Any attempt to treat chronic pain necessitates consideration not only of the nature of chronic pain but also why and how it persists. The gate control theory of pain [1] has spawned a plethora of investigations into peripheral and central pain mechanisms. New techniques such as imaging [2] have supported sophisticated and wide-ranging investigations of its neurobiological substrate [3] holding the promise of further enhancing our understanding of fundamental pain mechanisms. However the experience of pain is ultimately perceptual and the nature and impact of chronic pain cannot be understood without consideration of the psychological processes involved. Melzack’s conceptualization of a ‘body-self neuromatrix’, depicts a 3D model of pain perception [4]. The sensory-discriminative dimension is concerned with the location, timing and physical characteristic of pain. The affective-motivational dimension comprises the emotional impact and defensive behaviors such as escape or avoidance. Finally, the cognitive-evaluative dimension focuses on the meaning and consequences of injury and pain. These pathways are at the heart of modern pain psychology and offer a foundation for the biopsychosocial conceptualization of chronic pain. The pathways are differentially activated when people are instructed to pay attention either to the location of pain (sensory-discriminative dimension) or its unpleasantness (affective-motivational dimension) [5]. Experimental studies have demonstrated interactions between neurocognitive processes (such as memory, anticipation and attention) [6,7] and investigated the ways in which cognitive and behavioral responses to pain are acquired [8]. These wide-ranging findings provide a basis for the understanding of pain perception as well as offering a framework for understanding the process of chronicity [9], with implications for treatment and prevention. Chronic musculoskeletal pain does not seem to be simply a function of physically demonstrable disease, poor posture (caused by reflex spasm connected with high muscle tension) or trunk instability [10]; central mechanisms also seem to be involved. Fundamental to the understanding of chronic pain and its development is neurophysiology and the role of central sensitization [11] characterized by an enhanced pain response to normal stimuli (hyperalgesia), a decrease in pain threshold to normally nonpainful stimuli (allodynia) and an increase in spontaneous activity (spontaneous pain), with the involvement of a variety of neurotransmitters and neuromodulators. It manifests clinically as enhanced perceptual responses. There has also been progress in identifying combinations of genes associated with pain susceptibility, initiation, maintenance and aggravation of chronic pain [12]. There is now some evidence of functional and structural differences in the brain possibly associated with the development of chronic pain [2], although focused longitudinal clinical studies are needed to clarify uncertainties about the nature of the psychosocial/neurobiological interface and resolve debate about causal mechanisms. Within pain psychology there has always been a major focus on pain cognitions and pain behavior, but the precise role of emotional learning and cognitive factors still awaits elucidation [3], and the role of the hypothalamic–pituitary–adrenal axis in the development of chronicity is unclear [13]. Considered from an evolutionary perspective, emotion has an important role as a mediator between the perception of threat and the behavioral response [14]. Pain can be viewed as a driver to action, triggering pain behavior, some of which is ‘preprogrammed’ and some of which is learned, which can be adaptive, or in the case of chronic pain can become maladaptive [15]. The importance of the interaction between pain and emotion is further elaborated in discussion of the emotional process–pain (EP–P) model below. Although the precise elucidation of psychological mechanisms is still a “work in progress”, there is clinical evidence for the influence of psychological factors, such as fear, anxiety and catastrophizing, both on the perception of pain and on the response to it [16,17], but until the last decade, there has been relatively little investigation of the specific influence of emotions on pain experience, pain processing and pain behavior [18]. Understanding pain from a functional and evolutionary perspective requires acknowledgement of the social context and, in parallel with the elucidation of psychological factors, there has been an increasing focus on social factors. The social component of pain has always been an important component in the behavioral models [19] as illustrated in the communal coping model [20] and is central to the socio-communication model of pain [21], which is of particular relevance to the consultation. Although psychological and social factors have an important influence on the persistence of pain, its impact on function and wellbeing and on the development of pain-associated disability [22], elucidation of the psychological processes and mechanisms involved has proved a challenge and a number of theoretical models have been proposed. Linton and Shaw [23] highlight some of the core processes/mechanisms and intervention strategies identified in five of the most influential psychosocial models. These models and two additional models currently attracting research interest are described briefly below. The fear-avoidance model [16,24] is one of the most influential and extensively researched models. It is based principally on a