PreprintPDF Available

Chronic Pain Assessment: Scales, Methods, and their Psychometric Properties

Preprints and early-stage research may not have been peer reviewed yet.

Abstract and Figures

Chronic pain assessment, a psychometric properties review for phase-1 of the qualifying exam.
Content may be subject to copyright.
Chronic Pain Assessment: Scales, Methods, and their Psychometric Properties
Aditya Ponnada
Personal Health Informatics
Northeastern University, Boston, MA
Author’s Note
Aditya Ponnada, College of Computer and Information Science and Bouve College of
Health Sciences, Northeastern University, Boston, MA.
This paper is submitted as health sciences literature review part of qualifying exams for
Personal Health Informatics PhD program. Contact:,
Chronic pain is a prevalent health problem affecting overall quality of life. Unlike the acute
pain caused due to tissue damage, chronic pain persists 3- 6 months or longer even after the
treatment or healing period. As a result, it has become increasingly important for researchers
to accurately assess chronic pain and its symptoms to design effective pain management
interventions. Moreover, chronic pain being a subjective experience, cannot be directly
captured using objective sensing methods. Therefore, we rely heavily on self-report methods
of chronic pain assessments.
This paper reviews different methods of assessing chronic pain, its symptoms, and its
effect on everyday living. It covers qualitative interviews, pain drawings, observation guides,
self-report questionnaires, and in situ measurements such as ecological momentary
assessments. For each of the measurement method, its applications, limitations, and
psychometric properties including composite scores, internal reliability, validity, and reactivity
to measurements have been discussed. Finally, this paper finds that there is a scope of
periodically capturing chronic pain episodes in natural settings using self-report on advanced
mobile/wearable technology.
Keywords: Chronic Pain, Pain Assessment, Psychometrics, Self-report, Measurements.
Chronic Pain Assessment: Scales, Methods, and Psychometric Properties
Chronic pain is persistent and recurring for long periods of time, often 3-6 months or longer
resulting in suffering and interference with daily functioning (Merskey, 1986). Unlike acute
pain resulting from an injury, chronic pain often extends beyond the expected period of healing
or treatment (Bonica, 1953; Turk & Okifuji, 2001). In fact, it is a subjective experience reported
or explained by sufferers similar to the pain resulting from tissue damage (Merskey, 1986).
However, even though chronic pain can result from a treatment (as a side effect), a continuation
of an acute pain, or due to an infection or an injury, some individuals experience chronic pain
even without any previous evidence of injury or tissue damage (Pain, 2016), making it a
challenging health construct for research and measurement. Chronic pain is not just limited to
physical experience/sensation as defined by Williams and Craig (2016) as: “a distressing
experience associated with actual or potential tissue damage with sensory, emotional,
cognitive, and social components.
As many as1.5 billion people (~20% worldwide) suffer from chronic pain in their daily life
(Analytics, 2015). As of 2016, it is estimated that 11.2% of the adult popu1lation in the United
States suffers from chronic pain (Dowell, Haegerich, & Chou, 2016). According to the Institute
of Medicine of the National Academies, 100 million Americans are diagnosed with chronic
pain, a much larger number than the total number of Americans affected by diabetes (25.8
million), coronary heart disease (16.3 million), and cancer (11.9 million ), (Steglitz, Buscemi,
& Ferguson, 2012). In fact, 63% of the chronic pain sufferers require clinical help from their
health care providers (Steglitz et al., 2012). Studies have associated the prevalence of chronic
pain to the loss of more than 18 hours of productivity per week for each affected individual in
the United States alone (Stewart, Ricci, Chee, Morganstein, & Lipton, 2003). Finally, chronic
pain is known to affect about 20% of the adult Europeans (Häuser et al., 2014), 13% of the
Indians (Dureja et al., 2014), and 15.4% of the Japanese (Nakamura, Nishiwaki, Ushida, &
Toyama, 2011). Clearly, chronic pain needs should be treated as a public health priority
worldwide (Goldberg & McGee, 2011).
An interventionist who intends to design chronic pain management programs as well as an
epidemiologist who intends to describe chronic pain phenomena in a population, both would
need robust measurement of chronic pain experiences. As the definitions of chronic pain
change, the investigators should the newer definitions in their measurement strategies.
Therefore, in this paper, I present a review of different chronic pain assessment methods and
their psychometric properties. This literature review covers methods such as pain drawings
used in clinical settings, psychophysiological correlates of pain episodes, as well as self-report
methods including unidimensional scales, multi-dimensional scales, and repeated assessments
(ecological momentary assessment (EMA)). The current paper has two goals. First, to identify
different chronic pain assessment methods used in previous studies and their potential use in
more in-situ assessment methods such as EMA or more advanced context-sensitive-EMA
(Stephen S Intille, 2007). Second, to examine the assessment methods that measure pain
incorporating other aspects of chronic pain such as the cognitive, sensory, and emotional
aspects of chronic pain.
A common misunderstanding about the sources of chronic pain is that it always results
from structural abnormalities (e.g., bones rubbing against each other). However, studies have
shown that there is a poor relationship between these structural abnormalities and chronic pain
such as in case of osteoarthritis of knee (Creamer & Hochberg, 1997). There is little evidence
of a strong association between the intensity of injury/tissue damage (due to nociceptive input)
to the level of chronic pain. Some studies have even claimed that much of the chronic
musculoskeletal pain stems from our autonomic nervous system.
Chronic pain can be categorized as either peripheral neuropathic or central neuropathic
pain (Woolf & Mannion, 1999). Peripheral neuropathic chronic pain is caused due to the
damage to peripheral nerves, such as in case of diabetic neuropathic pain or carpel tunnel
syndrome (Woolf & Mannion, 1999). On the other hand, central neuropathic chronic pain is
caused due to the damage to our central pain processing system (Khedr et al., 2005), such as in
case of Fibromyalgia and headache related pain. Pharmacologically, neuropathic pain is
generated by C-fibres and -fibres (Dray, Urban, & Dickenson, 1994). Following tissue
damage or an injury, these C-fibres induce hypersensitivity in neural pathways, causing
persistent high sensation of pain (also known as hyperalgesia). Likewise, Aß-fibres induce
hypersensitivity with low intensity stimulus (e.g., touch or pressure) resulting in low intensity
chronic pain (also known as allodynia). Nevertheless, an individual can experience any
combination of the peripheral neuropathic and central neuropathic pain, making the assessment
of chronic pain challenging with self-report methods. As a result, the International
Classification of Diseases proposed a new multi-layered classification (Table 1 in the
appendix) of chronic pain based on pain aetiology, underlying pathophysiological mechanisms,
and body site/organ (Treede et al., 2015).
Chronic pain also diminishes sleep quality (Smith & Haythornthwaite, 2004) and
increases physical inactivity (Nilsen, Holtermann, & Mork, 2011), resulting in poorer quality
of life. Chronic pain affects an individual, their immediate social circles, quality of life, and
emotional well-being. However, advances in pain management and treatment in part depends
on our ability to accurately capture the episodes, antecedents, and after effects of chronic pain
(R. N. Jamison et al., 2001). These measurements are used for screening, monitoring as well
as evaluating the effect of pain management programs.
Pain assessment methods can also help us computationally model how biological,
psychosocial, and behavioural factors influence our perceptions of pain (Dansie & Turk, 2013).
Likewise, several treatments and prescriptions tend to have painful side effects resulting in
prolonged chronic pain such as breast cancer surgeries (Poleshuck et al., 2006) and thoracic
cancer treatments (Perttunen, Tasmuth, & Kalso, 1999). Thus, chronic pain assessment can
also help in evaluation of the effectiveness of these treatments in terms of the resulting side
effects. In addition, accurate pain assessments can also help in building more robust software
solutions that can support just-in-time interventions (e.g., (S.S. Intille, 2003; S.S. Intille & Ho,
2004) for pain management.
Chronic Pain Assessment: Overview
A keyword search for “chronic pain assessment” alone yields 11,009 citations in
PubMed. In fact, in the last 15 years, 9522 studies listed in the PubMed have used chronic pain
assessment tools. This not only highlights the importance of this construct but also displays an
active research community working in this domain.
Qualitative and Observational Methods
Researchers studying pain diagnosis and management using cognitive-behavioural
approach have used in person interviews to gather information on patients’ adherence to
therapeutic treatments, their emotional states, their beliefs about painful experiences as well as
current medications and their side effects (e.g. (Bradley, 1984)). An extension of the qualitative
interviews is observation-based methods such as Pain Behaviour Checklist (Richards,
Nepomuceno, Riles, & Suer, 1982), where pain-related behaviours and their frequencies are
assessed by observing the diagnosed patient.
Self-report Questionnaires
Self-report surveys are the most commonly used methods to assess chronic pain. These
surveys include unidimensional standard pain intensity questionnaires such as the Numeric
Rating Scale (NRS), the Verbal Rating Scale (VRS), and the Visual Analogue Scale (VAS)
(Price, Bush, Long, & Harkins, 1994), Facial Pain Scale (Bieri, Reeve, Champion, Addicoat,
& Ziegler, 1990), and the Pain Thermometer (K. Herr, Spratt, Garand, & Li, 2007). Some of
these tests measure pain intensity as well as the body location where the pain is perceived such
as the McGill Pain Questionnaire (Melzack, 1975) and the Neuropathic Pain Scale (Galer &
Jensen, 1997). Certain scales also measure the chronic pain interference in daily life such as
the Pain Disability Index (Pollard, 1984) measuring pain’s effect on functionality and social
interactions. Similarly, some multidimensional scales are disease specific such as the
Fibromyalgia Impact Questionnaire for arthritis (Burckhardt, Clark, & Bennett, 1991).
Pain Drawings
Clinical psychologists have used pain drawings as a method to assess chronic pain
experience of their patients (e.g., Phillips, Ogden, and Copland (2015), Margolis, Tait, and
Krause (1986) and Margolis, Chibnall, and Tait (1988)). In this method, patients are asked to
sketch an image of their pain intensity and location on a body chart. These images are later
used to infer patient’s experience and beliefs about their pain.
Physiological Methods
Even though the roots of chronic pain are a topic of debate, it certainly results in a
biological response causing physiological changes. In fact, recent neuroimaging studies have
shown evidence of chronic pain having an effect on central pathology (Tracey & Bushnell,
2009). Therefore, even though there are no independent objective sensors measuring chronic
pain, it is reasonable to assume that there could be psychophysiological correlates of chronic
pain that can be measured using objective sensors. For instance, Collins, Cohen, Naliboff, and
Schandler (1982) have suggested that chronic back pain patients exhibit similar or significantly
less paraspinal muscle activity and significantly higher frontalis electromyography activity and
skin conductance than normal individuals. Similarly, other physiological responses such as
heart rate, heart rate variability (HRV), and electro dermal activity (EDA) undergo changes
when a patient experiences pain.
Diary Methods and Ecological Momentary Assessments
Though researchers cannot measure pain directly using objective sensors, we can gather
information of similar density using repeatedly administered self-report methods such as
experience sampling or EMA. Prior studies have shown that daily diary studies for chronic
pain patients are a reliable approach to gather longitudinal data (Follick, Ahern, & Laser-
Wolston, 1984). Extending the traditional diaries to digital devices Stinson et al. (2006) and
Robert N. Jamison et al. (2001) have used a PDA-based digital diary to enable automatic time
stamping of pain episodes. Studies have shown that chronic pain is highly momentary in nature
and therefore requires repeated assessments (e.g., (Arthur A. Stone, Joan E. Broderick, Saul S.
Shiffman, & Joseph E. Schwartz, 2004).
Different approaches to assess chronic pain have their applications depending on the
patient population, purpose of the study, as well as the infrastructure available to administer
these tests. Similarly, as the measurement method changes, the statistical approaches and
psychometric properties of the tests also change. Therefore, it is essential to review the
measurement tests as well as their psychometric properties such as composite scores,
reliability, variability, and validity. Though the literature covered in this paper may not be
comprehensive, it provides a brief overview of different methods used so far. Moreover, the
scope of this literature review is limited to studies that used English versions of the
questionnaires. Evaluation of translations of the assessments is beyond the scope of this review.
Qualitative and Observational-based Assessments
There has been very limited research in the use of qualitative methods such as
interviewing and contextual inquiry in studying chronic pain events, intensities, and episodes.
Researchers interested in using cognitive-behavioural therapy in pain management have used
in-person semi-structured interviews to understand patient’s pain experiences. For instance,
Bradley (1984) has discussed in person interview from their experience with chronic pain
patients and using this information to assess their pain episodes, emotional responses to painful
experiences, and medication and treatment adherence. However, their work gives little
evidence of the effectiveness of cognitive behaviour therapy for chronic pain management.
Moreover, qualitative methods do not make use of statistical analysis techniques, and therefore
are often not directly suitable to mathematically model a research construct such as chronic
pain. Nevertheless, unlike self-report surveys (covered in later section), qualitative methods
can help researchers and clinicians explore the contextual factors associated with the pain. They
allow patients to explain their experiences in more detail than any self-report methods.
Like qualitative interviews, observation-based methods are used to assess pain and
related behaviours without directly engaging with the participants. For instance, Richards et al.
