Pain: Assessment and Measurement

Richard H Gracely. Encyclopedia of Perception. Editor: E Bruce Goldstein. 2010. Sage Publication.

Pain in the clinic cannot be physically observed and usually cannot be independently precisely controlled. Such control is possible in laboratory studies in which pain is induced by controlled stimulation. This entry discusses how pain is assessed and measured in the laboratory as well as in the clinic.

Measurement of Pain in the Laboratory

Experimental heat, mechanical, electrical, chemical, and other forms of stimulation are extremely useful in studies of pain in both man and animals. Pain is inferred in animals from a range of physiological measurements and from reactive and operant behaviors. Those uncomfortable with human verbal reports of pain have promoted the same type of measurements for pain in humans using physiological measurements (such as spinal reflexes, cortical evoked potentials, or functional neuroimaging) and behavioral indices ranging from observed measures of grimacing and bracing to measures of social and work activities. However, pain is only truly defined in humans, based on the unique ability to describe and evaluate private experience.

Laboratory methods provide a degree of control over the subjective measurement process that is not possible with clinical pain measurement. These methods demonstrate that human subjects can reliably rate pain, distinguish between dimensions of sensory intensity and unpleasantness, and discriminate different pain qualities. They also provide valuable information about rating scale behavior that can be applied to clinical pain assessment. Many traditional psychophysical procedures, which focus on relations between stimulus-evoked sensations, have been applied to the evaluation of pain sensation. These methods can be divided into those that assess the pain threshold and those that evaluate the suprathreshold range from pain threshold to pain tolerance.

The term pain threshold is often used to describe general pain sensitivity, such as “he has a high pain threshold.” The specific meaning of pain threshold refers to the distinction between sensory qualities of nonpainful and painful stimulation. As the magnitude of a potentially painful stimulus is increased, the pain threshold marks the transition from the absence of pain sensation to the presence of pain sensation, and is quantified as the amount of stimulus intensity needed to evoke a painful sensation. In the laboratory, the pain threshold is determined by established psychophysical procedures that minimize biases that influence the result. These include the classical method of limits and the method of constant stimuli. The method of limits uses alternating ascending and descending series of continuous or discrete stimuli. These alternating series control for directional errors such as anticipation, of indicating the presence of pain before the sensation becomes painful, and errors of habituation, of continuing to use the non-painful responses after the stimulus has become painful. To further control biases, both the starting point and amount of change in successive stimuli can be randomly varied. For assessment of the pain threshold, this method is usually modified to use only ascending series to prevent unacceptably painful stimulation at the beginning of a descending series. The method of constant stimuli presents discrete stimuli of varying intensities that span the range from certain judgments of nonpain to certain judgments of pain. Thus, these stimulus values span the “gray area” of uncertain responding. The result of this method is usually plotted as a probability function extending for 0 to 100% probability, and a stimulus value corresponding to a specific value (such as 50 or 75%) is defined as the pain threshold.

Threshold methods are limited in that they only assess sensitivity to weak pain sensations at the very bottom of the perceptual pain range. This sensitivity may not reflect sensitivity throughout the range from pain threshold tolerance, which can be assessed by methods that evaluate subjective judgments of suprathreshold pain sensations.

Suprathreshold methods deliver sensations over the entire perceptual range from threshold to tolerance. Most of these deliver a series of discrete stimuli that vary randomly in intensity sequence. The dependent variable is the response made using a subjective rating scale. The difference in the varying stimulus intensities should be small enough to create confusion between adjacent stimulus intensities, forcing the subject to make ratings based on judgments of sensation and not on identification of a specific stimulus (e.g., that is the middle stimulus that I call “5”). These procedures vary in the type of response scale and the type of analysis. The response scales range from the ubiquitous numerical 1 to 10 scale and simple category scales (mild, moderate, severe) to continuous measures such as the visual analog scale (VAS) that involves marking a line that represents a range of pain sensation. The most common example of a VAS is a 10-centimeter (cm) horizontal line labeled at the extremes by “no pain” and “most intense pain imaginable” or some other descriptor of intense pain. Another approach combines these methods by spacing quantified descriptors along a continuous scale or numerical category scale. These scales have traditionally been presented on paper but modern versions are delivered by computer or handheld PDA (personal digital assistant) devices. All of these methods have been shown to be reliable and valid by repeated use in pain studies. One of the issues in the selection of a method, especially with patients in pain, is the selection of the appropriate stimulus values. In addition, the use of a fixed stimulus set to evaluate analgesics provides an undesirable cue because all of the stimuli will be less intense. These issues have been addressed by another class of suprathreshold scaling methods that use interactive computer algorithms to continuously adjust varying stimulus intensities to produce the same level of subjective sensations in all subjects. These methods can be considered to be a second-generation approach because of their interactive nature. These methods automatically adjust the stimulus levels to each participant, avoiding problems of selecting appropriate stimulus levels for each person. They also provide the same subjective levels of pain before and after an analgesic intervention, minimizing extraneous clues that an active drug has been administered.

Further issues involved in selection of a supra-threshold scaling method include applicability to multisite international studies (a problem with verbal scales), ability to provide absolute judgments of pain magnitude (greater face validity with verbal scales), and the need for a motor response (line marking methods are problematic for telephone evaluations and for studies such as brain imaging in which a motor response is difficult or undesirable). Responses that are easily remembered, such as simple category scales, may not be the best choice for measures of repeat reliability, because it is not clear if the subject remembers the pain or the previous response. Another issue concerns the ability to rate significant pain dimensions.