learning–conditioning paradigm and assigns central roles to fear, catastrophizing and behavioral avoidance as mediators between pain and chronic disability. Its influence has been widespread not only from a theoretical perspective but also in the development of a specific exposure-based treatment approach for highly fearful pain patients. Recently an interesting interoceptive fear conditioning model of fear of pain has been proposed [25] in which the authors suggest that bodily sensations can function as predictors of pain and as conditioned stimuli become endowed with the capacity to elicit an (anticipatory) fear response. The model seems to require further investigation in the context of constructs such as catastrophization. Acceptance and commitment therapy (ACT), a variant of cognitive–behavioral therapy (CBT), is based on an analysis of both the present context and how it has developed over time. It incorporates consideration of overt behavior, cognition and emotion. ACT has been invoked as a framework for understanding the psychological processes underlying pain persistence [26] and for guiding interventions. It assigns a key role to psychological flexibility in pursuing personally relevant goals, reducing futile attempts to control pain and attempting to establish more contextually relevant behavioral changes. This model allocates a central role to cognitive processes in chronic pain [27]. Worries about pain and cognitive evaluations are viewed as attempts at problem solving, which, while offering an evolutionary advantage, can lead to the persistence of pain as a consequence of misdirected attempts to reduce pain, rather than address more useful and achievable adaptation. It aligns with the perseverance loop in the fear-avoidance model and within ACT as an alternative strategy to the cognitive restructuring at the heart of traditional CBT [9]. Although a general health model rather than a specific pain/disability model, the self-efficacy model has been highly influential in the pain field. It assigns a central role to cognition and highlights specific beliefs about control as potential mediators of clinical outcome. The construct of self-efficacy model is associated with readiness to change, is central to the development of effective pain coping strategies and is an important element in the motivational model of self-management [28]. The self-efficacy model has been shown to be more predictive of treatment response than fear of movement [29] and has been recommended as a core pain outcome measure [30]. Recently a scale specifically focused on self-efficacy in relation to return-to-work has been developed [31]. The stress–diathesis model of post-traumatic stress disorder recognizes the importance of psychological vulnerabilities and predispositions as determinants of the response to pain and how individuals cope with pain [32]. It highlights in particular the role of emotional factors (such as stress, depression and anxiety) in the development of disability. It states that difficulty in differentiating between emotions and physical sensations and, more specifically, in processing emotions at the conscious level (alexithymia) may result in emotionally-laden reports of pain [33]. The EP–P model is a significant further development of the alexithymia model [17]. It assumes that chronic pain develops as a consequence of a process of central sensitization [11,34] in which the mutual influence of emotion and pain perception is central. A distinction is made between primary emotional states (essentially biologically determined drives, motives and impulses), which are adaptive (from an evolutionary point of view), and secondary emotional states (such depression, anxiety and fear), which develop as a consequence of failure to recognize and deal effectively with the primary emotional states, and are essentially maladaptive [35]. In this model emotional processes are seen as essential mediators of the response to pain, and chronic pain is effectively the consequence of emotional dysregulation, solutions for which might include training in mindfulness [36] or emotional disclosure [37,38]. Although these approaches appear promising, further research is needed into their influence on clinical outcome [18]. Like the EP–P model, the predispositional model [37] also attempts to improve our understanding of the relationship between the nature of injury and the persistence of pain by identifying psychological factors which might influence the development of chronic symptoms. However, this model focuses primarily on predispositional traits associated with psychopathology as potential diatheses for chronic musculoskeletal pain. Five such traits have been identified: anxiety sensitivity (AS); illness/injury sensitivity; fear of negative evaluation; pain-related anxiety; and intolerance of uncertainty [39]. AS is characterized by heightened concern about symptoms, as evidenced in catastrophizing, AS can trigger further anxiety and avoidance in its own right, thus becoming part of a vicious self-perpetuating circle contributing to the persistence of chronic pain and pain-associated disability. Like other somatically-focused measures, it is distinct from general anxiety. The illness/injury sensitivity, fear of negative evaluation and intolerance of uncertainty constructs have not been as yet extensively researched in the context of pain, but pain-related anxiety has been investigated. Although AS would seem to account for a large proportion of the variance in pain-related anxiety [40], it has also been viewed as a more fundamental