(1982) have introduced “UAB pain behaviour scale”. This scale consists of 10 items targeting
10 pain related behaviours such as verbal/non-verbal vocal complaints, down time, standing
posture, facial grimaces, mobility, body language, stationary movement, medication, and use
of visible supportive equipment. The observers are asked to rate the frequency of these
behaviors on a test sheet. When used in an interdisciplinary team of pain researchers, the
interrater reliability (Cohen’s Kappa) was 0.95 (p < 0.01) and a test-retest reliability was r =
0.89 (p<0.01) for the latency of 1 day. However, this scale had a poor correlation with other
chronic pain self-report measure such as McGill Pain Questionnaire (Melzack, 1975), (0.16 at
the beginning of the pain management program and 0.55 at discharge time), but had significant
negative correlations (between -0.40 to -0.65) with wellness behaviours such as walking,
biking, and standing. This suggests that UAB might be measuring a pain-related construct
(such as pain-related disability) but not directly the symptoms of chronic pain. On the other
hand, Gramling and Elliott (1992) in their study with 48 chronic pain patients have identified
significant relationship between VAS of pain intensity with sensory and effective subscales of
UAB questionnaire. In fact, these VAS scores accounted for 27.7% (p<0.001) variance in UAB
scores. One limitation of the observation guides is that they must be administered in the
presence of an expert or a health professional trained to observe the nuances of chronic pain
related behaviours. However, observation guides are best suited to conditions when patients
are not able to clearly express their pain experiences. Finally, it is advised to conduct
observation first followed by self-report to avoid any observer biases with chronic pain
observational guides.
Pain Drawings
Pain drawings are widely used in clinical settings to diagnose the regions of chronic
pain on the body and their intensities (Palmer, 1949). Chronic pain patients are given a sheet
of paper with human anatomical diagram (as shown in the appendix figure 2), with marked
anatomical regions. Patients are then asked to shade the anatomical regions where they
experience pain. The intensity of pain (also known as pain modalities) is marked using different
shading codes indicating numbness, stabbing pain, burning pain, stiffness, pins and needles
like pain, dull aching, and cramps (Grunnesjo, Bogefeldt, Blomberg, Delaney, & Svardsudd,
2006). At the end of the pain drawing session, independent scorers rate pain intensities based
on these drawings. Inter-rater reliability is measured to gauge high agreement between the
scorers, before prescribing any medication or program to the patients. Pain drawings have
shown high interrater reliability between 0.7 and 0.8 (e.g., (Uden, Astrom, & Bergenudd,
1988). Researchers have also designed rating systems to evaluate and interpret pain-drawings
in clinical settings. For instance, Margolis et al. (1986) have proposed a body surface system
to quantify the data from pain-drawings, which was proved to be reliable even with non-expert
evaluators. It evaluates chronic pain experience based on the body surface area shaded by the
patient on the pain-drawings.
Pain drawings are best suited to study the localization of pain on the body. However,
the outcome from the pain drawings must not be confused with the actual causes of the chronic
pain in the patient. As discussed earlier, pain can be experienced on a site of the body,
irrespective of its origin. Nevertheless, several studies described here have used pain drawings
in addition to other pain experience scales such as pain disability index and VAS. A typical
pain drawing assessment includes:
Pain Drawing Score
Pain drawing score is measured as the mean number of areas with at least one mark in
the anatomical diagram. It highlights the spread of pain experience throughout the body. The
higher the pain drawing score, larger is the spread of pain on the body (Melzack, 1975).
Pain Modality Dominance
Each pain modality represents different intensity of pain and is ranked in terms of
dominance in patient’s pain experience. Pain modality dominance is estimated based on the
frequency of a pain modality used in an anatomical region. The higher the frequency of a pain
modality, higher is its dominance (Melzack, 1975).
As an alternative to pain drawings, Phillips et al. (2015) had participants describe their
own experiences of pain in the form of images and drawings. Unlike traditional pain-drawing
method, participants could express their pain feelings in their own drawings, not being
restricted by an anatomical figure. These drawings included abstract imageries such as fire and
anthropomorphic characters with real human-like expressions. However, more research is
needed to establish this method in clinical practice. New rating scales must be designed to
evaluate their reliability and validity.
Pain drawing methods make pain related communications easier between clinicians and
patients by reducing the assessment session to a single drawing sheet. Moreover, being a
graphical assessment tool, pain-drawings might not be affected due to language and age
barriers. Pain drawings have also been shown to have high test-retest reliability (Margolis et
al., 1988). However, this method has several limitations. Pain drawings have not been used in
a repeated assessment settings, where pain sensation tends to fluctuate regularly. Moreover,
pain-drawings rely on patients’ recall over a period (sometimes weeks and months). Therefore,
describing pain experience in one drawing sheet as a summary of an accumulated experience
may not be an easy task and may not result in a reliable pain estimate. Finally, research
community is yet to arrive at a consensus on pain drawing scores and their relationship with
patient’s psychological states.
Ginzburg, Merskey, and Lau (1988) have identified poor or limited correlations
between the body surface with pain experience and psychological states (e.g., anxiety r = 0.069
and depression r = 0.114). However, Hayashi et al. (2015) have identified a significant
relationship between pain drawing scores and psychological states based on the region of the
body in pain, especially in neck, shoulder, and lower back regions. Whereas, in a systematic
review Carnes, Ashby, and Underwood (2006) have identified very little evidence of pain-
drawings to be used as psychological assessment tools. Likewise, Von Baeyer, Bergstrom,
Brodwin, and Brodwin (1983) have identified that individuals with psychological involvement
in pain have been diagnosed as normal on pain-drawing tests. Therefore, future assessment
designs for chronic pain may utilize pain-drawings for ease of communication and identifying
the body surface in pain, but they must be wary of its limitations in capturing any psychological
mechanisms resulting in painful sensations.
Physiological and Passive Sensing Methods
Researchers are always interested in measuring health behaviours and states as
objectively as possible. This is primarily because any active participation of the patient comes
with the additional challenge of participation burden and recall bias. Even though chronic pain
is a subjective experience, our bodies do undergo physiological changes and therefore,
available sensing technologies can offer ways to capture chronic pain passively. For instance,
Hallman, Ekman, and Lyskov (2014) have identified diminished heart rate variability in
chronic pain patients, when heart rate was measured using a wrist band monitor (using optical
heart rate sensors). However, their results do not establish a causal link between chronic pain
and heart rate variability. Nevertheless, Martínez-Lavín, Hermosillo, Rosas, and Soto (1998)
have also identified decreased heart rate variability among patients suffering from
fibromyalgia. Likewise, an electrocardiography (ECG) of pain stimulation suggests an
association between pain experiences and lower heart rate variabilities (Appelhans & Luecken,
2008). However, in a controlled experiment Tousignant-Laflamme, Rainville, and Marchand
(2005) have found the heart rate to increase by 11% within 2 minutes of immersion into a pain
stimulation task. In fact, their study has identified a gender effect, where no strong associations
between heart rate and pain intensities and pain unpleasantness for female participants have
been found. While these studies highlight an association between pain experiences and heart
rate variability, they do not suggest a direct causal relationship.
In addition to heart rate variability, electro-dermal activity (EDA) has been shown to
be associated with chronic lower back pain (Peters & Schmidt, 1991), with chronic pain
patients having higher EDA. In fact, EDA has been found to have significant interaction effect
on the diagnosis of depression among the chronic lower back pain patients (Bonnet &
Naveteur, 2004). However, they did not find any significant difference in EDA levels of
chronic back pain patients, temporomandibular pain patients, and healthy individuals. EDA is
known to change with changing levels of anxiety. It is possible that occurrence of chronic pain
might lead to feelings of anxiety resulting in significant changes in EDA. In brief, there needs
to be more research studying the associations of EDA and chronic pain episodes. However, for
context-adaptive assessments, available sensing technologies for heart rate variability and EDA
may play a role. The sensing mechanisms can trigger self-report methods and gather pain
related information with lesser recall bias.
Self-report Questionnaires
The most common method of capturing chronic pain related subjective experience is
through self-report questionnaires. These questionnaires can be administered on paper or in
digital format. Self-report surveys include unidimensional scales (that measure only one
construct of pain), multidimensional scales measuring pain intensity and location, pain
experience scales that capture the effect of chronic pain in daily life, and disease specific pain
experience questionnaires. Table 1 (appendix) below summarizes different pain assessment
scales reviewed in this paper.
Unidimensional Surveys
Unidimensional surveys of pain have single item of pain-related experience. The most
common construct measured using unidimensional single item scale is pain intensity, i.e., the
perceived severity of pain. The five commonly used unidimensional scales include –VRS,
NRS, VAS, FPS, and IPT as shown in figure 3 (appendix).
Verbal rating scale (VRS). In verbal rating scale (Ohnhaus & Adler, 1975), pain intensity
continuum is divided in 5 equal segments with 6-point verbal indicators (in order) – “No pain”,
“Mild pain”, “Moderate pain”, “Severe pain”, “Very severe pain”, and “Worst possible pain”.
Survey respondents select the option most appropriately indicating their perceived pain
intensity. The first point, “No pain” is coded as 0 to provide ratio properties to the scale.
Visual analogue scale (VAS). Visual Analogue Scale (Briggs & Closs, 1999) has a 10-cm long
line with only two ends specified with values. The left-most end (at 0 mm) is referred to as “No
pain at all” and the right-most end (at 100mm) is referred to as “Worst possible pain”. No other
markers are provided on the line. The respondents cross or put a tick-mark at a point on this
axis, which they perceive to be closest to their pain intensity. This scale has ratio properties
and responses are measured as the distance between the 0mm mark and the response mark.
Numeric rating scale (NRS). Just like VAS, NRS (Pagé et al., 2012) has a 10-cm long line but
with markers specified at 1-cm each. NRS has points from 0 to 10, with 0 being no pain to 10
being the worst imaginable pain. Hence, NRS also follows ratio properties with an absolute 0.
However, NRS does not necessarily have to be 10-sm long if the answer options are separated
by 1 unit and the scale is divided into 10 equal parts. Due to equal separation of points, the
measurement properties of NRS are not affected by the length of the scale. This makes NRS
more adaptable on electronic devices with smaller screens.
Faces pain scale (FPS). In faces pain scale (Bieri et al., 1990), the pain intensity continuum
comprises of either five or seven points. Each point is illustrated as a facial expression
resembling the pain experience. The left most point represents a “No pain at all” or 0 to provide
ratio properties. The end points resemble the “worst pain possible”. Participants are required
to select the facial expression that closely resembles their pain sensation.
Iowa pain thermometer (IPT). The Iowa pain thermometer is a combination of VRS and a
colour gradient, where a vertical thermometer with a colour gradient is presented. The colour
gradient indicates the direction of pain intensity with darker shades of gradient indicating
higher pain intensity. Along with the thermometer, there are verbal indicators (from VRS)
placed in the same direction of the gradient.
Several studies have established test-retest reliability of unidimensional scales. For
instance, Ware, Epps, Herr, and Packard (2006) in their study with cognitively impaired and
intact minority adults have found high test-retest reliabilities with r = 0.87 for NRS, r = 0.86
for VRS, r = 0.81 for IPT, and r = 0.76 for FPS. Similarly, Ware et al. (2015) have established
test-retest reliability of 0.79 for IPT and 0.80 for NRS for cognitively intact and impaired older
minority adults. Bieri et al. (1990) in their original work proposing FPS have also reported high
test-retest reliability of FPS with children. Moreover, in a comparison of these scales with
patients admitted to ICU, a high reliability (r = 0.84) was found across NRS and VAS.
Likewise, in a systematic review, Williamson and Hoggart (2005), have identified high test-
retest reliability of VAS (r ~0.97 – 0.99).
Pain intensity scales also have high concurrent and convergent validity. For instance,
Ferreira-Valente, Pais-Ribeiro, and Jensen (2011) in their validity study with Portuguese
population have found strong inter-scale correlations of above 0.79 between all the pain
intensity scales (excluding IPT). Similarly, Ware et al. (2006) in their study have found
spearman correlations between all these scales to range between 0.64 to 0.90. Ware et al. (2015)
in a recent evaluation of IPT have found high convergent validity of 0.95 between ITP and
NRS for current pain, 0.94 for recalled pain, and 0.97 for reassessed pain. Concurrent validity
has been established between FPS and NRS for older adults (Kim & Buschmann, 2006).
Finally, in a controlled experimeriment with a cold presser test, Ferreira-Valente et al. (2011)
have established that NRS, VAS, VRS, and FPS were able to distinguish between different
cold temperatures (that induce pain) in cold presser test. There has been an exhaustive literature
review of the use unidimensional surveys by Hjermstad et al. (2011), where they have reported
on the validity of unidimensional scales across different pain conditions.
However, different unidimensional scales have their own advantages determining their
use in pain research. For instance, FPS has been dedicatedly designed for use with children
(Bieri et al., 1990). Similarly, FPS has been preferred over other unidimensional scales by older
adults (K. A. Herr, Mobily, Kohout, & Wagenaar, 1998) as well as individuals with cognitive
impairments (Ware et al., 2006). This could be because FPS does not demand high verbal
ability from its respondents. However, weaker correlation of FPS with other rating scales could
suggest that FPS measured a construct broader than pain intensity (e.g., overall affect resulting
from pain), since differentiating a pain facial expression from an emotional response is a
challenging task (Ware et al., 2006). Nevertheless, FPS had the least failure rate (i.e. failure to
understand and respond) as compared to other scales used for cognitively impaired minority
adults. FPS is also known to be used and easily reproducible in cross-cultural settings, where
language could be a barrier for researchers and clinicians (e.g., (Ferreira-Valente et al., 2011).
