without medication,
all other items, had an item–total correlation of 0.7 or
greater.
The factorial structure of the PSEQ was also confirmed with the RNSH sample of 1306 heterogeneous
chronic pain patients. Principal components analysis
yielded a single factor solution (Eigen value > 1) that
accounted for 60.9% of variance in scores.
A subset (n = 130, comprising a consecutive series of
patients who attended the program over a 2-year period)
of the RNSH sample was followed-up 3–4 years following their attendance at the pain management program
(Beeston, 2001). Comparison of mean PSEQ scores (by
t tests) at each occasion according to use/non-use of
medication for pain revealed significant differences in
M.K. Nicholas / European Journal of Pain 11 (2007) 153–163
159
Table 4
Comparison of mean PSEQ (SD) scores from a sample of heterogeneous chronic pain patients attending a pain management program at RNSH
according to whether they were taking medication or not at pre-treatment, post-treatment and 3–4 year follow-up (n = 130)
Medication status
Occasions
Pre-treatment
Taking
Not taking
Significance
Post-treatment
F/U (3–4 years)
N
PSEQ (SD)
N
PSEQ (SD)
N
PSEQ (SD)
114
16
19.3 (10.8)
26 (10.4)
P < 0.021
28
102
32 (12.6)
38.6 (12.8)
P < 0.015
43
77
26.2 (13)
36.3 (13.6)
P < 0.001
PSEQ scores between the two groups at each stage (see
Table 4). Those not taking medication for pain reported
significantly higher self-efficacy levels at each stage (at
pre-treatment, t = 2.34, df = 128, p < 0.021; at posttreatment, t = 2.47, df = 128, p < 0.015; at follow-up,
t = 3.97, df = 118, p < 0.001).
Overall, there is close correspondence between the
psychometric properties of the PSEQ with this large
heterogeneous sample of chronic pain patients and the
(relatively small) original Westmead sample of low back
pain patients.
4. Discussion
This paper sought to examine a measure designed to
assess self-efficacy beliefs in people with chronic pain.
Unlike most published measures of self-efficacy beliefs
in people with chronic pain, the PSEQ explicitly asks
respondents to take their pain into account.
The results obtained from a number of studies reveal
that the PSEQ has strong psychometric properties. A
high degree of reliability was reflected in high internal
consistency and high stability across time (up to 3months) under conditions of pain-focussed treatment
and no change in pain or disability (Asghari and Nicholas, 2001; Williams et al., 1996).
The PSEQ’s validity was reflected in high correlations
(in expected directions) with measures of pain-related disability and different coping strategies. Importantly also, a
number of studies have found scores on the PSEQ to be
related to a range of complex behaviours amongst different chronic pain samples. These inlcude dropout from a
pain management program (Coughlan et al., 1995), pain
behaviours (Asghari and Nicholas, 2001), as well as work
status (Cohen et al., 2000; Adams and Williams, 2003),
medication use (Ralphs et al., 1994; Beeston, 2001), and
interference in daily activities observed by the patients’
significant-others (Nicholas et al., 1992). Strong et al.
(2002) also found the PSEQ to be a better predictor of
functional outcomes than the Pain Stages of Change
Questionnaire (PSOCQ; Kerns et al., 1997) after a pain
management program, providing further support for the
PSEQ’s construct validity.
The evidence of the PSEQ’s sensitivity to change
under conditions when it would be expected to change
also provides support for its construct validity (Murphy
and Davidshofer, 1988). Nicholas et al. (1992) reported
significant improvements in PSEQ, SIP (spouse-rated),
medication-use, and CSQ in low back pain patients after
cognitive-behavioural treatment (that included exercises) compared to a control condition. Similarly, Williams et al. (1993) reported that patients attending an
inpatient pain management program achieved significant improvements on measures of mood, medication
use, disability, specific activities (e.g., stairs climbed)
and the PSEQ, but not on pain severity ratings.