fear than AS, in that nociceptive stimulation appears to be inherently noxious and does not require contextual cues or catastrophic concerns about future events. There is clearly overlap as well as distinctiveness amongst these various constructs, and further research is required into their individual and combined effects. Dispositional characteristics appear in various guises in the current pain and disability models, and seem to hold the promise not only of clarification of the mechanisms underlying the development of chronic pain, but also of influencing the design of more effective psychological interventions. Like the fear-avoidance model, the avoidance–endurance model [41] is supported by an extensive program of research into its components, their neural correlates and their relationship with pain [42]. The avoidance–endurance model postulates that two different response patterns influence the maintenance of pain and disability. The distress endurance pattern is characterized by marked thought suppression, symptoms of depression and task persistence at a behavioral level in spite of pain. The eustress endurance pattern is characterized by ignoring pain sensations, minimizing the meaning of pain experiences and marked task persistence, often accompanied by positive mood. The model maps onto long established ideas of adaptive and maladaptive coping, but from a much more focused perspective. However, at its present stage of development the avoidance–endurance model is perhaps more a model of pain persistence than as a pain and disability model per se, although in the identification of subgroups it would certainly seem to have potential as a basis for screening and targeting (see below). While these models differ in specifics, they all emphasize the powerful influence of psychological factors on the persistence of pain and the development of pain-associated disability, and underscore the need for a patient-centered approach in clinical management. In addition to illuminating the nature of pain and the role of psychological factors, these models have energized the design and development of psychologically oriented interventions. Over the last decade there has been a shift in focus in clinical care from illness to wellness, with a concomitant move towards more patient-centered approaches to medicine. This in turn has stimulated interest in positive adaptation to chronic pain with attention to factors such as resilience rather than vulnerabilities [43,44] and recent constructs such as pain acceptance, hope and optimism have been identified as potential moderators or mediators of clinical outcome. Interestingly they appear to be associated not with perceived pain intensity per se but with adjustment to pain [45] and would seem to merit further investigation. Although there have been significant advances in identification of psychological influences on outcome, variation in clinical presentation however still offers a significant challenge to clinical decision making in the individual case. There have been many attempts to identify types of patient, based on combinations of presenting characteristics (such as personality profile, clusters of symptoms, pain behaviors, perceived etiology and psychophysiological response profiles) [46]. The most extensively validated tool is the Multidimensional Pain Inventory [47], yielding a classification based on scores on six different scales measuring pain and its psychosocial impact, which assigned patients to one of three clusters: dysfunctional, interpersonally distressed and adaptive copers. More recently it has been shown that the Multidimensional Pain Inventory subtypes are associated with distinct patterns of fear-avoidance responses and endurance related responses [40,46]. Such approaches to ‘psychological phenotyping’, since derived from an analysis of psychological mechanisms appear to hold more potential as a basis for interventions than general psychological symptoms unrelated to the outcome. Further sharpening in terms of focus is evident in the design of the STarTBack screening tool [48] based on known prognostic factors for low back pain outcome (discussed below) and the recently developed Pain Recovery Inventory of Concerns and Expectations [49] designed to identify patients at risk of developing back disability. Both of these tools require further investigation. In summary, psychological and social factors appear to be deeply implicated in the experience of pain and the development of pain-associated disability. The psychological components identified in the pain and disability models are all potential targets for intervention. Chronic pain patients consulting in primary care may be referred to clinical psychologists for treatment of concurring mental health problems, but if referred to pain psychologists, the primary focus of therapy (often CBT) is likely to be on psychological adaptation or adjustment to their chronic pain and its associated impact. Pain-specific treatment may include a range of specific cognitive and behavioral techniques [9,50], the most common of which are: disclosure; cognitive restructuring; activity scheduling; relaxation and biofeedback; hypnosis; and exposure interventions, although, as aforementioned, there has been recent developing therapeutic interest in new ways of addressing the emotional component [37] and in ACT [26]. If chronic pain patients require more intensive and skilled psychological approaches, they may be referred to specialist pain clinics for an interdisciplinary pain management program as part of a coordinated