Even though there are very minute differences in VAS and NRS, NRS is easier to
reproduce and administer (Salaffi, Stancati, Silvestri, Ciapetti, & Grassi, 2004). VAS has a
fixed length of 100mm, which often gets restricted when administering VAS on electronic
devices such as smartphones and PDA. Whereas, in NRS the gap between two consequent
points must be the same irrespective of the total length. As a result, NRS is a more scalable
alternative when used on digital devices of varying resolutions. Moreover, NRS is easier to use
by respondents as they find it easier to identify their perceived pain intensity. In fact, in a recent
systematic review NRS and VRS had much higher compliance (survey completion response
rate) than VAS which has recorded least compliance (Hjermstad et al., 2011). However, VAS
offers a more continuous measurement than discrete points on NRS. Nevertheless, VAS and
NRS both had higher effect sizes than other unidimensional scales when distinguishing pain
perception in controlled cold presser tests (Ferreira-Valente et al., 2011). In addition, NRS has
been found to have better discriminant properties than VAS and other scales. For instance,
NRS is better than VRS in distinguishing between background and peak pain intensities
(Brunelli et al., 2010). Finally, NRS has been found to have highest reliability in both literate
and illiterate respondents (e.g., (Ferraz et al., 1990)) and has been identified to be the most
responsive for current pain (Bolton & Wilkinson, 1998) and in chronic pain in general (Grotle,
Brox, & Vøllestad, 2004). This high responsiveness for current pain makes NRS suitable for
use in repeated in-situ assessments.
Although VRS is designed to have ratio properties with “no pain” category being
equivalent to 0, there has been disagreement among researchers about the use of VRS in
chronic pain assessments. Ferreira-Valente et al. (2011) in their study have suggested that VRS
does not have ratio properties and the intervals between two labels might not be perceived as
equal by respondents. Their study has also found VRS to be less responsive than VAS and
NRS. In fact, Ekblom and Hansson (1988) have empirically verified that the VRS categories
are not equidistant in terms of intensity changes in other scale scores (i.e. VAS and NRS), a
property necessary for interval data types. Similarly, Jensen, Karoly, and Braver (1986) have
suggested that VRS possesses more ordinal properties and therefore is more suitable for
parametric tests. However, there has been limited empirical evidence on the distribution of
responses of VRS. Finally, Ohnhaus and Adler (1975) have found that VRS tends to augment
the feeling of pain much more than VAS, resulting in recall biases.
Jamison et al. (2002) have validated an electronic version of VAS (implemented on
palm-top computers) with the paper version of VAS. In a controlled lab study with 24-
participants using test-weights, participants reported pain intensity on both electronic-VAS and
paper-VAS. High correlations (r = 0.91, p <0.0001) between electronic-VAS and paper-VAS
was found. Similarly, for the same sensory stimulus electronic-VAS and paper-VAS had a
correlation of r = 0.98 (p<0.0001). This suggests that VAS (just like NRS) can be translated
into their digital format. However, a formal comparison of using NRS and VAS in electronic
versions is worthwhile.
Multiple-Item/Multidimensional Surveys
Multiple item chronic pain assessments measure more than one constructs related to
pain experience including pain location, pain intensity, and feelings associated with the pain
experience. These assessment scales provide a more comprehensive assessment of pain suitable
for one time measurement and allow researchers to study different correlates of pain in the
same study. The most popular multiple item pain surveys include McGill Pain Questionnaire
(Melzack, 1975), Regional Pain Scale (Wolfe, 2003), and Neuropathic Pain Scale (Galer &
Jensen, 1997). In addition, there are several pain assessment questionnaires that are specific to
diseases and disabilities such as Fibromyalgia Impact Questionnaire for arthritis (Burckhardt
et al., 1991), WOMAC (Bellamy, Buchanan, Goldsmith, Campbell, & Stitt, 1988), Pain
Disability Index (Pollard, 1984), and Brief Pain Inventory (Cleeland & Ryan, 1994).
McGill Pain Questionnaire. McGill Pain Questionnaire has 4 subscales. The first subscale is
a pain drawing sheet as depicted in figure 2. This subscale captures the location of the pain as
well as lets respondents report if the pain in that region is internal, external or both. The second
subscale captures the perceived sensation of the pain. In this subscale, certain qualifying
keywords are presented in 20 categories. The number of words in each category can vary from
2 to 6 and are grouped based on their qualitative similarity. Each keyword in these categories
describes a feeling associated with pain that can fall under either sensory, affective or
evaluative responses. For instance, feelings such as “pinching”, “shooting”, and “cramping”
are sensory feelings. Likewise, “sickening”, “punishing”, and “killing” are affective feelings,
and “discomforting”, “annoying”, and “intense” are evaluative feelings. Respondents are
required to select only one keyword from each of the twenty categories. However, they are free
to uncheck a category if it does not describe their pain feelings appropriately. This subscale
measures their Pain Rating Index (PRI). PRI is estimated in two parts - PRI (S) based on
patient’s mean scale values and PRI (R) based on the rank values of the words. Along with this
questionnaire a cheat sheet is provided to the researchers with the scale values of all the
keywords associated with the pain included in this subscale. PRI (R) is captured using the rank
values assigned to each of the words in different sub-categories. For instance, the word
describing least pain is ranked as 1, the word following that is ranked 2 and so forth. Finally,
the total number of words chosen is also used as a descriptive statistic of the scale.
The third subscale captures the summary of change of pain with time using three items.
First, respondents are asked to select keywords (e.g., Continuous, rhythmic, and brief) from a
corpus that describe the nature of their change in pain followed by the events or activities that
relieve or increase their pain. The fourth 6-item subscale measures the present pain intensity
(PPI) with each item having 5-answer choices (1 – Mild, 2 – Discomforting, 3 – Distressing, 4
– Horrible, and 5 – Excruciating). The items include pain intensity of the current pain, worst
pain, least painful experience, worst toothache, worst headache, and worst stomach ache.
McGill Pain Questionnaire has been used to study pre-and post-treatment pain
sensations. The original work by Melzack (1975) has demonstrated high interclass correlations
(~ r = 0.90) in PRI(S) between sensory, affective, and evaluative descriptors of pain. However,
even though PPI correlations with number of words chosen and PRI are significant, they lie
between 0.3 and 0.4. Much of the variance in PPI could be explained by the fact that each
patient responding to PPI can have their own perception of an appropriate pain intensity. For
instance, a discomforting pain for an individual could be mild for another individual. Likewise,
PPI could also be influenced by external factors such as mood, anxiety, and even the most
recent state of pain. Nevertheless, authors did identify consistency in the choice of words
among patients to describe their pain.
Finally, the PRI subscale scores (scale and rank scores) have shown strong correlation
in various chronic pain problems including menstrual, arthritis, cancer, dental, and lower back
pain (Melzack, 1975). McGill Pain Questionnaire, thus, is suitable to describe a patient’s
chronic pain related experiences irrespective of the condition they are in. It can describe the
location, pain intensity as well as pain frequency of the painful episodes.
However, this questionnaire takes 10 – 15 minutes to complete and stronger vocabulary
skills for an unfamiliar patient and 5 – 10 minutes for a patient familiar with this procedure.
Therefore, researchers have introduced a short-form of McGill Pain Questionnaire, also SF-
MPQ, which takes only 2 – 5 minutes to complete (Dudgeon, Raubertas, & Rosenthal, 1993).
SF-MPQ consists of 15 word descriptors related to sensory and affective dimensions of pain.
Pain intensity in SF-MPQ is measured using present pain intensity of McGill Pain
Questionnaire and VAS described before. SF-MPQ has been particularly found useful for
chronic pain in cancer patients and similar populations who have difficulty concentrating for
prolonged periods of time. It has also been adapted in electronic pain assessments such as Pain-
QUILT (Lalloo, Kumbhare, Stinson, & Henry, 2014).
Regional Pain Scale. The Regional Pain Scale (RPS) is a 19-item scale that captures the pain
occurrences and severity at 19 different locations on the body (Wolfe, 2003). The scale consists
of two dimensions – the location (left and right for each body joint) and severity of the pain.
There are 19 different joint locations including shoulders, elbows, wrists, hand knuckles, finger
knuckles, hips, knees, ankles, ball of foot, heels, foot arch, jaws, lower back, upper back, neck,
upper arms, lower arms, upper legs, lower legs, headache, chest, and abdomen. Each of these
joints is rated for the severity of pain on a 4-point scale where 1 is none, 2 is mild, 3 is moderate,
and 4 is severe. This scale requires participants to recall their pain experience over the past 7-
days. Wolfe (2003) have validated this scale with 12,799 patients suffering from rheumatoid
arthritis, osteoarthritis, and fibromyalgia. In addition to capturing the severity in rheumatic
diseases, RPS can also differentiate fibromyalgia patients from osteoarthritis and rheumatoid
arthritis. Unlike McGill Pain Questionnaire, RPS can only capture location and pain severity
in that location and has been used only for patients with rheumatoid arthritis.
Neuropathic Pain Scale. The Neuropathic Pain Scale (NPS) is a 10-item scale that captures
the pain occurrences and severity of neuropathic pain syndromes, i.e., pain or numbness in
peripheral nervous system (Galer & Jensen, 1997). The scale items measure intensity,
sharpness, hotness, dullness, coldness, sensitivity, and itchiness of the neuropathic pain. The
item 8 in NPS captures the time duration of the pain. Finally, 9th and 10th items of the scale
measure unpleasantness and intensity of deep and surface pain. Each of the 9 items (except
item 8) in the scale are NRS with 11 points (0 to 10).
NPS items have shown to have strong discriminant validity (Galer & Jensen, 1997). In
NPS, pain sensations such as “deep”, “Itchy”, “cold”, and “dull” share less than 25%
(correlation coefficient of 0.50) of the variance with other NPS items. This suggests that NPS
items can distinguish between distinct pain experiences for peripheral nervous system.
Similarly, NPS items have also been shown to have good predictive validity. NPS items could
distinguish post-herpetic neuralgia from other neuropathic pains resulting from reflex
sympathetic dystrophy, diabetic neuropathy, and peripheral nerve injury. However, NPS may
not be able to distinguish between the pain caused due to standing/walking as compared to the
ones caused by lying down or being stationary. Nevertheless, Galer and Jensen (1997) in their
original research have also identified that NPS can also evaluate the effectiveness of different
treatment/medication (e.g., lidocaine vs phentolamine) available on neuropathic pain.
Fibromyalgia Impact Questionnaire. The 10-item Fibromyalgia Impact Questionnaire (FIQ)
measures the impact of fibromyalgia (musculoskeletal pain) in everyday activities, its impact
on our body functions, negative affect, and productivity (Burckhardt et al., 1991). The first
item is a subscale with 10 sub-items measuring the patients’ ability to perform everyday
activities (physical functioning) including shopping (“Were you able to do shopping?”),
laundry (“Were you able to do laundry?”), vacuuming (“Were you able to do vacuuming?”),
and driving a car (“Were you able to drive a car?”). Each sub-item is a 4-point question with 0
as always, 1 as most times, 2 as occasionally, and 3 as never. Later, an additional sub-item on
climbing stairs has been added to FIQ (Bennett, 2005). The next two questions ask the number
of days the participants felt good (scored in reverse) and the number of days they missed work
because of fibromyalgia, respectively. Finally, question numbers 4 through 10 ask about the
severity of pain interference with work, pain intensity, tiredness, feelings after waking up,
stiffness, anxiety, and depression using a VAS discussed above.
Exploratory factor analysis on FIQ reveals that all the sub-items of the first item load
on a single factor with factor loadings ranging from 0.5 to 0.95. No other items of the scale
load significantly on this factor. This suggests that all the sub-items in first item are additive to
form a composite sub-score. The physical functioning subscale had a test-retest reliability of
0.95 and other FIQ items had test-retest reliability of 0.56. The low test-retest reliability of
fibromyalgia syndromes (pain, morning tiredness, stiffness, and fatigue) could suggest the
temporal nature of these symptoms, highlighting the need for a more repeated assessment like
EMA. Burckhardt et al. (1991) have also checked for convergent construct validity for FIQ
with an existing Arthritis Impact Measurement Scale (AIMS) (Meenan, Mason, Anderson,
Guccione, & Kazis, 1992), where FIQ items where moderately to highly correlated with AIMS
items. However, the number of tender points, a common diagnostic approach for fibromyalgia
showed a poor correlation with FIQ items except physical functioning (r = 0.61) and missing
work (r = 0.74). Nevertheless, FIQ is used in several studies as a measure of outcomes in
fibromyalgia trials (Bennett, 2005). In fact, Bennett (2005) in their review have highlighted
that FIQ has been effective in distinguishing fibromyalgia with other regional pain syndromes
(i.e. good discriminant validity).
Pain Disability Index. The Pain Disability Index (PDI) measures the disabilities caused due to
chronic pain in our everyday lives (Pollard, 1984). PDI is a 7-item scale with NRS for each
item. Items measure disabilities caused by pain in household activities, recreational activities,
social activities, occupation-related tasks, sexual behaviour, self-care, and life support
activities. Each item is measured on a 10-point NRS with 0 - no disability and 10 - total
disability. No disability refers to complete ability to perform the tasks normally. The composite
score ranges from 0 to 70.
In Tait, Pollard, Margolis, Duckro, and Krause (1987)’s study, they have established a
strong internal consistency (with Cronbach’s alpha = 0.87, p < 0.0001), nevertheless, PDI
showed weaker test-retest reliability (r = 0.44, p < 0.001) over a two months period (Tait,
Chibnall, & Krause, 1990), suggesting that PDI can be a poor assessment of stability of pain-
related disabilities. On performing principle component analysis, two factors have been
identified namely voluntary/discretionary activities including household activities, recreation,
social activities, occupation-related tasks, and sexual behaviour and obligatory
activities/functions including self-care and life support activities (Tait et al., 1987).
