Introduction

Electrical stimulation for the treatment of pain dates back to 46 ad when Scribonius Largus described the use of torpedoes, a fishlike animal capable of emitting an electric discharge, placed over the area of pain for relief from intractable headache and arthritis. In 1745 the Leyden jar allowed physicians to control electrical current, and its use spread rapidly. Electrical stimulation of the brain was noted in 1950. Shealy, Mortimer, and Reswick¹ reported on the use of cardiac pacemaker technology to deliver electric current to the spinal cord via surgically implanted electrodes in 1967. Remarkable surgical and technological advances over the ensuing half century have resulted in various types of percutaneous and surgically implanted paddle leads capable of delivering thousands of different stimulating patterns using totally internalized, radiofrequency coupled, or rechargeable pulse generators. This technological flexibility has dramatically broadened the horizon of clinical application. The emphasis in this chapter is on the use of electrical stimulation in the treatment of chronic pain.

Descartes’ explanation of pain mechanisms and processing² put forth in his 1664 book Treatise of Man held sway from the 1600s until Melzack and Wall presented their gate control theory in 1965.³ The latter theory, which has now undergone many revisions, allowed for the role of psychological factors in the modulation of pain. The ingenious model of chronic pain put forth by Apkarian, Baliki, and Geha⁴ provides for even greater clarification of the role of psychological factors. Spawned by the revealing work on pain mechanisms and system reorganization,⁵ pain processing,^6,7 and neuroimaging,^8,9 the model is neurophysiological in nature. It highlights the brain as a “…dynamical network wherein detailed connectivity is consistently modified by the instantaneous experience of the organism.”^4(p95) Although the involvement of the nociceptive transmission system (i.e., spinal thalamic pathways) is acknowledged, activity at the cortical level is central to the theory.

Along with changes in our understanding of the neurophysiological aspects of “pain” (I use pain inside quotation marks because its definition and our understanding of its nature [i.e., disease vs. symptom vs. syndrome vs. emergent phenomenon] continue to evolve), the psychological/psychiatric conceptualization has changed as well. Psychoanalytical theories of pain based on the work of Sigmund Freud were popular in the psychiatric and psychological communities in the 1960s.¹⁰ The psychodynamic approach of Freud held pain to be a means of controlling the expression of unwanted and unconscious desires or motivations. Engel¹¹ followed by detailing the “pain-prone personality.” The essential features of the psychodynamic approach included (a) pain as a common conversion system, often with symbolic meaning; (b) unpleasant affect, usually guilt, hostility, resentment, or conflict is converted to bodily pain; (c) the choice of the symptom is determined by precipitating events; and (d) frequently there is a hereditary influence, most always a physical substrate, if only muscular.^10,12

Although largely replaced by psychological approaches based on learning theory, the legacy of psychodynamic theory lives on in the classifications of Somatization Disorder and Pain Disorder found in the revised fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) of the American Psychiatric Association.¹³ Because of the tendency of such terms to continue to invigorate the notion of pain as a psychogenic phenomenon, Merskey¹⁴ has called for their removal from the DSM. In fact, the term Complex Somatic Symptom Disorder has been proposed to replace Somatization Disorder, Pain Disorder, Hypochondriasis, and Undifferentiated Somatoform Disorder in the upcoming DSM-V^13a and eliminate “medically unexplained symptoms” as a diagnostic criterion (www.dsm5.org/ProposedRevisions). The DSM Working Group is hopeful that this change will eliminate any unintended reference to mind-body dualism and believe that it is more in keeping with the identification of somatic symptoms and cognitive distortions as shared features among the four existing disorders listed.

Early on, C. Norman Shealy¹⁵ recognized the importance of psychological factors in the treatment of pain as illustrated by his recommendation of (a) the absence of elevations on the Minnesota Multiphasic Personality Inventory (MMPI), except on the depression scale; (b) emotional stability; and (c) involvement in a rehabilitation program such as patient selection criteria in the application of spinal cord stimulation (SCS). Indeed, the psychological status of the patient was noted by Long¹⁶ to be the most common reason for failure of stimulation therapy. One would be hard pressed to find any reputable publication on the subject that does not put forth psychological factors as important to the outcome of SCS therapy for the treatment of pain. In fact, the National Institute for Health and Clinical Excellence (NICE)¹⁷ strongly recommends that SCS therapy for chronic pain be carried out within a multidisciplinary setting.