rehabilitative approach involving other professionals such as physicians (specializing in pain), physiotherapists, occupational therapist and nurses [50,51]. The psychological intervention may be the primary focus of therapy, whether as a precursor to, or instead of, a multidisciplinary pain program, but may also be built in as part of a comprehensive pain management program. There are five major types of intervention with significant psychological content: education/information (back schools); behavioral interventions; CBT programs; stress management programs; and risk factor targeted interventions [50]. The early, fairly ‘light touch’ approach of the early back schools has evolved into more sharply focused educational approaches such as neuroeducation, a cognitive-based education intervention that aims to reduce pain and disability by helping patients gain an increased understanding of the underlying neurophysiological processes [52] rather than on biomechanics and tissue injury per se[53]. Changes in belief may be important mediators of clinical outcome, and neuroeducation merits further investigation in controlled trials but the principal emphasis to date appears to have been on education in pain physiology, rather than on specific psychosocial features. A comprehensive review of literature on behavioral approaches is presented elsewhere [9] as is literature on stress reduction programs [50], but in the treatment of chronic pain, both feature normally as part of multifaceted programs rather than as standalone treatments. The primary focus of the present article therefore will be on CBT programs and risk factor targeted interventions in which the interventions have a major psychological focus. The modern CBT program has its roots in the behaviorally-oriented rehabilitative approach, with its particular emphasis on pain behavior it was originally developed at the University of Washington in Seattle (USA) in the 1960s. The later incorporation, more specifically of cognitive factors, laid the foundation for modern psychologically oriented pain management. Typically this comprises a variety of methods for directly and indirectly producing behavior change, including methods based on cognitive and behavioral therapy, learning and conditioning processes, skills training, physical exercise, and education [9]. The general approach has been widely adopted in the USA and elsewhere. Recent developments in CBT include contextual CBT, described as a group-based treatment format based on ACT, in which the therapeutic focus is on the influence of thought and feelings on behavior, not on the extent to which these feelings are uncomfortable, frightening or unwanted [54]. The emphasis is on contextual change stimulating the patient to become more aware of the process of thinking rather than its content, thus enabling a greater flexibility of response by disentangling the content of thought from its context within an overall framework of acceptance. At this time, there is a strong evidence base for ‘traditional’ CBT [55], but recent research suggests that ACT may be a feasible alternative [56]. However, as yet there is insufficient data to determine whether treatment matching to one or other of the approaches might enhance clinical effectiveness, or the extent to which differences in pain processing, or patient preference, might underpin differences in outcome. The primary focus of all these techniques has been on the reduction of pain per se or on reducing the impact of pain in order to facilitate optimal function and participation, but can we do anything to prevent unnecessary pain-associated disability? Kendall and colleagues coined the term ‘yellow flags’ to encompass psychological risk factors and social and environmental risk factors for prolonged disability and failure to return to work as a consequence of musculoskeletal symptoms [57]. The original yellow flags monograph provided a guide to the assessment of yellow flags and advice on how to incorporate cognitive–behavioral change principles into early management. In recent years, the focus of research on yellow flags has been complemented by an additional, more specific, focus on occupational contexts and perceptions of work [58] and finally integrated into a system for management of musculoskeletal symptoms in the workplace in which consideration of the yellow (clinical) flags and blue (perceptions of work) flags is complemented by the inclusion of more objective contextual factors (or black flags), including aspects of the work and the working environment [59], and the wider social context. Following their literature review of the strength of psychosocial factors as predictors and of studies attempting to address them in interventions, Nicholas and colleagues concluded: “When candidates are carefully selected on the basis of the presence of yellow flags and when an intervention known to address these factors is competently applied, good outcomes are to be expected. On the contrary, when patients are not selected for yellow flags and psychological interventions are provided indiscriminately, the outcomes tend to be disappointing” [60]. Before attempting to address the link between screening and targeting, it is relevant to revisit the development of cognitive–behavioral interventions. Chronic pain is a reason for consultation throughout the healthcare system and is not just a matter for tertiary rehabilitation services. Addressing beliefs, expectations, behavioral responses to pain and, perhaps to a lesser extent, emotional responses, is an integral part of