To assess the relevant dimensions of pain, measurement systems must go beyond the assessment of sensation. Pain is not only a sensation but a negative feeling state that motivates behavior. This is a problem also for the assessment of other sensations with homeostatic consequences, such as temperature, taste, and olfaction that contain both negative and positive feeling states. Measurement methods used for these homeostatic sensations have been adapted to the assessment of pain, distinguishing between the distinct dimensions of pain intensity and pain unpleasantness, and have shown how these dimensions differ among different types of pain and respond differently to pain control interventions. For example, there is evidence that verbal scales with words descriptive of a single pain dimension (e.g., mild, moderate, intense for the dimension of pain intensity and uncomfortable, annoying, distressing for pain unpleasantness) facilitate the discrimination of that dimension. The types of studies demonstrating these properties have shown that an intravenous opioid reduces pain intensity with variable effects on unpleasantness and that the amount of unpleasantness associated with a specific level of sensory intensity is less in experimental situations and more in chronic pain conditions, which show further differences between conditions such as fibromyalgia and cancer.

Pain psychophysical methods are often criticized because the laboratory environment does not duplicate the sensory and emotional aspects of the clinical situation. While true to some extent, these procedures have been extremely useful for increasing the knowledge about the neural mechanisms that mediate pain and analgesia. The application of these methods to patients, termed quantitative sensory testing (QST), essentially extends the neurological examination to discover abnormalities in simple pain perception and in processing of prolonged pain. QST has greatly increased the capabilities of clinical pain assessment, leading to both improved diagnosis and treatment and to improved knowledge of poorly understood pain mechanisms. QST involves the same methods as used in the laboratory, with the complexity of these methods matched to the alteration of symptoms. Gross changes can be detected by simple procedures, such as raising the temperature of a probe slowly until a report of pain is made. More subtle symptoms may require the more elaborate procedures previously described. The modalities commonly used are thermal (both hot and cold), pressure (both blunt and punctuate), and vibration. Other methods include nonpainful and painful electrical stimulation and methods such as repeated heat stimuli that activate spinal summation mechanisms. These methods differentially activate classes of primary afferent fibers that mediate touch (large diameter A-beta fibers activated selectively by vibration, pressure, and electrical stimulation), pricking pain (thinly myelinated A-delta fibers activated by punctuate, heat, cold, and electrical stimulation), and diffuse, often burning pain (unmyelinated fibers activated by heat, blunt pressure, and electrical stimulation).

Measurement of Pain in the Clinic

Despite the complexity of pain, most clinical measures still treat it as a single dimension and use simple scales, such as a number from 1 to 10 or marking a VAS. These measures are used to assess pain in the clinic and often as the primary endpoint in extensive clinical trials of analgesic agents. Unlike other physical signs or symptoms, these measures are based on a person’s evaluation and report of a private experience. Individuals may be instructed to evaluate the pain experience “right now” or asked to integrate the average pain over the last day, week, or other interval. Home pain diaries are used to capture fluctuations in pain and provide a more objective integration of pain over time. Concern for compliance has led to the use of electronic diaries that record the time of every entry and that provide a reminder function. On the other end of the objective-subjective continuum, measures of global perception of change in clinical trials are receiving attention because the judgments include subjectivity about overall quality of life.

The previous measures of pain magnitude are useful for the goals of testing analgesics and monitoring treatment efficacy. Other measures are useful for diagnosis and choice of treatment. One of the most popular is the McGill pain questionnaire (MPQ), which uses multiple verbal categories to assess pain intensity and unpleasantness, but also to evaluate the many varied qualities of pain sensation. These qualities include four types of pressure (compression, traction, lacerating, punctuate, described by words such as squeezing, pulling, cutting, and drilling), thermal qualities of heat, warm, cool, cold, and freezing, temporal qualities such as flickering or pounding, locations such as superficial or deep, and spatial qualities such as spreading or radiating. The profile for an individual is useful for diagnosis and for the selection of available treatments. This information can be supplemented by simple physical bedside QST measures or more extensive QST procedures if needed.

Clinical pain measurement is concerned with pain, the person, and pain impact. Psychosocial variables, such as depression and anxiety, accompany pain and need to be evaluated for treatment planning and treatment efficacy. Ability to follow a behavioral regime is an important factor, as are cognitive styles such as catastrophizing. This style refers to the tendency to view all possibilities in the darkest terms. Functional brain imaging studies have found brain activity related to depression in the insular cortex and amygdala, whereas catastrophizing is associated with multiple regions related to the anticipation of pain (contralateral medial frontal gyrus, ipsilateral cerebellum) and attention to pain (anterior cingulate cortex [ACC], bilateral dorsolateral prefrontal cortex), as well as to emotional (ipsilateral claustrum, interconnected to the amygdale) and motor responses (contralateral lentiform nuclei). These findings are particularly interesting because catastrophizing is a significant negative prognostic indicator of treatment success that fortunately can be treated by cognitive behavioral methods. In specific cases, this aspect of treatment may be as important as the treatment of pain pathology. One interesting question is whether the activity in the previously mentioned brain regions is altered by successful therapy, suggesting a mediating mechanism, or not altered by successful therapy, suggesting a risk factor for catastrophizing.

The measurement of pain depends on the goal of the measurement, including diagnosis, choice of treatment, efficacy of treatment, efficacy of a particular component of treatment (clinical trial), and basic research into underlying mechanisms. Another goal, not discussed, is the medicolegal determination of disability. The tools required depend on the goal, ranging from quick clinical procedures, extensive psychophysical methods, physiological measures, behavioral assessment, and interview and questionnaire assessment of pain and of the person.