Voluntary/discretionary accounts for 59.3% of the variance and obligatory activities share
14.3% of the variance in total pain disability. In their study, they have found significant
differences in pain disabilities of outpatient and inpatients groups, with inpatients group
reporting higher pain disabilities. In fact, outpatients reported significantly lower disabilities
in voluntary as well as obligatory activities.
Similarly, Tait et al. (1990) have found that people with high PDI score (i.e. higher pain
related disabilities) have reported higher pain (using McGill Pain Questionnaire) and high
distress in everyday life. However, no significant differences were found in pain-related history
(e.g., pain surgery and pain duration) between high and low PDI patients. In their study, Tait
et al. (1990) have found that patients with high PDI scores have exhibited higher rates of pain
behaviour, in particular, verbal complaints, grimaces, mobility issues, body language, and
stationary behaviour. Finally, these studies also suggest that PDI is not as suitable for
intermittent pains such as migraine headache as compared to chronic lower back pain.
Brief Pain Inventory. Brief Pain Inventory (BPI) measures the severity and effect of pain in
cancer patients for a recall period of 1 week (Cleeland & Ryan, 1994). The long form of BPI
is a 29-item inventory comprising of patient pain-history questions, regional pain questionnaire
(pain drawings), pain intensity questions, pain medication, and pain interference in everyday
life. Pain intensity is measured using 4-NRSs mearing pain intensity at its least, worst, average,
and momentary pain. Three open-ended questions in this inventory capture measures that make
the pain feeling better, worse, and medications that patient is receiving. Like McGill Pain
Questionnaire, this inventory provides a list of 15 adjectives and patients are asked to check if
that adjective represents their pain. Finally, 7 items capture pain’s interference with general
activity, mood, walking ability, household work, social relationships, sleep, and enjoyment in
life using an NRS with 0 as “does not interfere” and 1 as “completely interferes”.
Tan, Jensen, Thornby, and Shanti (2004) in their study have identified a strong internal
consistency (Cronbach’s alpha = 0.85 for intensity subscale and 0.88 for interference subscale)
of BPI with chronic non-malignant pain. Similarly, Keller et al. (2004) and Zelman, Gore,
Dukes, Tai, and Brandenburg (2005) in their studies have identified high internal consistency
(Cronbach’s alpha = 0.94) of BPI with cancer related chronic pain and neuropathic pain
respectively. Test-retest reliabilities in cancer patients for pain intensity items was 0.98 (p <
0.05) and for pain interference was 0.97 (p <0.05) for a retest duration between 30 – 60 minutes
(Radbruch et al., 1999). In fact, 1 day to 1 week test-retest reliability pain intensity ratings
range between 0.93 and 0.59, suggesting BPI is more stable when measured between shorter
intervals (Daut, Cleeland, & Flanery, 1983).
Factor analysis in Tan et al. (2004) verifies its two factor structure of pain intensity and
pain interference. This study established a significant correlation of pain interference (r = 0.57,
p <0.01) and pain intensity (r = 0.40, p < 0.01) of BPI with Roland-Morris Disability
Questionnaire (RMDQ) (Roland & Fairbank, 2000). Similarly, Zelman et al. (2005) have also
identified two factors in principle factor analysis (eigenvalues > 1) consistent with previous
studies. Pain intensity of BPI was highly correlated (r = 0.63, p < 0.001) with Bodily Pain
Short Form composite scores (McHorney, Ware, Lu, & Sherbourne, 1994) and VRS of pain
intensity (r = 0.74, p < 0.001).
WOMAC Osteoarthritis Index. Western Ontario and McMaster Universities (WOMAC)
Osteoarthritis Index measures the clinical symptoms of osteoarthritis and their severity
(Bellamy, 1989). WOMAC-OI is a 24-item scale measuring three constructs pain intensity in
different activities, feelings of stiffness, and difficulty levels of different physical functions
(like pain disability). Pain intensity is measured on a 5-point scale (0 = None, 1 = light, 2 =
Moderate, 3 = Very, 4 = Extremely) walking, climbing stairs, nocturnal activities, resting, and
weight bearing. Stiffness is measured on a 5-point scale for morning and later day pain. Finally,
17 everyday pain related disabilities in physical functions are captured on a 5-point scale.
When measured for symptoms at two time-points between 6-months, high test-retest
reliability was identified by Bellamy (1989). Moreover, high internal consistency (Cronbach’s
alpha > 0.70) and high responsiveness (p < 0.001) have been identified when measured for
surgery, “Isoxicam”, and “Piroxicam” trials. However, WOMAC-OI also identified significant
variability in perceived pain at different time points in a day.
Electronic Diaries and Ecological Momentary Assessments
Studies in the past have shown that chronic pain is a fleeting, frequently occurring, and
varying (non-constant) experience throughout the day (Morren, van Dulmen, Ouwerkerk, &
Bensing, 2009). The previous methods of pain assessment described so far are administered in
one setting. As a result, they impose high recall burden introducing bias in responses. In fact,
this response bias can also result in overestimation of the pain (e.g., (Sorbi et al., 2006a).
Therefore, it is important to capture episodes of chronic pain repeatedly in natural settings and
in the moment, so that participants are asked to log their pain experience at or around the
moment when the pain occurs. This allows researchers to get as close as possible to the true
pain scores as against recalled scores. Electronic diary studies (Bolger, Davis, & Rafaeli, 2003)
and ecological momentary assessment (EMA; (A.A. Stone & Shiffman, 1994) or experience
sampling method (Csikszentmihalyi & Larson, 1987) are two of the most widely used methods
that allow such in situ data collection to measure behaviours, contexts, and states.
Diary data collection methods and EMA are implemented on paper/pen as well as
electronic devices. In recent times, especially post-2003, mobile-device based EMA and diary
methods have been gaining traction to measure momentary changes in behaviour. The devices
used in pain EMA/diary studies include (but not limited to) PDAs (Heiberg et al., 2007),
feature-mobile phones (for SMS-based assessment, e.g., (Alfven, 2010)), smartphones(e.g.,
(Garcia-Palacios et al., 2014)), and smartwatches/watch-type computers (e.g., (Kikuchi et al.,
2006)). In addition to reliable prompting, these devices also offer accurate time-stamping of
participant responses, thereby reducing back-dated or forward-dated entries. Since repeated
assessments also allow for measurement of more than one variables, we are limiting the scope
of this review only to the studies that have chronic pain as the primary variable. We excluded
studies where pain was a secondary variable in addition to other primary variables of interest.
In electronic diary studies or EMA, a participant is prompted one or more times a day
on the electronic device with a set of multiple choice questions related to the research construct
of interest (e.g., chronic pain). Often these prompts are either signal contingent (based on an
alarm on the device) or event contingent (based on a specific event, e.g., chronic back pain).
Diary studies and EMA can allow studying of chronic pain and related states in a longitudinal
(e.g., multiweek) study. As a result, these methodologies not only allow summarization of pain-
related variables over a period, but also capture the within-subject changes in the construct. In
most EMA/Diary studies, the electronic devices are loaned to the study participants.
Nevertheless, with increasing popularity of smartphones researchers can deploy their studies
directly on participants’ personal devices.
End-of-the-day Diaries
A common repeated in situ assessment is end of the day dairy, where participants are
required to narrate their pain experience once a day (typically end of the day). For instance,
Robert N. Jamison et al. (2001) have conducted a 1-year study with 36 chronic lower back pain
patients, among which 20 patients used a palmtop computer as well as paper diaries. The
purpose of this study was to compare compliance, validity, and reliability of momentary pain
assessment of electronic diaries compared to paper diaries. Patients reported hourly pain
intensities using VAS, however, it was optimized for the palmtop computer to the length of
only 5.7 cm. The agreement between paper diary scores and electronic diary scores were tested
using a measurement error model: Paper score = intercept + (score X electronic diary score).
This is to ensure that paper diary scores reflect the direct measure of pain intensity. A
correlation of r = 0.88 (p<0.0001) was found between weekly telephonic surveys and average
weekly electronic diary assessment. Finally, electronic diaries had a compliance rate of 89.9%.
This study presents an early evidence that repeated assessments on electronic diaries measure
the same behaviour as traditional paper diaries.
Ecological Momentary Assessments (EMA)
Researchers often use electronic diary studies interchangeably with ecological
momentary assessment. However, for this review, we intend to distinguish these two
methodologies. As against end-of-the-day diaries, ecological momentary assessment (also
known as Real Time Data Collection) refers to data collection where participants are prompted
multiple times a day with a set of multiple choice questions with an intention to gather data
with high temporal density. Compared to end-of-the-day diary studies, EMA questions are
asked amid everyday activities, thereby reducing recall bias (A. A. Stone and Broderick (2007).
In particular, electronic diaries just like recall questionnaires often suffer from “peak-end-
effects” (Redelmeier & Kahneman, 1996), which bias responses towards the most intense
painful episode occurring in the recall period. Similarly, they also suffer from “duration-
neglect”, which leads respondents to ignore episodes without pain when recalling their pain
(Fredrickson & Kahneman, 1993). EMA can reduce these biases by asking respondents on their
immediate chronic pain experiences as against recall over a period. Finally, the self-report
questionnaires discussed before (especially McGill Pain Questionnaire, Regional Pain Scale,
and Brief Pain Inventory) require a clinical staff or a healthcare professional to administer the
test with the patients. Due to high costs of human resources, this limits researchers’ abilities to
capture pain experiences for longitudinal studies in natural settings such as precision medicine
initiative (National Institutes of Health, 2015). It also limits the patients’ ability to voluntarily
report their pain experience.
In fact in the early 90s, Lewis, Lewis, and Cumming (1995) have introduced PIPER
(Prompting intensity of pain, Electronic Records) where participants respond to momentary
pain intensity on a 7-point NRS. Here, 0 represents no pain and 6 represents extreme pain. In
three follow-up studies, they establish the reliability and validity of measuring momentary pain
using electronic diaries. For a period of two months, participants responded to NRS prompts
4-times a day. In their first study, 40 chronic pain patients were asked to respond to PIPER
pain intensity scale (PPS) along with a standard VAS. PPS and VAS scales had construct
validity of r = 0.91 at the original administration and r = 0.93 for repeated assessment
suggesting a strong test-retest reliability. Their second study, as part of their first experiment
involved a 4-controlled movement of painful knee within 30-minute period. The pain ratings
recorded using PPS and VAS correlated at r = 0.93 (p < 0.001), providing support for
convergent validity. Similarly, Sprott and Miiller (1998) have studied symptoms of
Fibromyalgia using electronic diaries for a period of 18 days. Using the electronic diaries, they
have observed a decrease in pain intensity over 18 days of treatment period.
However, EMA administration is not like traditional surveys administered at one point
in time. Therefore, it is important to check for other biases introduced due to repeated
assessments such as study burden, low study compliance, and reactivity. For instance, Arthur
A. Stone et al. (2003) have conducted a 2-weeks long chronic pain assessment study with 91
chronic pain patients to gauge the effect of frequent momentary assessments on reactivity,
compliance, and patient satisfaction. They manipulated the temporal density of prompting
participants with surveys to measure its effect on study compliance. This study used digital
diaries (PDAs) to ask, “Are you in pain right now?” with yes/no response options. If the patient
responds yes, a follow-up question asking, “How much pain are you in right now?” that
measures pain-intensity on 100mm VAS is prompted. In a between-subject design, participants
were divided into EMA and no-EMA groups. Within the EMA group, participants were sub-
divided into 3 prompts/day, 6 prompts/day, and 12 prompts/day sub-groups. On examining
(using between- and within-subject multi-level models) the average proportion of times
patients felt pain across the three EMA groups, they found no main effect of the group, no
linear effect of time, and no interaction between group and time. However, there was a
significant interaction of group and linear component of time when comparing VAS ratings for
episodes when patients felt the pain. The 12 prompts/day prompting strategy had a flat slope
over time, 6-prompts/day had a positive temporal trend (i.e., increasing pain), and 3-
prompts/day had a negative temporal trend (i.e., decreasing pain). While it may suggest that
there is an indication of reactivity considering the group*time interaction, it does not indicate
any trend (e.g., increasing prompt density increases reactivity). Likewise, there was no effect
of momentary pain assessment on the recall of pain over the same period when tested for both
between- and within-subject effects. In addition, no significant effect of prompting density on
pain recall was observed using the same analysis. Moreover, there were no significant
differences in EMA groups (based on prompt density) on study compliance (~94%). However,
it is hard to explain if the compliance was also influenced by financial incentives ($100) offered
to patients, which may not be available for large scale studies. Finally, the authors observed
that participants with 12-prompts/day perceived the study to be more burdensome and more
interfering than other EMA groups.
Likewise, it is imperative to examine if repeated assessments in the moment give any
different information from the traditional weekly recall of pain. It is crucial to check if high
temporal density of pain related information can give more information on the within-subject
changes than summarized between-subject scores. For instance, Arthur A. Stone et al. (2004)
have compared the weekly recall of pain with the weekly average obtained from momentary
pain assessment over a 2-week period. In this study, 68 patients diagnosed with chronic pain
were asked to report their pain experiences for 4-weeks. However, only in the last two weeks
they were required to report momentary pain every day in addition to weekly recall of pain.
For EMA, participants reported chronic pain experience using a palmtop computer. Similar to
Arthur A. Stone et al. (2003), participants were first asked if they felt the pain (“Are you in
pain right now?” – Yes/No). If they responded yes, they were asked a follow-up question on
pain intensity using VAS. If the patients responded no to the first question, it was also recorded
as 0 pain intensity. In addition, 17 questions measure patient’s location, current activity, mood,
affect, and sensory characteristics of the pain. Since this was a follow up study to Arthur A.