musculoskeletal management, although not usually an explicit focus of clinical practice. Most treatment is not delivered by healthcare professionals with specific training in psychological principles or practice and could not be regarded as full-blown CBT. However aspects of the cognitive–behavioral approach can be seen in the development of secondary prevention programs. The earliest programs were developed from the back school approach (originally developed in occupational settings with the objective of primary prevention), but the fairly didactic educational approach also characterized the early secondary prevention programs [61]. The initial programs had a primary biomedical focus, characterized by emphasis on biomechanical and ergonomics, with little emphasis per se on psychological factors. The description of later programs suggested a more specific focus on psychological factors had been included and that education about fear and avoidance may have been important [62,63] but the extent to which such factors specifically were targeted is not reported. During the last decade, programs have become more wide-ranging in scope (similar in breadth to tertiary pain management programs, but less intensive). In one of the first such studies, Hay and colleagues demonstrated efficacy, comparable with manual therapy with a simple, individualized cognitive–behavioral pain management approach, thus demonstrating that the approach was acceptable to patients and could be delivered by trained physiotherapists [64]. Such studies offered encouragement for a psychologically-oriented approach to reactivation, but considered overall, the results (in terms of size of effect) of randomized controlled trials of secondary prevention were somewhat disappointing and van der Windt and colleagues concluded the following [65]:
Patient classification presents a significant methodological (and sometimes logistical) challenge [66] but patient profiling based on the psychological constructs most strongly associated with development of disability such as pain-related fear, catastrophizing and depression [67] has been undertaken and five distinct profiles, suggestive of a number of different potential pathways to chronicity, have been identified [68]. However, in one of the first studies of stratified care based on the identification of clinical (mainly psychological) risk factors, patients were allocated on the basis of a 9-item screening tool [48] into three risk groups (low, medium and high) for which treatment protocols, with associated clinical pathways, had been developed by clinical experts. The stratified approach in the intervention arm was compared with three comparable groups whose risk status had been determined, but who all received best practice usual care (without stratification). Those allocated to the low risk intervention received a single 30 min assessment and treatment session delivered by their GP according to an agreed protocol. Medium-risk patients were referred for standardized physiotherapy to address symptoms and function. High-risk patients were referred for psychologically informed physiotherapy to address physical symptoms and function, and also psychosocial obstacles to recovery [69]. The overall stratified approach proved superior in terms of clinical and economic outcomes to nonstratified treatment, and is now being widely adopted [70]. There is a wide range of social perspectives of potential relevance to pain and its treatment. There may be lessons to be learned from the determinants of healthcare consulting for common health conditions in the general population but the topic is beyond the remit of the present article. However, an overview of the specific influence of cultural, economic and occupational factors is presented elsewhere [50] and it would appear that there is wide variation across cultures in the prevalence of disabling musculoskeletal pain between occupational groups carrying out similar tasks [71]. In the concluding section of the present article, the social aspects of chronic pain will be considered from a social-learning/developmental perspective, and then implications for treatment will be discussed: firstly in terms of the role of significant others in pain rehabilitation and secondly in terms of social aspects of consultation. By the time chronic pain is established there will have been many factors that have had an influence on the nature of chronic pain, and how it is communicated and expressed. A distinction has been made between two types of pain response: reflexive behavioral reactions, which are unintentional and probable manifestations of hard-wired evolutionary responses [15], and responses which are mediated by higher levels of central processing (including volitional control), are established by a process of social learning and heavily influenced by the social context [21]. These pain communications elicit responses in others, responses which are also partly preprogrammed, as evidenced by parental responses to pain expression by children, but also by empathic responses of significant others and healthcare professionals to chronic pain patients [72]. The process is perhaps best depicted in the sociocommunications model of pain [21], a general communications framework in which the internal experience of pain is coded in pain expression by the pain sufferer and then decoded by others who may respond to the pain behavior in a variety of ways. An important distinction is made between the experience and the expression of pain, each of which has its determinants and influences, and which