On surveying the outcome studies from the 1970s to the1980s, Bedder¹⁸ concluded an estimated 40% success rate for SCS therapy. In a 2005 article Taylor, van Buyten, and Bucher¹⁹ noted a significant pain reduction in 62% of patients with 40% returning to work, 53% discontinuing their use of opioids, and 70% expressing satisfaction with SCS therapy. Any concern over significant activity restrictions in patients with a properly secured SCS device appears resolved by reports, including one in 2009, documenting the return to active duty in combat zones of military personnel following internalization of SCS.²⁰

Less encouraging are the results of a prospective controlled study that compared SCS therapy, pain clinic, and usual care for failed back surgery syndrome (FBSS) in patients under worker compensation. By way of summarizing their data, the authors concluded that “…the SCS group did not differ from the other groups at 12 or 24 months on any outcome, including leg pain intensity, physical function, back pain intensity, and mental health. Outcomes were poor in all groups … fewer than 6% of patients achieved success on the primary outcome (a composite index of improvement in pain, function, and medication use); fewer than 10% were working; and more than twice as many patients reported a decline as report improvement in ability to perform everyday tasks.”^21(p23)

May, Banks, and Thomson²² reported that, of the 100% of patients reporting success 16 months after implantation, only 59% did so at 58 months. Disease progression, adaptation, or tolerance to the stimulation; “a regression to the mean”; misinterpretation of the screening trial; and psychological variables may be contributing factors. These very disparate results combined with the loss of effect over time suggests that factors other than those related to surgical technique and device function are contributing to therapeutic outcomes. Indeed, North and Shipley emphasized that, “Technical success, however, is not sufficient to ensure clinical success.”^23(pS202)

Psychological Evaluation

In performing a psychological evaluation, it is important to consider several basic assumptions of pain management. First, persistent pain, regardless of its associated physical pathology (e.g., malignant tumor, degenerative disc disease, or osteoarthritis), is multidimensional and therefore influenced by psychosocial factors. Second, these psychosocial factors can influence the outcomes of pain relief–oriented therapies; in addition, co-use of behavior/psychological therapies along with somatic therapies can enhance pain relief and functioning. Finally, the prominence and priority of psychosocial factors can vary according to the type and degree of pain and pain-related pathology.²⁴

Several years ago my colleagues and I examined the literature and, along with our own experience, listed a number of hypothesized positive patient characteristics thought to indicate an appropriate patient to proceed to a screening trial.²⁵ These positive indicators included (a) generally stable psychologically; (b) cautious, sufficiently defensive, self-confident; (c) self-efficacy, ability to cope with setbacks without responding in emergent fashion; (d) realistic concerns regarding illness often associated with a congruent mild depression; (e) generally optimistic regarding outcome, with the patient and significant other having appropriate expectations; (f) comprehends instructions and has a demonstrated history of compliance with previous treatments; (g) appropriately educated regarding procedure and device, supportive and treatment-educated family/support member(s); (h) behavior and complaints consistent with pathology and a behavioral/psychological evaluation consistent with the patient’s complaints and reported psychosocial status; and (i) able to tolerate electrical stimulation, perhaps evidenced by a trial of transcutaneous electrical nerve stimulation (TENS).

Patient beliefs which were thought to be associated with a less positive outcome include (a) pain is a purely physical phenomenon, (b) psychosocial factors play little role in pain and treatment outcome, (c) chronic pain means loss of productive life, (d) pain can only be relieved if the medical cause (e.g., arthritis, scar tissue) is eliminated, and (e) medical technology holds the solution. Those appearing to correlate with a positive outcome are: (a) pain is multidimensional and multi-factorial, (b) attitudes and behaviors can affect treatment outcomes, (c) coping skill (e.g., relaxation, distraction, goal setting) can be helpful, (d) an active participant in therapeutic decision, (e) support systems that reinforce positive behavioral change are useful, and (f) proper expectations influence outcomes.