Stone et al. (2003), participants received prompt 3 times a day, 6 times a day, and 12 times a
day. For weekly pain recall, authors used “Weekly Pain Questionnaire” administered once a
week. Authors found that average momentary levels of pain (44/100) were lower than levels
of weekly recall pain (58/100) and the differences for both the weeks were significant
(p<0.0001). Higher average of weekly recall pain could be due to “duration neglect” or “peak-
end-effects”. Accounting for these biases in the analysis, only those episodes were included in
the comparison where patients reported “Yes” to the first question in EMA. Even though the
difference in weekly average was still significant, the difference was very small with average
momentary assessment (53.0/100) and weekly recall (53.8/100). This suggests that there is a
close correspondence of perceived pain intensities for episodes when patients reported as to
feeling pain (neglecting the episodes of no pain).
Since EMA captures chronic pain intensity in the moment, it can offer a more detailed
information on within-subject variability of pain. Using Pearson’s correlation on the averages
of momentary pan and weekly recall pain may ignore the effects of the within-subject
variability. Therefore, in this work, authors also explore “intra-class” correlations (in addition
to inter-class correlations) to account for the variability and differences in average pain
intensity between momentary assessments and weekly recall. The intra-class correlations are
measured based on the change in weekly recall pain and change in the average momentary
pain. The inter-class correlation was found to be above 0.75 (p<0.001) for both the weeks. This
only suggested that patients who reported higher pain intensity in weekly recall also reported
higher pain intensity in momentary assessments. However, the intra-class correlations were
less than 0.4 (even though significant; p<0.001), suggesting that the change in pain measured
by momentary assessment was weakly correlated with weekly recall pain. Finally, patients also
reported their perceived/judged change in pain using a five point-categorical scale. Using Chi-
squared test, it was found that momentary pain had a stronger association with the
judged/perceived change in pain than weekly recall of pain. However, one-way ANOVA
presented significant differences between judged change in pain and computed averages from
momentary assessment and weekly recall. Since, judged pain was measured using categorical
variables, it is hard to assess if patients who judged the pain change in certain way reported it
in the same way using EMA.
EMA’s ability to gather information repeatedly from natural settings also allows it to
collect information on a range of behaviours in a single study. This allows researchers to study
within-subject associations between these variables and build more robust prediction models
of behaviour. In case of chronic pain, EMA can be used to study different pain related responses
such as mood, affect, anxiety, and locomotion. For instance, in a two-part study, Sorbi et al.
(2006a) have gathered information on pain intensity, psychological pain responses, pain
disabilities. Psychological pain responses include fear avoidance (avoiding pain causing
behaviours due to the fear of pain), measured using a 7-item 7-point scale, cognitive responses
(beliefs over chronic pain and its effects) measured using a 7-item 7-point scale, and spousal
responses (partner’s reactions towards pain behaviour) measured using 6-item 7-point scale.
Likewise, pain disabilities measure physical capacity, mental capacity, pain interference, and
immobility. Participants were prompted for 4-weeks and 4-times per day.
When capturing these variables using EMA, there are two kinds of changes in these
variables – within subject and between-subject changes. Therefore, it is important to use multi-
level modelling techniques that account for both between-subject (level-I) and within-subject
(level-II) variables. Within subject variables are often time varying variables and between
subject variables are often time-invariant variables. In part-I of their study, they found that
psychological pain responses such as fear avoidance and cognitive responses explain 28% of
total variance in pain intensity. Spousal responses explained only 9% of the variance. However,
this study also found that momentary differences in pain intensity within patients explained
16% of the pain intensity variance. In part-II of their study (Sorbi et al., 2006b), they found
that pain intensity explained 8 to 19% of the pain disability variance. However, psychological
pain responses only explain 7 to 16% of the variance in pain disability. Nevertheless, spousal
responses predicted immobility due to pain better than other disabilities. These studies
demonstrate how different pain-related variables gathered in EMA could help us understand
their associations and design robust prediction models that account for between as well as
within subject changes in variables.
Administering EMA requires technology that can reliably prompt or remind
participants with surveys and offer usable input methods to record these responses. Different
technology platforms have been used to assess chronic pain using EMA in natural settings.
Stinson et al. (2006) have developed a palmtop computer application called “e-Ouch” to
capture pain intensity and location. Pain intensity is measured using VAS and pain location is
captured using a pain drawing. Authors have found that VAS scale on a touch device was too
sensitive to respond, suggesting a need for a more controllable slider to reduce touch
sensitivity. They also found that users felt it easier to respond to a prompt asking about the
momentary pain as against end-of-the-day pain recall task. Moreover, e-Ouch takes 9 minutes
to complete at each prompt. However, in their study, participants were asked to rate the pain
intensity before identifying the pain location. It is unclear if this unconventional sequence of
questions introduces any bias in the responses.
Similarly, Lalloo et al. (2014) proposed Pain-QUILT – a web-application that measures
pain location, pain intensity, and pain interference with everyday life (akin to short form
McGill Pain Questionnaire). In Pain-QUILT, participants are first asked to locate the site of
pain on body, followed by pain intensity, and overall pain interference. On comparing pain-
QUILT with paper based McGill Pain Questionnaire and Brief Pain Inventory, authors found
that pain-QUILT is significantly easier to use taking only 5 minutes to complete. Moreover,
pain-QUILT responses were significantly correlated with McGill Pain Questionnaire and Brief
Pain Inventory (r = 0.70, p < 0.01). However, pain-QUILT is a web-based application that
depends on user’s proactivity to log their pain experiences.
Garcia-Palacios et al. (2014) have assessed chronic pain in fibromyalgia using a
smartphone. Here participants respond to EMA prompts three times a day measuring pain
intensity using an NRS. Instead of a linear rating scale, their application displays all the
numbers of NRS in a grid to comfortably fit the smartphone screen. In addition, it also asks for
the current mood using an FPS. When compared with paper pen diary in a within-subject
experiment of 2-weeks duration (1-week on EMA and 1-week on paper-pen based diary), the
aggregated pain intensities were strongly correlated r = 0.79, p < 0.001. Authors found that
the aggregated EMA pain intensity values were lower than that of recall methods, suggesting
strong “peak-effects” in chronic pain reporting, despite having a strong correlation (r = 0.59,
p < 0.001). Finally, smartphone-based EMA was found to be easier to use for chronic pain
reporting and more useful compared to paper-pen based EMA.
Even though smartphones are pervasive in the present times, EMA can be administered
on feature phones using SMS service. For instance, Alfven (2010) proposed “SMS pain diary”,
a real time data capture for chronic pain in children. Pain was reported in terms of pain intensity
(using NRS), pain duration (in minutes), and pain related disabilities. Children were asked to
respond to messages six times a day for 1 week. The tool was evaluated for test-retest reliability
and construct validity. Similarly, measuring systematic disagreement and random individual
disagreement (relative change in position and random individual changes), authors found very
insignificant disagreements. This corroborates strong test-retest reliability using SMS as a dairy
method for chronic pain. Finally, the construct validity measured using monotonic agreement
(concordance = 0.77), suggests a strong construct validity of the pain measures. However, SMS
requires high involvement text-input from the users. For the instances of painful experiences,
such a high involvement text typing can introduce additional burden. Authors in this paper
have not addressed this bias, and this is a research opportunity for the future.
Recent advances in smartwatches make them a viable option for EMA data collection.
Methods such as micro-EMA (S. Intille, Haynes, Maniar, Ponnada, & Manjourides, 2016),
which use short microinteractions to gather data at high temporal densities may be useful in
the assessment of chronic pain. Kikuchi et al. (2006) have used a watch-type computer, which
is a wrist worn device to capture pain intensity. Even though their study measured tension type
headache intensity, their approach could be replicable for chronic pain assessments. In this
study, participants are asked to respond to watch type prompts for 7-days. Prompting include
signal contingent recordings (using an alarm on the watch), participant initiated recordings
(even contingent), and a peak headache recording. Headache intensity was measured using a
VAS adjusted to watch screen with an increment of multiples of 5 instead of unit increments
in traditional VAS. In fact, as discussed above, device touch sensitivity could hurt participant
responses, therefore, a discrete increment could be easier to use in case of watch-type display
(and a physical joystick). However, such changes to VAS should be compared to NRS as the
fundamental properties of both are the same. The mean compliance using watch-type computer
was 96% for signal contingent recordings. Intra-individual standard variability for pain
intensity varied between 6.26 to 35.49, suggesting how EMA can capture within-subject
changes in pain intensity. Consistent with previous study, recall headache intensity at the end
of the week was higher than the aggregate EMA headache intensity (i.e. peak end effects). In
fact, there was no significant difference between the aggregated extreme headache intensities
and recalled headache intensity, supporting the presence of peak effects in recall based pain
assessments. Inter- and intra-class correlations were calculated between headache intensities
recorded using watch-type computers and weekly recall. Inter-class correlation was r = 0.70
(p < 0.001), suggesting that at an aggregate level EMA and weekly recall measure in the same
direction. However, low intra-class correlations suggest that the level differences in NRS in
weekly recall and watch-type computer could be measuring constructs differently at within-
subject level. However, authors have not explored if the reactivity to survey prompts was also
influenced by the novelty of device, especially the fact that devices were loaned. Effect of
novelty on response rates is still in its nascent stages with conflicting results. For instance,
Walsh and Brinker (2016) have found an increase in compliance when loaning a mobile device
to EMA participants. On the other hand, Ponnada, Haynes, Maniar, Manjourides, and Intille
(2017) have identified no difference in the response rates of smartphone-based EMA and
smartwatch-based EMA for the same behaviour measurement task.
Chronic pain is a subjective experience making its assessment a challenging task for
clinicians and health care practitioners. In this paper, I reviewed different chronic pain
assessment methodologies including qualitative and observational methods,
psychophysiological sensing, self-report questionnaire, and in situ ecological momentary
assessments (EMA). For each of the methods discussed in the paper, I reviewed their
applications, administrative challenges, and psychometric properties including composite
scores, internal reliability, and validity. The choice of an assessment method depends on 1)
purpose of the measurement (e.g., capturing pain disability vs treatment effect), 2) Population
or patient groups (e.g., young adults vs older adults), and 3) scale or length of the assessment
(e.g., one-time assessment vs multiweek monitoring).
For instance, interview based qualitative methods are often used in clinical settings in the
presence of a clinician or a health care professional. While this method is helpful in gathering
the first-hand experiences of chronic pain patients, it has several limitations. First, to conduct
these interviews, a clinical staff is always required. These staff members are not always
available at large scale, restricting qualitative studies to small sample sizes. Second, qualitative
chronic pain assessments do not have statistical properties restricting their use in building
computational models of chronic pain experiences. Nevertheless, qualitative assessments are
best used when the researcher or clinical staff has no prior understanding of patients’
symptoms. In fact, to design new quantitative assessment tools, qualitative assessment can help
researchers identify the key constructs and behaviours that inform their decisions.
There are no direct or independent psychophysiological sensing methods available for
chronic pain assessment. However, chronic pain experiences have systematic effects on heart
rate and galvanic skin response. Prior work shows that during the chronic pain experience, the
heart rate and galvanic skin response tend to increase and heart rate variability tend to decrease.
However, these physiological parameters can also vary during other behavioural responses
such as anxiety, positive mood, and exertion during the physical activity. Even though these
measures are not available for chronic pain independently, little research has been done to use
them in tandem with the self-report. Multiple sources of information often help in triangulating
the measurement. Therefore, there is a scope of using self-report methods to improve the
recognition using passive sensing methods or enable context-sensitive prompting.
This paper also covered a unique assessment called “pain drawings” that ask participants
to identify the region on the body where they feel the pain along with the verbal descriptors of
that pain (e.g., burning, stabbing or pinching pain). Pain drawings are best suited to easily
localize the pain experience and identify the region of the body. In fact, many multidimensional
surveys such as McGill Pain questionnaire use pain drawings as a subscale to capture the
regional location of chronic pain. However, so far pain drawings have been administered only
in the presence of clinical staff with an exception of PainQUILT, where an interactive pain
drawing is used to capture pain location (Lalloo et al., 2014). In fact, there is little evidence
that pain drawings can capture the psychological properties of chronic pain. This restricts their
use as the only chronic pain assessment tool and are often used alongside other self-report
methods that assess other factors associated with pain e.g., mood.
A large section of this review was dedicated to self-report methods. This paper reviews
unidimensional surveys (with single items), multidimensional surveys (measuring multiple
constructs with multiple items), and condition/disability specific pain scales. The
unidimensional surveys have been mainly used to capture pain intensity. All the pan intensity
scales have high test-retest reliability (r>0.70) and strong convergent validity (r>0.70).
However, different unidimensional scales are suitable for different populations based on the
complexity of responding to the survey. For instance, FPS has been found to be more sensitive
and suitable to older adults, individuals with cognitive impairments, and children. However,
just like VRS, FPS is argued for not having ratio properties. It is often difficult to mark no pain
in FPS. Similarly, the most popular pain intensity scales are NRS and VAS, which are shown
to be easy to mark pain intensity. In fact, research suggests that NRS and VAS have a very
strong correlation (r > 0.85) suggesting that these scales can be used interchangeably. Even
though researchers suggest that VAS must be used with a 100mm scale, several researchers
using VAS in ecological momentary assessment (EMA) have modified it (e.g., 50 mm VAS)
to fit the screen sizes. However, often the modifications in the properties of VAS make it
behave like NRS, with fixed points to move the slider on. Research studies modifying or
shortening VAS do not delve deep into the possible changes in their psychometric properties.