can be understood both in terms of the immediate context and also from a developmental or historical perspective. Thus both the expression of pain and the response to the pain behavior in others can be understood partly in evolutionary terms as biologically programmed but also as further shaped by social learning. Since the advent of behavioral perspectives on pain [19] there has been interest in the role of significant others in patients presenting with chronic pain although there are significant methodological challenges in this field of enquiry [73]. Marital distress as well as maladaptive interaction is associated with pain, interference and depression in patients with chronic pain [74]. Early observational studies focused on the role of operant factors [75], but there has been interest more recently also in cognitive factors [76]. Significant others are an integral part of the communal coping model of catastrophizing [20]. The pain-interaction process has been suggested as a mechanism which might explain the progression of acute to chronic pain in couples [77] and the nature of empathy between pain patients and their spouses has also been a specific focus of research [78,79]. Finally, interest in the role of significant others has simulated new approaches to intervention such as spouse-assisted coping skills training [80] and integrative behavioral couple therapy [81]. Patient beliefs and expectations have an important influence on the decision to consult, response to treatment and clinical outcome, but they are frequently not specifically addressed during consultations [82]. Although there is a large amount of literature on the teaching of consultation skills, disentangling the effective components is a major challenge, particularly in the context of patient-centered medicine in which there is supposed to be less emphasis on information transmission per se and more emphasis on responsiveness to the patient agenda. Establishing a therapeutic climate and facilitating self-disclosure underpin successful identification of patients’ beliefs and expectations. Addressing specific concerns, and clarifying mistaken beliefs at the outset will facilitate the development of an agreed plan of action. All patients should be provided with a credible but simple explanation of the difference between acute and chronic pain, the importance of central pain mechanisms and the development of disability. This explanation must be provided using language and terminology that patients understand, to ensure that it shapes their beliefs and expectations and optimizes their pain coping strategies [83]. Although much has been learned from the teaching of doctor–patient communication skills about the nature of communication, perhaps surprisingly the burgeoning literature on the nature and effectiveness of psychosocial interventions has generally failed to address either the importance of therapeutic competency or the appraisal of treatment fidelity. A more systematic and targeted approach to pain communication would still seem to be needed. In developing a clinically relevant research agenda, a specific focus on observational learning has been suggested as a framework in which to investigate whether particular types of pain-related beliefs or behaviors are more affected by learning from others, and whether particular environmental or observer characteristics moderate observational learning for some beliefs or behaviors and not for others [84]. In conclusion, in the clinical management of patients presenting with chronic pain, appreciation of the social context, and in particular the nature of the communication process, would appear to be particularly important and seem to merit further investigation. The purpose of this review has been to highlight the nature and influence of psychosocial factors in patients presenting with chronic pain. Although research has helped identify some of the more powerful factors and investigate their influence on patient outcomes, we are still clearly at an early stage in identifying moderators and mediators influencing the persistence of pain and the development of disability. Nonetheless, a number of conclusions can be drawn:
The challenges over the next 5–10 years are to further develop our understanding of the role of psychosocial factors in the development and prevention of unnecessary chronicity; clarify the influence of psychosocial factors as moderators and mediators of response to treatment; and design better interventions which incorporate not only the psychological processes, but also the social context of pain, with reference not only to the environment of the pain sufferer, but also within the consultation itself. There is every reason to be optimistic that progress will be made, but such an agenda will require not only more clearly focused research but improved research methodology.
CJ Main owes a considerable debt to M Hasenbring, A Rusu and D Turk; and to H Flor and D Turk, for their excellent textbooks (2012 and 2011, respectively), to both of which CJ Main has made frequent reference. Thank you also to L Goubert and M Lumley for generous assistance. However, the author takes full responsibility for any inadvertent inaccuracies or misinterpretations. This work was supported by Arthritis Research UK Primary Care Centre (grant numbers 18139 and 20202) and Arthritis Research UK who funded the STarTBack study (grant code 17741). The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript. Papers of special note have been highlighted as: ▪ of interest ▪▪ of considerable interest References
|