The PSEQ’s validity was also supported by a high correlation with another, more activity-specific measure of
self-efficacy beliefs (Self-Efficacy Scale; SES; Kaivanto
et al., 1995) (Gibson and Strong, 1996). Significantly, in
the light of Bandura’s proposition about performance
in the face of adversity, the PSEQ was more strongly associated with perceived work capacity in injured workers
with chronic pain than the SES, which did not incorporate the presence of pain as a context. This finding could
be seen as supportive of the importance of taking into
account the context (i.e., pain) in which the tasks or functions were to be performed when assessing self-efficacy
beliefs, rather than just the task itself – similar to findings
in the social anxiety literature (Alden et al., 2004).
The finding that correlations between the PSEQ and
pain experience (pain severity and MPQ subscales),
and other somatic perceptions (MSPQ), were generally
quite low suggests that while pain experience is reflected
in self-efficacy beliefs (in an inverse relationship), other
factors contribute to self-efficacy beliefs as well – as
argued originally by Bandura and by more recent psychological accounts about adjustment to pain (e.g.,
Turk, 2002b; Keefe et al., 2004). Treatment studies that
have shown improvements in self-efficacy (PSEQ) and
disability, but not pain severity, would also be consistent
with this view (Nicholas et al., 1992; Williams et al.,
1993, 1996).
Bandura’s self-efficacy theory would predict that a
generalised measure of self-efficacy as a resilient selfbelief system in the face of difficulties would change
in the light of personal achievements (in performance),
observation of others performing relevant behaviours,
and verbal persuasion. All three elements could be
said to exist in (group) cognitive-behavioural pain
management programs. As already noted, such
160
M.K. Nicholas / European Journal of Pain 11 (2007) 153–163
changes in self-efficacy beliefs measured by the PSEQ
have been repeatedly found in a number of cognitive-behavioural treatment studies. These findings also
suggest that the lack of correspondence between selfefficacy beliefs and functional activities reported by
Altmaier et al. (1993) may well have been due to their
measure’s item content rather than more fundamental
questions about the relationship between self-efficacy
beliefs and activities when the presence of pain is
taken into account.
It might also be expected that self-efficacy for performing tasks despite pain would have an inverse relationship to fears about pain and possible injury. This
relationship has been reported in two studies (Watson
et al., 1997; Ayer and Tyson, 2001). Interestingly, Ayer
and Tyson found that the PSEQ accounted for a greater
proportion of the variance in disability scores (in
patients with chronic low back pain) than fear-avoidance beliefs assessed by the Fear-Avoidance Beliefs
Questionnaire (FABQ; Waddell et al., 1993). As both
self-efficacy and fear-avoidance beliefs have been shown
to be associated with disability in chronic pain patients
it would be important to explore the nature of these relationships. One possible scenario is that a person in
chronic pain could be fearful of an activity they expected
to aggravate their pain, but whether or not they performed the activity might depend on how confident they
were in being able to do it despite the pain.
It is also likely that pain self-efficacy beliefs would
interact with acceptance of pain (e.g., McCracken and
Eccleston, 2003). For example, it might be predicted
that even if a person accepted the chronic nature of their
pain, but their pain self-efficacy was low, the chances of
their performing an activity that could aggravate their
pain would be low. On the other hand, it might also
be predicted that if pain acceptance was low and pain
self-efficacy was high, the peson might be more inclined
to seek pain-relieving treatment despite having few functional limitations (they might say ‘‘I just don’t want to
live in pain’’). These predictions also remain to be tested.
Apart from theoretical considerations, the findings
presented here provide support for the use of the PSEQ
in clinical settings as both a screening instrument to
determine patients’ beliefs about a fundamental aspect
of pain management (i.e., confidence in performing normal activities despite pain) and in evaluating outcomes
after treatment. As a screening instrument, it can provide an indication of their receptivity to an intervention
(like a pain program) that did not offer significant pain
relief. A very low score (e.g., <17, from Coughlan
et al., 1995) could be interpreted as the patient believing
pain relief was necessary before s/he could become more
active. In this case, these beliefs would need to be
addressed prior to such a person starting a pain management program. On the other hand, a high PSEQ score at
initial assessment (say, over 40, from Williams et al.,
1996; Cohen et al., 2000) would appear incongruent with
someone needing a pain management program. This
should prompt further examination of the reasons for
being there.