Multidimensional pain surveys not only capture pain intensity and location, but also capture
the disabilities associated with the pain and enable participants to provide a detailed account
of their experience through representative keywords. McGill Pain Questionnaire (MPQ)
measures pain experience using 7 categories of verbal descriptors or keywords. Similarly,
regional pain scale captures pain location and intensity without using a pain drawing. However,
regional pain scale is limited to specific regions of the body. As a result, it doesn’t allow
participants to freely highlight the regions where they are feeling the pain. As a result, MPQ
has been most widely used chronic pain assessment tool used by clinicians and researchers.
One limitation of the MPQ and related scales is that they do not capture pain’s interference in
our everyday life. To capture chronic pain’s effect on the overall quality of life, Pain Disability
Questionnaire and disease specific scales such as Fibromyalgia Impact Questionnaire are used.
This suggests that there is no comprehensive and yet short chronic pain assessment
questionnaire that can capture pain location, intensity, experience, and pain related disabilities.
This is a future opportunity for researchers to design a comprehensive, short, and easily
administrable scale capturing all aspects of a chronic pain episode.
Finally, chronic pain is a dynamic experience with changing intensity at different time
points. Therefore, researchers have used EMA and end-of-the-day diary studies to measure
within-subject changes in pain episodes. VAS has been most commonly used to capture pain
intensity in EMA using mobile phones, PDAs, and watch-type computers. Researchers have
found that EMA measurements can overcome the limitations of recall scales caused due to
“peak effects” and “duration neglect”. Even though EMA-based chronic pain assessment had
strong correlation with the recall based scales, EMA can capture more variability in chronic
pain assessment. This suggests that chronic pain can be reliably captured using self-report on
mobile/wearable devices for long periods of time.
Responding to the EMA prompts can often be very burdensome to the participants. In the
prior studies measuring chronic pain intensity, participants have been prompted several times
a day. However, most studies have formally measured the burden of participating in the study,
with an exception of Arthur A Stone, Joan E Broderick, Saul S Shiffman, and Joseph E
Schwartz (2004)’s work. While perceived burden is a generic property of an EMA study, this
burden could also be affected due to the domain of measurement. Chronic pain patients have
an added disability induced due to the pain. This disability can hinder the participation in an
EMA study, thereby making EMA assessments more burdensome. Moreover, it is also possible
that EMA burden can alter participants’ responses to the prompts. Therefore, it is essential to
systematically study EMA perceived burden to measure chronic pain.
Another important aspect of EMA is reactivity. Though there have been some studies
measuring reactivity of EMA capturing chronic pain, there is dearth of research modelling
EMA reactivity and self-report responses. So far, researchers have not found any evidence of
increasing interruption being related to increasing reactivity in pain assessment. If chronic pain
is a subjective experience, it is possible that repeated assessments could trigger patients to think
more about pain resulting in skewed responses.
One way to reduce EMA burden is to make each EMA prompt easy to respond taking very
little time. For instance, S. Intille et al. (2016) have proposed micro-EMA where all the EMA
prompts are reduced to single questions with yes/no answers to measure behaviors with high
temporal density. This approach reduces response burden drastically and significantly
improves compliance. However, such a quick EMA response requires less cognitive
complexity of the EMA questions. Therefore, it is essential for the researchers to study the
cognitive complexity of different pain intensity scales when used in EMA. For instance, it is
possible that for a quick response the NRS and VAS could be more cognitively taxing for the
participants to process the answer options. Unlike the evidence shown on paper-pen tests,
VRSs and FPS could be cognitively less demanding for a quick EMA response. A formal
comparison of perceived cognitive complexity of different pain intensity scales for EMA is a
worthwhile endeavour.
One limitation of the chronic pain assessment methods explored in this work is that they
have not measured the individual subjective assessment of pain intensity. For instance, when
using NRS for pain intensity, a person may rate an excruciating pain as 8, based on the prior
experience with the condition This within subjects’ difference is only captured in repeated
assessment studies. However, even in repeated assessments, the subjective feeling of pain can
drastically influence participants’ responses. For instance, if a participant was feeling
excruciating pain and has reported 10 on an NRS. That participant can rate as low as 3 on NRS
the moment the there is a minute reduction in the feeling of pain. A slight decrease in the
intensity from an excruciating pain can be highly rewarding to participant. This phenomenon
is hard to capture and therefore, requires a formal investigation.
One way to address this limitation could be to make assessments relative to each other. For
example, using micro-EMA, participants can be asked if their current feeling or intensity of
chronic pain is more or is less than their previous experience. One way to achieve that could
be to ask participants if “the current is more intense than before” with options “Yes/No/Same”.
Another way to capture this difference is to allow participants to view their previous response,
and adjust that response to register their current feelings of pain. This may allow participants
to accurately report their changing pain intensities relative to their own baseline. However,
such a measure would require novel statistical approaches to model chronic pain experiences.
As discussed above, a combination of passive sensing from heart rate and galvanic skin
response sensors could be used to trigger EMA prompts for chronic pain assessments. These
context-sensitive EMA prompts (Stephen S Intille, 2007) can further reduce recall biases of
self-report and enable participants to report their chronic pain episodes more accurately. Such
context-sensitive chronic pain assessments can improve automatic recognition of chronic pain
episodes using passive sensors. Likewise, there is a scope of capturing chronic pain episodes
using more low cost methods such as “day reconstruction method”. Finally, chronic pain
definitions are getting updated every year and most measurements are now expected to capture
a broader set of pain-related constructs. In fact, measuring pain-intensity alone may not be
sufficient for a holistic pain management interventions (Ballantyne & Sullivan, 2015) and
assessment methods should include social, psychological, physical, and biological constructs
of chronic pain.
I am sincerely thankful to Prof. Rob Volpe for valuable discussions and his recommendations
on evaluating psychometric properties. I am thankful to Prof. Stephen Intille for his valuable
comments and suggestions on manuscript writing.
Alfven, G. (2010). SMS pain diary: a method for real-time data capture of recurrent pain in
childhood. Acta Paediatr, 99(7), 1047-1053. doi:10.1111/j.1651-2227.2010.01735.x
Analytics, G. I. (2015). Pain management: A global strategic business report. Global Industries
Analytics, Inc. Retrieved from
Appelhans, B. M., & Luecken, L. J. (2008). Heart rate variability and pain: associations of two
interrelated homeostatic processes. Biological psychology, 77(2), 174-182.
Ballantyne, J. C., & Sullivan, M. D. (2015). Intensity of Chronic Pain--The Wrong Metric? N
Engl J Med, 373(22), 2098-2099. doi:10.1056/NEJMp1507136
Bellamy, N. (1989). Pain assessment in osteoarthritis: experience with the WOMAC
osteoarthritis index. Semin Arthritis Rheum, 18(4 Suppl 2), 14-17.
Bellamy, N., Buchanan, W. W., Goldsmith, C. H., Campbell, J., & Stitt, L. W. (1988).
Validation study of WOMAC: a health status instrument for measuring clinically
important patient relevant outcomes to antirheumatic drug therapy in patients with
osteoarthritis of the hip or knee. J Rheumatol, 15(12), 1833-1840.
Bennett, R. (2005). The Fibromyalgia Impact Questionnaire (FIQ): a review of its
development, current version, operating characteristics and uses. Clin Exp Rheumatol,
23(5 Suppl 39), S154-162.
Bieri, D., Reeve, R. A., Champion, G. D., Addicoat, L., & Ziegler, J. B. (1990). The Faces Pain
Scale for the self-assessment of the severity of pain experienced by children:
development, initial validation, and preliminary investigation for ratio scale properties.
Pain, 41(2), 139-150.
Bolger, N., Davis, A., & Rafaeli, E. (2003). Diary methods: Capturing life as it is lived. Annual
Review of Psychology, 54, 579-616.
Bolton, J. E., & Wilkinson, R. C. (1998). Responsiveness of pain scales: a comparison of three
pain intensity measures in chiropractic patients. J Manipulative Physiol Ther, 21(1), 1-
Bonica, J. J. (1953). The management of pain. Philadelphia: Lea & Febiger.
Bonnet, A., & Naveteur, J. (2004). Electrodermal activity in low back pain patients with and
without co-morbid depression. International Journal of Psychophysiology, 53(1), 37-
Bradley, L. A. (1984). Pain and behavioral medicine: A cognitive-behavioral perspective.
General Hospital Psychiatry, 6(3), 235-236. doi:10.1016/0163-8343(84)90047-1
Briggs, M., & Closs, J. S. (1999). A Descriptive Study of the Use of Visual Analogue Scales
and Verbal Rating Scales for the Assessment of Postoperative Pain in Orthopedic
Patients. Journal of Pain and Symptom Management, 18(6), 438-446.
Brunelli, C., Zecca, E., Martini, C., Campa, T., Fagnoni, E., Bagnasco, M., . . . Caraceni, A.
(2010). Comparison of numerical and verbal rating scales to measure pain
exacerbations in patients with chronic cancer pain. Health Qual Life Outcomes, 8, 42.
Burckhardt, C. S., Clark, S. R., & Bennett, R. M. (1991). The fibromyalgia impact
questionnaire: development and validation. J Rheumatol, 18(5), 728-733.
Carnes, D., Ashby, D., & Underwood, M. (2006). A Systematic Review of Pain Drawing
Literature: Should Pain Drawings Be Used for Psychologic Screening? Clin J Pain,
22(5), 449-457. doi:10.1097/
Cleeland, C. S., & Ryan, K. (1994). The Brief Pain Inventory. Available onli ne: https://www/.
mdanderson. org/education-andresearch/departments-programs-and-
tools/BPI_UserGuide. pdf (accessed on 1 March 2016).
Collins, G. A., Cohen, M. J., Naliboff, B. D., & Schandler, S. L. (1982). Comparative analysis
of paraspinal and frontalis EMG, heart rate and skin conductance in chronic low back
pain patients and normals to various postures and stress. Scandinavian Journal of
Rehabilitation Medicine, 14(1), 39-46.
Creamer, P., & Hochberg, M. C. (1997). Why does osteoarthritis of the knee hurt--sometimes?
Rheumatology, 36(7), 726-728. doi:10.1093/rheumatology/36.7.726
Csikszentmihalyi, M., & Larson, R. (1987). Validity and reliability of the Experience-
Sampling Method. The Journal of Nervous and Mental Disease, 175(9), 526-536.
Dansie, E. J., & Turk, D. C. (2013). Assessment of patients with chronic pain. Br J Anaesth,
111(1), 19-25. doi:10.1093/bja/aet124
Daut, R. L., Cleeland, C. S., & Flanery, R. C. (1983). Development of the Wisconsin Brief
Pain Questionnaire to assess pain in cancer and other diseases. Pain, 17(2), 197-210.
Dowell, D., Haegerich, T. M., & Chou, R. (2016). Cdc guideline for prescribing opioids for
chronic pain—united states, 2016. JAMA, 315(15), 1624-1645.
Dray, A., Urban, L., & Dickenson, A. (1994). Pharmacology of chronic pain. Trends in
Pharmacological Sciences, 15(6), 190-197. doi:
Dudgeon, D., Raubertas, R. F., & Rosenthal, S. N. (1993). The short-form McGill Pain
Questionnaire in chronic cancer pain. J Pain Symptom Manage, 8(4), 191-195.
Dureja, G. P., Jain, P. N., Shetty, N., Mandal, S. P., Prabhoo, R., Joshi, M., . . . Phansalkar, A.
A. (2014). Prevalence of chronic pain, impact on daily life, and treatment practices in
India. Pain Pract, 14(2), E51-62. doi:10.1111/papr.12132
Ekblom, A., & Hansson, P. (1988). Pain intensity measurements in patients with acute pain
receiving afferent stimulation. J Neurol Neurosurg Psychiatry, 51(4), 481-486.
Ferraz, M. B., Quaresma, M. R., Aquino, L. R., Atra, E., Tugwell, P., & Goldsmith, C. H.
(1990). Reliability of pain scales in the assessment of literate and illiterate patients with
rheumatoid arthritis. J Rheumatol, 17(8), 1022-1024.
Ferreira-Valente, M. A., Pais-Ribeiro, J. L., & Jensen, M. P. (2011). Validity of four pain
intensity rating scales. PAIN®, 152(10), 2399-2404.
Follick, M. J., Ahern, D. K., & Laser-Wolston, N. (1984). Evaluation of a daily activity diary
for chronic pain patients. Pain, 19(4), 373-382. doi:
Fredrickson, B. L., & Kahneman, D. (1993). Duration neglect in retrospective evaluations of
affective episodes. J Pers Soc Psychol, 65(1), 45-55.
Galer, B. S., & Jensen, M. P. (1997). Development and preliminary validation of a pain
measure specific to neuropathic pain: the Neuropathic Pain Scale. Neurology, 48(2),
Garcia-Palacios, A., Herrero, R., Belmonte, M. A., Castilla, D., Guixeres, J., Molinari, G., &
Banos, R. M. (2014). Ecological momentary assessment for chronic pain in
fibromyalgia using a smartphone: a randomized crossover study. Eur J Pain, 18(6),
Ginzburg, B. M., Merskey, H., & Lau, C. L. (1988). The relationship between pain drawings
and the psychological state. Pain, 35(2), 141-146. doi:
Goldberg, D. S., & McGee, S. J. (2011). Pain as a global public health priority. BMC public
health, 11, 770-770. doi:10.1186/1471-2458-11-770
Gramling, S. E., & Elliott, T. R. (1992). Efficient pain assessment in clinical settings. Behav
Res Ther, 30(1), 71-73.
Grotle, M., Brox, J. I., & Vøllestad, N. K. (2004). Concurrent Comparison of Responsiveness
in Pain and Functional Status Measurements Used for Patients With Low Back Pain.
Spine, 29(21), E492-E501.
Grunnesjo, M., Bogefeldt, J., Blomberg, S., Delaney, H., & Svardsudd, K. (2006). The course
of pain drawings during a 10-week treatment period in patients with acute and sub-
acute low back pain. BMC Musculoskelet Disord, 7, 65. doi:10.1186/1471-2474-7-65
Hallman, D. M., Ekman, A. H., & Lyskov, E. (2014). Changes in physical activity and heart
rate variability in chronic neck-shoulder pain: monitoring during work and leisure time.
Int Arch Occup Environ Health, 87(7), 735-744. doi:10.1007/s00420-013-0917-2
Häuser, W., Wolfe, F., Henningsen, P., Schmutzer, G., Brähler, E., & Hinz, A. (2014). Untying
chronic pain: prevalence and societal burden of chronic pain stages in the general
population - a cross-sectional survey. BMC public health, 14(1), 352.
Hayashi, K., Arai, Y.-C. P., Morimoto, A., Aono, S., Yoshimoto, T., Nishihara, M., . . . Ushida,
T. (2015). Associations Between Pain drawing and Psychological Characteristics of
Different Body Region Pains. Pain Practice, 15(4), 300-307. doi:10.1111/papr.12173
Heiberg, T., Kvien, T. K., Dale, O., Mowinckel, P., Aanerud, G. J., Songe-Moller, A. B., . . .
Hagen, K. B. (2007). Daily health status registration (patient diary) in patients with
rheumatoid arthritis: a comparison between personal digital assistant and paper-pencil
format. Arthritis Rheum, 57(3), 454-460. doi:10.1002/art.22613
Herr, K., Spratt, K. F., Garand, L., & Li, L. (2007). Evaluation of the Iowa pain thermometer
and other selected pain intensity scales in younger and older adult cohorts using
controlled clinical pain: a preliminary study. Pain Med, 8(7), 585-600.
Herr, K. A., Mobily, P. R., Kohout, F. J., & Wagenaar, D. (1998). Evaluation of the Faces Pain
Scale for use with the elderly. Clin J Pain, 14(1), 29-38.
Hjermstad, M. J., Fayers, P. M., Haugen, D. F., Caraceni, A., Hanks, G. W., Loge, J. H., . . .
Kaasa, S. (2011). Studies Comparing Numerical Rating Scales, Verbal Rating Scales,
and Visual Analogue Scales for Assessment of Pain Intensity in Adults: A Systematic
Literature Review. Journal of Pain and Symptom Management, 41(6), 1073-1093.
Intille, S., Haynes, C., Maniar, D., Ponnada, A., & Manjourides, J. (2016). µEMA:
Microinteraction-based ecological momentary assessment (EMA) using a smartwatch
Proceedings of the 2016 ACM International Joint Conference on Pervasive and
Ubiquitous Computing (pp. 1124-1128): ACM.
Intille, S. S. (2003). Ubiquitous computing technology for just-in-time motivation of behavior
change Proceedings of the UbiHealth Workshop.
Intille, S. S. (2007). Technological innovations enabling automatic, context-sensitive
ecological momentary assessment. In A. Stone, S. Shiffman, A. Atienza, & L. Nebeling
(Eds.), The Science of Real-Time Data Capture: Self-Report in Health Research (pp.
308-337). New York, NY: Oxford University Press.
Intille, S. S., & Ho, J. (2004). A just-in-time persuasive interface for healthy aging. Retrieved
from Cambridge, MA:
Jamison, R. N., Gracely, R. H., Raymond, S. A., Levine, J. G., Marino, B., Herrmann, T. J., . .
. Katz, N. P. (2002). Comparative study of electronic vs. paper VAS ratings: a
randomized, crossover trial using healthy volunteers. Pain, 99(1–2), 341-347.
Jamison, R. N., Raymond, S. A., Levine, J. G., Slawsby, E. A., Nedeljkovic, S. S., & Katz, N.
P. (2001). Electronic diaries for monitoring chronic pain: 1-year validation study. Pain,
91(3), 277-285.
Jamison, R. N., Raymond, S. A., Levine, J. G., Slawsby, E. A., Nedeljkovic, S. S., & Katz, N.
P. (2001). Electronic diaries for monitoring chronic pain: 1-year validation study. Pain,
91(3), 277-285. doi:
Jensen, M. P., Karoly, P., & Braver, S. (1986). The measurement of clinical pain intensity: a
comparison of six methods. Pain, 27(1), 117-126. doi:
Keller, S., Bann, C. M., Dodd, S. L., Schein, J., Mendoza, T. R., & Cleeland, C. S. (2004).
Validity of the Brief Pain Inventory for Use in Documenting the Outcomes of Patients
With Noncancer Pain. Clin J Pain, 20(5), 309-318.
Khedr, E. M., Kotb, H., Kamel, N. F., Ahmed, M. A., Sadek, R., & Rothwell, J. C. (2005).
Longlasting antalgic effects of daily sessions of repetitive transcranial magnetic
stimulation in central and peripheral neuropathic pain. Journal of Neurology,
Neurosurgery & Psychiatry, 76(6), 833-838. doi:10.1136/jnnp.2004.055806
Kikuchi, H., Yoshiuchi, K., Miyasaka, N., Ohashi, K., Yamamoto, Y., Kumano, H., . . .
Akabayashi, A. (2006). Reliability of recalled self-report on headache intensity:
Investigation using ecological momentary assessment technique. Cephalalgia, 26(11),
1335-1343. doi:10.1111/j.1468-2982.2006.01221.x
Kim, E. J., & Buschmann, M. T. (2006). Reliability and validity of the Faces Pain Scale with
older adults. Int J Nurs Stud, 43(4), 447-456. doi:10.1016/j.ijnurstu.2006.01.001
Lalloo, C., Kumbhare, D., Stinson, J. N., & Henry, J. L. (2014). Pain-QuILT: clinical feasibility
of a web-based visual pain assessment tool in adults with chronic pain. J Med Internet
Res, 16(5), e127. doi:10.2196/jmir.3292
Lewis, B., Lewis, D., & Cumming, G. (1995). Frequent measurement of chronic pain: an
electronic diary and empirical findings. Pain, 60(3), 341-347.
Margolis, R. B., Chibnall, J. T., & Tait, R. C. (1988). Test-retest reliability of the pain drawing
instrument. Pain, 33(1), 49-51. doi:
Margolis, R. B., Tait, R. C., & Krause, S. J. (1986). A rating system for use with patient pain
drawings. Pain, 24(1), 57-65. doi:
Martínez-Lavín, M., Hermosillo, A. G., Rosas, M., & Soto, M.-E. (1998). Circadian studies of
autonomic nervous balance in patients with fibromyalgia: A heart rate variability
analysis. Arthritis & Rheumatism, 41(11), 1966-1971. doi:10.1002/1529-
McHorney, C. A., Ware, J. E., Jr., Lu, J. F., & Sherbourne, C. D. (1994). The MOS 36-item
Short-Form Health Survey (SF-36): III. Tests of data quality, scaling assumptions, and
reliability across diverse patient groups. Med Care, 32(1), 40-66.
Meenan, R. F., Mason, J. H., Anderson, J. J., Guccione, A. A., & Kazis, L. E. (1992). AIMS2.
The content and properties of a revised and expanded Arthritis Impact Measurement
Scales Health Status Questionnaire. Arthritis Rheum, 35(1), 1-10.
Melzack, R. (1975). The McGill Pain Questionnaire: major properties and scoring methods.
Pain, 1(3), 277-299.
Merskey, H. (1986). Classification of chronic pain: Descriptions of chronic pain syndromes
and definitions of pain terms. Pain, Suppl 3, 226.
Morren, M., van Dulmen, S., Ouwerkerk, J., & Bensing, J. (2009). Compliance with
momentary pain measurement using electronic diaries: a systematic review. Eur J Pain,
13(4), 354-365. doi:10.1016/j.ejpain.2008.05.010
Nakamura, M., Nishiwaki, Y., Ushida, T., & Toyama, Y. (2011). Prevalence and characteristics
of chronic musculoskeletal pain in Japan. J Orthop Sci, 16(4), 424-432.
National Institutes of Health. (2015). Precision Medicine Iniative Cohort Program. Retrieved
Nilsen, T. I. L., Holtermann, A., & Mork, P. J. (2011). Physical Exercise, Body Mass Index,
and Risk of Chronic Pain in the Low Back and Neck/Shoulders: Longitudinal Data
From the Nord-Trøndelag Health Study. American Journal of Epidemiology, 174(3),
267-273. doi:10.1093/aje/kwr087
Ohnhaus, E. E., & Adler, R. (1975). Methodological problems in the measurement of pain: A
comparison between the verbal rating scale and the visual analogue scale. Pain, 1(4),
379-384. doi:
Pagé, M. G., Katz, J., Stinson, J., Isaac, L., Martin-Pichora, A. L., & Campbell, F. (2012).
Validation of the Numerical Rating Scale for Pain Intensity and Unpleasantness in
Pediatric Acute Postoperative Pain: Sensitivity to Change Over Time. The Journal of
Pain, 13(4), 359-369. doi:
Pain, C. (2016). Chronic Pain Management. Retrieved from
Palmer, H. (1949). Pain charts; a description of a technique whereby functional pain may be
diagnosed from organic pain. N Z Med J, 48(264), 187-213.
Perttunen, K., Tasmuth, T., & Kalso, E. (1999). Chronic pain after thoracic surgery: a follow-
up study. Acta Anaesthesiologica Scandinavica, 43(5), 563-567. doi:10.1034/j.1399-
Peters, M. L., & Schmidt, A. J. (1991). Psychophysiological responses to repeated acute pain
stimulation in chronic low back pain patients. Journal of Psychosomatic Research,
35(1), 59-74.
Phillips, J., Ogden, J., & Copland, C. (2015). Using drawings of pain-related images to
understand the experience of chronic pain: A qualitative study. British Journal of
Occupational Therapy, 78(7), 404-411. doi:10.1177/0308022614562791
Poleshuck, E. L., Katz, J., Andrus, C. H., Hogan, L. A., Jung, B. F., Kulick, D. I., & Dworkin,
R. H. (2006). Risk Factors for Chronic Pain Following Breast Cancer Surgery: A
Prospective Study. The Journal of Pain, 7(9), 626-634.
Pollard, C. A. (1984). Preliminary validity study of the pain disability index. Percept Mot
Skills, 59(3), 974. doi:10.2466/pms.1984.59.3.974
Ponnada, A., Haynes, C., Maniar, D., Manjourides, J., & Intille, S. (2017). Microinteraction
ecological momentary assessment response rates: Effect of microinteractions or the
smartwatch? Proc. of the ACM Journal on Interactive, Mobile, Wearable, and
Ubiquitous Technology, 1(3).
Price, D. D., Bush, F. M., Long, S., & Harkins, S. W. (1994). A comparison of pain
measurement characteristics of mechanical visual analogue and simple numerical rating
scales. Pain, 56(2), 217-226.
Radbruch, L., Loick, G., Kiencke, P., Lindena, G., Sabatowski, R., Grond, S., . . . Cleeland, C.
S. (1999). Validation of the German version of the Brief Pain Inventory. J Pain
Symptom Manage, 18(3), 180-187.
Redelmeier, D. A., & Kahneman, D. (1996). Patients' memories of painful medical treatments:
real-time and retrospective evaluations of two minimally invasive procedures. Pain,
66(1), 3-8.
Richards, J. S., Nepomuceno, C., Riles, M., & Suer, Z. (1982). Assessing pain behavior: the
UAB Pain Behavior Scale. Pain, 14(4), 393-398.
Roland, M., & Fairbank, J. (2000). The Roland–Morris Disability Questionnaire and the
Oswestry Disability Questionnaire. Spine, 25(24), 3115-3124.
Salaffi, F., Stancati, A., Silvestri, C. A., Ciapetti, A., & Grassi, W. (2004). Minimal clinically
important changes in chronic musculoskeletal pain intensity measured on a numerical
rating scale. Eur J Pain, 8(4), 283-291. doi:10.1016/j.ejpain.2003.09.004
Smith, M. T., & Haythornthwaite, J. A. (2004). How do sleep disturbance and chronic pain
inter-relate? Insights from the longitudinal and cognitive-behavioral clinical trials
literature. Sleep Medicine Reviews, 8(2), 119-132.
Sorbi, M. J., Peters, M. L., Kruise, D. A., Maas, C. J., Kerssens, J. J., Verhaak, P. F., & Bensing,
J. M. (2006a). Electronic momentary assessment in chronic pain I: psychological pain
responses as predictors of pain intensity. Clin J Pain, 22(1), 55-66.
Sorbi, M. J., Peters, M. L., Kruise, D. A., Maas, C. J., Kerssens, J. J., Verhaak, P. F., & Bensing,
J. M. (2006b). Electronic momentary assessment in chronic pain II: pain and
psychological pain responses as predictors of pain disability. Clin J Pain, 22(1), 67-81.
Sprott, H., & Miiller, W. (1998). Functional Symptoms in Fibromyalgia. Clinical Bulletin of
Myofascial Therapy, 3(1), 61-67. doi:10.1300/J425v03n01_07
Steglitz, J., Buscemi, J., & Ferguson, M. J. (2012). The future of pain research, education, and
treatment: a summary of the IOM report "Relieving pain in America: a blueprint for
transforming prevention, care, education, and research". Transl Behav Med, 2(1), 6-8.