The finding that high PSEQ scores following pain
management programs are commonly strongly associated with clinically significant functional gains provides
a potentially useful guage for evaluating outcomes in
chronic pain patients. In this context the PSEQ score
may provide an indication of the likely maintenance of
any behavioural changes or even likelihood of resumption of work given the presence of pain. Thus, scores of
around 40 post-treatment (as found in injured workers
who had returned to work by Cohen et al. (2000), Adams
and Williams (2003); and in patients who generally maintained their treatment gains at 6- and 12-month followups: Williams et al., 1993, 1996) might be considered to
have reached a type of threshold where maintenance of
gains or return to work was reasonably likely. On the
other hand, scores around 30 post-treatment might suggest that things were more in the balance and less likely
to be predictive of return to work or maintenance of
treatment gains (e.g., Coughlan et al., 1995).
Bandura’s self-efficacy theory would predict that
those patients who make behavioural changes in a pain
management treatment but do not report a sufficient
increase their level of pain self-efficacy would be at risk
of relapse or drop-out. Coughlan et al.’s (1995) findings
could be seen as consistent with that. Council et al.
(1988) also described this phenomenon with a different
self-efficacy scale. While this issue needs further investigation, it would clearly be a potentially useful means of
identifying possible ‘at risk’ cases at the end of a pain
management treatment who could be offered further
help as a means of preventing possible drop-out or
relapse.
To conclude, the findings obtained with the PSEQ
across a number of studies with different pain populations provide support for the idea originally
espoused by Bandura (1989) that it is useful to conceptualise self-efficacy as a reflection of a ‘resilient
self-belief system’ in the face of obstacles. By specifying the nature of the obstacles to be faced (i.e., pain)
the PSEQ may provide more clinically useful information than simply asking someone about their confidence in performing an activity in isolation. The
PSEQ psychometric strengths provide support for its
use in both clinical and research settings (Jensen,
2003).
Acknowledgements
I acknowledge the assistance of the following people
with aspects of the development of this paper: Amanda
Williams, Mohsen Dehghani, Ali Asghari, Toby Newton-John and the anonymous reviewers.
M.K. Nicholas / European Journal of Pain 11 (2007) 153–163
Appendix A
PAIN S-E QUESTIONNAIRE (PSEQ)
M.K.Nicholas, 1989
NAME: __________________________________________ DATE: __________________
Please rate how confident you are that you can do the following things at present, despite the pain.
To indicate your answer circle one of the numbers on the scale under each item, where 0 = not at all
confident and 6 = completely confident.
For example:
0
Not at all
Confident
1
2
3
4
5
6
Completely
confident
Remember, this questionnaire is not asking whether of not you have been doing these things, but rather
how confident you are that you can do them at present, despite the pain.
1.
2.
3.
4.
I can enjoy things, despite the pain.
0
1
2
3
4
5
6
Not at all
Completely
Confident
confident
I can do most of the household chores (e.g., tidying-up, washing dishes, etc.), despite the
pain.
0
1
2
3
4
5
6
Not at all
Completely
Confident
confident
I can socialise with my friends or family members as often as I used to do, despite the
pain.
0
1
2
3
4
5
6
Not at all
Completely
Confident
confident
I can cope with my pain in most situations.
0
Not at all
Confident
5.
6.
7.
1
2
3
4
5
6
Completely
confident
I can do some form of work, despite the pain. (“work” includes housework, paid and
unpaid work).
0
1
2
3
4
5
6
Not at all
Completely
Confident
confident
I can still do many of the things I enjoy doing, such as hobbies or leisure activity, despite
pain.
0
1
2
3
4
5
6
Not at all
Completely
Confident
confident
I can cope with my pain without medication.
0
1
2
3
4
Not at all
Confident
5
6
Completely
confident
8.
I can still accomplish most of my goals in life, despite the pain.
0
1
2
3
4
5
6
Not at all
Completely
Confident
confident
9.
I can live a normal lifestyle, despite the pain.
0
1
2
3
4
Not at all
Confident
10.
5
6
Completely
confident
I can gradually become more active, despite the pain.
0
Not at all
Confident
1
2
3
4
5
6
Completely
confident
161
162
M.K. Nicholas / European Journal of Pain 11 (2007) 153–163
References
Adams JH, Williams ACdeC. What affects return to work for
graduates of a pain management program with chronic upper limb
pain? J Occ Rehab 2003;13:91–106.
Alden LE, Mellings TMB, Laposa JM. Framing social information
and generalized social phobia. Behav Res Ther 2004;42:585–600.
Altmaier EM, Russel DW, Kao CF, Lehmann TR, Weinstein JN. Role
of self-efficacy in rehabilitation outcome among chronic low back
pain patients. J Counsel Psychol 1993;40:335–9.
Anderson KO, Dowds BN, Pelletz RE, Edwards WT, PeetersAsdourian C. Development and initial validation of a scale to
measure self-efficacy beliefs in patients with chronic pain. Pain
1995;63:77–84.
Arnstein P, Caudill M, Mandle C, Norris A, Beasley R. Self-efficacy as
a mediator of the relationship between pain intensity, disability and
depression in chronic pain patients. Pain 1999;80:483–91.
Asghari A, Nicholas MK. Pain self-efficacy beliefs and pain behaviour:
a prospective study. Pain 2001;94:85–100.
Ayer M, Tyson GA. The role of self-efficacy and fear-avoidance beliefs
in the prediction of disability. Aust Psychol 2001;36:250–3.
Bandura A. Self-efficacy: toward a unifying theory of behavioural
change. Psychol Rev 1977;84:191–215.
Bandura A. Perceived self-efficacy in the exercise of personal agency.
The Psychologist: Bul Brit Psychol Soc 1989;10:411–24.
Bandura A, O’Leary A, Taylor CB, Gauthier J, Gossard D. Perceived
self-efficacy and pain control: opioid and nonopioid mechanisms. J
Pers Soc Psychol 1987;53:563–71.
Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory
for measuring depression. Arch Gen Psychiat 1961;4:561–71.
Beeston L. Long term follow-up of the cognitive-behavioural pain
management programme, ADAPT. Unpublished MSc (Med; Pain
Management) thesis, University of Sydney; 2001.
Bergner M, Bobbitt RA, Carter WB, Gilson BS. The Sickness Impact
Profile: development and final revision of a health status measure.
Med Care 1981;19:787–805.
Brown GK, Nicassio PM. Development of a questionnaire for the
assessment of active and passive coping strategies in chronic pain
patients. Pain 1987;31:53–64.
Budzynski T, Stoyva J, Adler LS, Mullaney DJ. EMG biofeedback
and tension headache: a controlled outcome study. Psychosom
Med 1973;35:484–96.
Cohen M, Nicholas MK, Blanch A. Medical assessment and management of work-related low back or neck-arm pain: more questions than
answers. J Occ Health Safety, Aust New Zealand 2000;16:307–17.
Coughlan GM, Ridout KL, Williams ACdeC, Richardson PH.
Attrition from a pain management program. Brit J Clin Psychol
1995;34:471–9.
Council JR, Ahern DK, Follick MJ, Kline CL. Expectancies and
functional impairment in chronic low back pain patients. Pain
1988;33:323–32.
Dehghani M, Sharpe L, Nicholas MK. Modification of attentional
biases in chronic pain patients: a preliminary study. Eur J Pain
2004;8:585–94.