Stewart, W. F., Ricci, J. A., Chee, E., Morganstein, D., & Lipton, R. (2003). Lost productive
time and cost due to common pain conditions in the us workforce. JAMA, 290(18),
2443-2454. doi:10.1001/jama.290.18.2443
Stinson, J. N., Petroz, G. C., Tait, G., Feldman, B. M., Streiner, D., McGrath, P. J., & Stevens,
B. J. (2006). e-Ouch: Usability Testing of an Electronic Chronic Pain Diary for
Adolescents With Arthritis. Clin J Pain, 22(3), 295-305.
Stone, A. A., & Broderick, J. E. (2007). Real-time data collection for pain: appraisal and current
status. Pain Med, 8 Suppl 3, S85-93. doi:10.1111/j.1526-4637.2007.00372.x
Stone, A. A., Broderick, J. E., Schwartz, J. E., Shiffman, S., Litcher-Kelly, L., & Calvanese,
P. (2003). Intensive momentary reporting of pain with an electronic diary: reactivity,
compliance, and patient satisfaction. Pain, 104(1–2), 343-351.
Stone, A. A., Broderick, J. E., Shiffman, S. S., & Schwartz, J. E. (2004). Understanding recall
of weekly pain from a momentary assessment perspective: absolute agreement,
between- and within-person consistency, and judged change in weekly pain. Pain,
107(1–2), 61-69. doi:
Stone, A. A., Broderick, J. E., Shiffman, S. S., & Schwartz, J. E. (2004). Understanding recall
of weekly pain from a momentary assessment perspective: absolute agreement,
between- and within-person consistency, and judged change in weekly pain. Pain,
107(1-2), 61-69. doi:10.1016/j.pain.2003.09.020
Stone, A. A., & Shiffman, S. (1994). Ecological momentary assessment: Measuring real world
processes in behavioral medicine. Annals of Behavioral Medicine, 16, 199-202.
Tait, R. C., Chibnall, J. T., & Krause, S. (1990). The Pain Disability Index: psychometric
properties. Pain, 40(2), 171-182.
Tait, R. C., Pollard, C. A., Margolis, R. B., Duckro, P. N., & Krause, S. J. (1987). The Pain
Disability Index: psychometric and validity data. Arch Phys Med Rehabil, 68(7), 438-
Tan, G., Jensen, M. P., Thornby, J. I., & Shanti, B. F. (2004). Validation of the Brief Pain
Inventory for chronic nonmalignant pain. J Pain, 5(2), 133-137.
Tousignant-Laflamme, Y., Rainville, P., & Marchand, S. (2005). Establishing a Link Between
Heart Rate and Pain in Healthy Subjects: A Gender Effect. The Journal of Pain, 6(6),
341-347. doi:10.1016/j.jpain.2005.01.351
Tracey, I., & Bushnell, M. C. (2009). How Neuroimaging Studies Have Challenged Us to
Rethink: Is Chronic Pain a Disease? The Journal of Pain, 10(11), 1113-1120.
Treede, R.-D., Rief, W., Barke, A., Aziz, Q., Bennett, M. I., Benoliel, R., . . . Wang, S.-J.
(2015). A classification of chronic pain for ICD-11. Pain, 156(6), 1003-1007.
Turk, D. C., & Okifuji, A. (2001). Pain terms and taxonomies of pain. Bonica’s management
of pain (3 ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Uden, A., Astrom, M., & Bergenudd, H. (1988). Pain drawings in chronic back pain. Spine
(Phila Pa 1976), 13(4), 389-392.
Von Baeyer, C. L., Bergstrom, K. J., Brodwin, M. G., & Brodwin, S. K. (1983). Invalid use of
pain drawings in psychological screening of back pain patients. Pain, 16(1), 103-107.
Walsh, E. I., & Brinker, J. K. (2016). Should participants be given a mobile phone, or use their
own? Effects of novelty vs utility. Telematics and Informatics, 33(1), 25-33.
Ware, L. J., Epps, C. D., Herr, K., & Packard, A. (2006). Evaluation of the Revised Faces Pain
Scale, Verbal Descriptor Scale, Numeric Rating Scale, and Iowa Pain Thermometer in
older minority adults. Pain Manag Nurs, 7(3), 117-125.
Ware, L. J., Herr, K. A., Booker, S. S., Dotson, K., Key, J., Poindexter, N., . . . Packard, A.
(2015). Psychometric Evaluation of the Revised Iowa Pain Thermometer (IPT-R) in a
Sample of Diverse Cognitively Intact and Impaired Older Adults: A Pilot Study. Pain
Manag Nurs, 16(4), 475-482. doi:10.1016/j.pmn.2014.09.004
Williams, A. C. d. C., & Craig, K. D. (2016). Updating the definition of pain. Pain, 157(11),
2420-2423. doi:10.1097/j.pain.0000000000000613
Williamson, A., & Hoggart, B. (2005). Pain: a review of three commonly used pain rating
scales. Journal of Clinical Nursing, 14(7), 798-804. doi:10.1111/j.1365-
Wolfe, F. (2003). Pain extent and diagnosis: development and validation of the regional pain
scale in 12,799 patients with rheumatic disease. J Rheumatol, 30(2), 369-378.
Woolf, C. J., & Mannion, R. J. (1999). Neuropathic pain: aetiology, symptoms, mechanisms,
and management. The Lancet, 353(9168), 1959-1964.
Zelman, D. C., Gore, M., Dukes, E., Tai, K. S., & Brandenburg, N. (2005). Validation of a
modified version of the brief pain inventory for painful diabetic peripheral neuropathy.
J Pain Symptom Manage, 29(4), 401-410. doi:10.1016/j.jpainsymman.2004.06.018"
Table 1.
Classification of chronic pain (Treede et al., 2015)
Pain classification
Chronic primary pain
Chronic cancer pain
Chronic post-traumatic pain
Chronic neuropathic pain
Chronic headache/orofacial
Chronic visceral pain
Chronic musculoskeletal pain
Table 2.
Self-report pain assessment questionnaires
Pain Assessment Scale
Construct Measured
Unidimensional Scales
Verbal Rating Scale
Pain Intensity
Ohnhaus and Adler
Numeric Rating Scale
Pain Intensity
Price et al. (1994)
Visual Analogue Scale
Pain Intensity
Ohnhaus and Adler
Faces Pain Scale
Pain Intensity
Bieri et al. (1990)
Iowa Pain Thermometer
Pain Intensity
K. Herr et al. (2007)
Multidimensional Scales
McGill Pain Questionnaire
Pain intensity, severity,
subjective pain description and
pain location
Melzack (1975)
Regional Pain Scale
Pain intensity and pain location
Wolfe (2003)
Neuropathic Pain Scale
Neuropathic pain intensity
Galer and Jensen (1997)
Fibromyalgia Impact Questionnaire
Pain disabilities associated
with Fibromyalgia
Burckhardt et al. (1991)
Pain disability, mobility, and
quality of life
Bellamy et al. (1988)
Pain Disability Index
Pain related disabilities
Pollard (1984)
Brief Pain Inventory
Pain intensity and effect in
cancer patients
Cleeland and Ryan
Figure 1. Example of pain drawings, borrowed from Grunnesjo et. al. (2006)
Figure 2. Unidimensional surveys VRS (top-left), VAS (top-right), NRS (bottom-left), and
FPS (bottom-right); (Right) IPT
ResearchGate has not been able to resolve any citations for this publication.
Conference Paper
Full-text available
Ecological Momentary Assessment (EMA) is a method of in situ data collection for assessment of behaviors, states, and contexts. Questions are prompted during everyday life using an individual's mobile device, thereby reducing recall bias and increasing validity over other self-report methods such as retrospective recall. We describe a microinteraction-based EMA method ("micro" EMA, or μEMA) using smartwatches, where all EMA questions can be answered with a quick glance and a tap -- nearly as quickly as checking the time on a watch. A between-subjects, 4-week pilot study was conducted where μEMA on a smartwatch (n=19) was compared with EMA on a phone (n=14). Despite an =8 times increase in the number of interruptions, μEMA had a significantly higher compliance rate, completion rate, and first prompt response rate, and μEMA was perceived as less distracting. The temporal density of data collection possible with μEMA could prove useful in ubiquitous computing studies.
Chronic pain self-management has long since been accepted as a necessary component for the successful management of chronic pain. Whilst recognition of its necessity continues, a standard definition which provides certainty as to what chronic pain selfmanagement actually is has yet to be developed. Clear definitions for a phenomenon such as chronic pain self-management are imperative to clinical practice. They aide professional-patient-carer communication, and lead to the standardisation of appropriate clinical and service outcome measures, allowing us to assess the effectiveness of interventions. In this chapter we aim to 1) illustrate the value of defining a phenomenon such as chronic pain self-management; 2) discuss the consequences of failing to develop a standard definition, including the role of Health Professionals and; 3) explore the additional benefits of developing a standard definition, and their link to clinical guidelines. We conclude that until an agreed definition for chronic pain self-management is reached, the development of professional guidelines for the delivery of services supporting chronic pain self-management will be hindered, and health professionals will remain unclear as to their contribution. This lack of clarity negatively impacts upon clinical practice, attitudes towards practice, outcome measurement, service improvement, and consequently, patient outcomes.
Mobile-based ecological-momentary-assessment (EMA) is an in-situ measurement methodology where an electronic device prompts a person to answer questions of research interest. EMA has a key limitation: interruption burden. Microinteraction-EMA(µEMA) may reduce burden without sacrificing high temporal density of measurement. In µEMA, all EMA prompts can be answered with ‘at a glance' microinteractions. In a prior 4-week pilot study comparing standard EMA delivered on a phone (phone-EMA) vs. µEMA delivered on a smartwatch (watch-µEMA), watch-µEMA demonstrated higher response rates and lower perceived burden than phone-EMA, even when the watch-µEMA interruption rate was 8 times more than phone-EMA. A new 4-week dataset was gathered on smartwatch-based EMA (i.e., watch-EMA with 6 back-to-back, multiple-choice questions on a watch) to compare whether the high response rates of watch-µEMA previously observed were a result of using microinteractions, or due to the novelty and accessibility of the smartwatch. No statistically significant differences in compliance, completion, and first-prompt response rates were observed between phone-EMA and watch-EMA. However, watch-µEMA response rates were significantly higher than watch-EMA. This pilot suggests that (1) the high compliance and low burden previously observed in watch-µEMA is likely due to the microinteraction question technique, not simply the use of the watch versus the phone, and that (2) compliance with traditional EMA (with long surveys) may not improve simply by moving survey delivery from the phone to a smartwatch.
Importance Primary care clinicians find managing chronic pain challenging. Evidence of long-term efficacy of opioids for chronic pain is limited. Opioid use is associated with serious risks, including opioid use disorder and overdose.Objective To provide recommendations about opioid prescribing for primary care clinicians treating adult patients with chronic pain outside of active cancer treatment, palliative care, and end-of-life care.Process The Centers for Disease Control and Prevention (CDC) updated a 2014 systematic review on effectiveness and risks of opioids and conducted a supplemental review on benefits and harms, values and preferences, and costs. CDC used the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) framework to assess evidence type and determine the recommendation category.Evidence Synthesis Evidence consisted of observational studies or randomized clinical trials with notable limitations, characterized as low quality using GRADE methodology. Meta-analysis was not attempted due to the limited number of studies, variability in study designs and clinical heterogeneity, and methodological shortcomings of studies. No study evaluated long-term (≥1 year) benefit of opioids for chronic pain. Opioids were associated with increased risks, including opioid use disorder, overdose, and death, with dose-dependent effects.Recommendations There are 12 recommendations. Of primary importance, nonopioid therapy is preferred for treatment of chronic pain. Opioids should be used only when benefits for pain and function are expected to outweigh risks. Before starting opioids, clinicians should establish treatment goals with patients and consider how opioids will be discontinued if benefits do not outweigh risks. When opioids are used, clinicians should prescribe the lowest effective dosage, carefully reassess benefits and risks when considering increasing dosage to 50 morphine milligram equivalents or more per day, and avoid concurrent opioids and benzodiazepines whenever possible. Clinicians should evaluate benefits and harms of continued opioid therapy with patients every 3 months or more frequently and review prescription drug monitoring program data, when available, for high-risk combinations or dosages. For patients with opioid use disorder, clinicians should offer or arrange evidence-based treatment, such as medication-assisted treatment with buprenorphine or methadone.Conclusions and Relevance The guideline is intended to improve communication about benefits and risks of opioids for chronic pain, improve safety and effectiveness of pain treatment, and reduce risks associated with long-term opioid therapy.
Borrowing treatment principles from acute and end-of-life pain care, particularly a focus on pain intensity, has had harmful consequences for patients with chronic pain. Multimodal therapy, by contrast, aims to reduce pain-related distress, disability, and suffering.
The measurement of subjective pain intensity continues to be important to both researchers and clinicians. Although several scales are currently used to assess the intensity construct, it remains unclear which of these provides the most precise, replicable, and predictively valid measure. Five criteria for judging intensity scales have been considered in previous research: ease of administration of scoring; relative rates of incorrect responding; sensitivity as defined by the number of available response categories; sensitivity as defined by statistical power; and the magnitude of the relationship between each scale and a linear combination of pain intensity indices. In order to judge commonly used pain intensity measures, 75 chronic pain patients were asked to rate 4 kinds of pain (present, least, most, and average) using 6 scales. The utility and validity of the scales was judged using the criteria listed above. The results indicate that, for the present sample, the scales yield similar results in terms of the number of subjects who respond correctly to them and their predictive validity. However, when considering the remaining 3 criteria, the 101-point numerical rating scale appears to be the most practical index.