Dolce JJ, Crocker MF, Moletteire C, Doleys DM. Exercise quotas,
anticipatory outcome and self-efficacy expectancies in chronic pain:
a preliminary report. Pain 1986;24:365–72.
Eccleston C. The attentional control of pain: methodological and
theoretical concerns. Pain 1995;63(1):3–10.
Edwards R, Telfair J, Cecil H, Lenoci J. Reliability and validity of a
self-efficacy instrument specific to sickle cell disease. Behav Res
Ther 2000;38:951–63.
Estlander A-M, Vanharanta H, Moneta GB, Kaivanto K. Anthropmetric variables, self-efficacy beliefs, and pain and disability ratings
on the isokinetic performance of low back pain patients. Spine
1994;19:941–7.
Frost H, Klaber Moffett J, Moser J, Fairbank J. Evaluation of a fitness
programme for patients with chronic low back pain. Brit Med J
1993;310:151–4.
Gibson L, Strong J. The reliability and validity of a measure of
perceived functional capacity for work in chronic back pain. J Occ
Rehab 1996;6:159–75.
Gottlieb BS. Development of the Pain Belief Questionnaire: a
preliminary report. In Paper presented at the Association for the
Advancement of Behavioural Therapy, Philadelphia, PA; 1984.
Jensen MP. Questionnaire validation: a brief guide for readers of the
research literature. Clin J Pain 2003;19:345–52.
Jensen MP, Turner JA, Romano JM. Self-efficacy and outcome
expectancy relationship to chronic pain, coping strategies and
adjustment. Pain 1991;44:263–9.
Kaivanto K, Estlander A-M, Moneta GB, Vanharanta H. Isokinetic
performance in low back pain patients: The predictive power of the
Self-Efficacy Scale. J Occ Rehab 1995;5:87–99.
Keefe FJ, Lefebvre JC, Maixner W, Salley AN, Caldwell DS. Selfefficacy for arthritis pain: relationship to perception of thermal
laboratory pain stimuli. Arthritis Care Res 1997;10(3):177–84.
Keefe FJ, Rumble ME, Scipio CD, Giordano LA, Perri LM.
Psychological aspects of persistent pain: current state of the
science. J Pain 2004;5(4):195–211.
Kerns RD, Haythornthwaite J. Depression among chronic pain
patients: cognitive-behavioral analysis and effect on rehabilitation
outcome. J Consult Clin Psychol 1988;56:870–6.
Kerns RD, Turk DC, Rudy TE. The West Haven Yale Multidimensional Pain Inventory (WHYMPI). Pain 1985;23:345–56.
Kerns RD, Rosenberg R, Jamison RN, Caudill MA, Haythwornthwaite J. Readiness to adopt a self-management approach to chronic
pain: the pain stages of change questionnaire. Pain 1997;72:227–34.
Lackner JM, Carosella AM. The relative influence of perceived pain
control, anxiety and functional self-efficacy on spinal function
among patients with chronic low back pain. Spine
1993;24:2254–60.
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.
Lorig K, Chastain RL, Shoor EUS, Holman HR. Development and
evaluation of a scale to measure perceived self-efficacy in people
with arthritis. Arthrit Rheumat 1989;32:37–44.
Main CJ. The modified somatic perception questionnaire. J Psychosom Res 1983;27:503–14.
McCracken LM, Eccleston C. Coping or acceptance: what to do about
chronic pain? Pain 2003;105:197–204.
Melzack R. The McGill Pain Questionnaire: major properties and
scoring methods. Pain 1975;1:277–99.
Murphy KR, Davidshofer CO. Psychological testing: principles and
applications. London: Prentice-Hall; 1988.
Nicholas MK. Self-efficacy and chronic pain. In Paper presented at the
annual conference of the British Psychological Society, St.
Andrews, Scotland; 1989.
Nicholas MK, Wilson PH, Goyen J. Comparison of cognitive
behavioural group treatment and an alternative non-psychological
treatment for chronic low back pain. Pain 1992;48:339–47.
Nunnally JC, Bernstein IH. Psychometric theory. 3rd ed. New
York: McGraw-Hill; 1984.
Ozer EM, Bandura A. Mechanisms governing empowerment effects: a
self-efficacy analysis. J Pers Soc Psychol 1990;58(3):472–86.
Pincus T, Morley S. Cognitive appraisal. In: Linton SJ, editor. New
avenues for the prevention of chronic musculoskeletal pain and
disability. Pain research and clinical management, vol. 12. Amsterdam: Elsevier; 2002. p. 123–42.
Ralphs JR, Williams ACdeC, Richardson PH, Pither CE, Nicholas
MK. Opiate reduction in chronic pain patients: a comparison of
M.K. Nicholas / European Journal of Pain 11 (2007) 153–163
patient-controlled reduction and staff controlled cocktail methods.
Pain 1994;56:279–88.
Roland M, Morris S. A study of the natural history of back pain. Part
I: development of a reliable and sensitive measure of disability in
low-back pain. Spine 1983;8:141–4.
Rosenstiel AK, Keefe FJ. The use of coping strategies in chronic low
back pain patients: relationship to patient characteristics and
current adjustment. Pain 1983;17:33–44.
Rudy TE, Lieber SJ, Boston JR, Gourley LM, Baysal E. Psychosocial
predictors of physical performance in disabled individuals with
chronic pain. Clin J Pain 2003;19:18–30.
Spielberger C, Gorsuch R, Luschere N. Manual for the State Trait
Anxiety Inventory. Palo Alto (CA): Consulting Psychology Press;
1970.
Strong J, Westbury K, Smith G, McKenzie I, Ryan W. Treatment
outcome in individuals with chronic pain: is the Pain Stages of
Change Questionnaire (PSOCQ) a useful tool? Pain 2002;97:
65–73.
Turk DC. A diathesis-stress model of chronic pain and disability
following traumatic injury. Pain Res Manage 2002;7:9–19.
Turk DC. Clinical effectiveness and cost-effectiveness of treatments for
patients with chronic pain. Clin J Pain 2002;18:355–65.
Turner JA. Comparison of group progressive-relaxation training and
cognitive-behavioural group therapy for chronic low back pain. J
Consult Clin Psychol 1982;50:757–65.
Turner JA, Calsyn DA, Fordyce WE, Ready LB. Drug utilization
patterns in chronic pain patients. Pain 1982;12:357–63.
163
Turner JA, Clancy S. Comparison of operant behavioural and
cognitive-behavioural group treatment for chronic low back pain.
J Consult Clin Psychol 1988;56:261–6.
Turner JA, Jensen MP, Romano JM. Do beliefs, coping, and
catastrophizing independently predict functioning in patients with
chronic pain? Pain 2000;85:115–25.
Waddell G, Newton M, Henderson I, Sommerville D, Main CJ. A
fear-avoidance beliefs questionnire (FABQ) and the role of fearavoidance beliefs in chronic low back pain and disability. Pain
1993;52:157–68.
Watson PJ, Booker CK, Main CJ. Evidence for the role of
psychological factors in abnormal paraspinal activity in
patients with chronic low back pain. J Musculoskel Pain
1997;5:41–56.
Williams ACdeC, Nicholas MK, Richardson PH, Pither CE, Justins
DM, Chamberlain JH. Evaluation of a cognitive behavioural
programme for rehabilitating patients with chronic pain. Brit J Gen
Pract 1993;43:513–8.
Williams ACdeC, Richardson PH, Nicholas MK, Pither CE,
Harding VR, Ralphs JA. Inpatient versus outpatient pain
management: results of a randomised controlled trial. Pain
1996;66:13–22.
Williams ACdeC, Nicholas MK, Richardson PH, Pither CE, Fernandes J. Does randomisation affect the generality of findings from a
controlled trial? The effects of patient preference versus randomisation on inpatient versus outpatient chronic pain management.
Pain 1999;83:57–65.