In addition, we proposed a set of clinician beliefs and attitudes thought to be more closely linked to a negative or positive outcome:

Assessing and ensuring that the perspective SCS patient has proper, realistic, and appropriate expectations is seen as part of the evaluation process. The term expectation is often used in somewhat of a parochial manner, perhaps because its clinical impact may tend to be underestimated. Imaging studies have examined the effect of both positive and negative expectations. Expectancies have been demonstrated to affect the activity of the pain modulatory system, including the anterior cingulated cortex (ACC), thalamus, prefrontal cortex, and insula cortex.^26,27 Expectations related to anticipated good or bad motor performance influenced the outcome of subthalamic nucleus stimulation in patients undergoing deep brain stimulation.²⁸

Furthermore, expectations contribute to the placebo effect. Price, Finiss, and Benedetti²⁹ noted the average placebo effect to be 2 U on a 0 to 10 pain scale in their total study population. However, it was as high as 5 U in the placebo responders (which varied from 20% to 55%, depending on the study). Expectations accounted for 49% of the variance. A series of studies by Dolce and associates^30–32 reported the impact of self-efficacy and expectancy on pain and exercise tolerance. It might be of interest to study the relationship between patient expectations and the reported loss of analgesia over time, despite a functioning SCS device. The author is unaware of any validated psychological instrument designed to evaluate expectations, particularly as they related to neuromodulation. However, the meeting of patient expectations has been shown to be central to good clinical outcomes.³³ This highlights the importance of appropriate expectations. Here again one encounters the necessity of the clinical interview.

How might the awareness of the patient’s expectations and the possibility of the placebo effect influence the screening trial? For one, the clinical interview could obtain information regarding the level of pain acceptable to the patient, allowing him or her to be more active (i.e., the functional pain level). The patient could then be encouraged to outline their functional goals (e.g., sitting longer, walking further, traveling more), which can be individualized for each patient, depending on his or her particular anatomical limitations. Second, the screening trial might then examine both cognitive or perceptual effects of SCS (i.e., pain ratings) and functional changes. The pleasure gained from being able to perform reinforcing activities again and the accolades from others may be important to long-term improvement. It is likely to be more difficult for a patient to perceive himself or herself as more functional, especially if one involves and solicits observations from a significant other in the screening process, than it is to experience a reduction in pain. In a sense the preimplant screening trial can function as an extension of the interview. Agreed on therapeutic goals can be addressed during the trial and used as a means to determine the desirability or appropriateness of proceeding to implantation. Logic would suggest that the more closely the trial circumstances mimic the final outcome, the less chance there is of a false-positive trial. This type of a screening trial mimics the N-of-1 approach illustrated by Cepeda and colleagues.³⁴

North and Shipley²³ published what might arguably be one of the most comprehensive reviews of the SCS literature. Over 20 participating experts reviewed some 300 articles spanning 40 years from 1967 to 2007. The document summarized some psychologically relevant information. Regarding psychological predictors, it was noted that, “We lack sufficient information to predict SCS outcome from the result of a pretreatment psychological evaluation, but SCS, as is the case for every interventional pain treatment, is reserved for patients with no evident unresolved major psychiatric co-morbidity.”^23(pS233) Concerning the benefits of a psychological evaluation, they stated that it “…provides patient selection information by identifying the small percentage of patients who might benefit from psychological treatment before undergoing SCS therapy or in whom SCS therapy might be complicated by psychological factors.”^23(pS234) The literature was interpreted to suggest that the psychological evaluation be conducted before the screening trial when a surgical lead was being used, before anchoring if percutaneous leads are used, and before internalization. The various tests that had been used in the studies reviewed included the MMPI with Wiggins content scales^35–37; Symptom Checklist-90-R³⁸; Derogatis Affects Balance Scale³⁹; Chronic Illness Problem Inventory⁴⁰; Spielberger State-Trait Anxiety Inventory (STAI) Scale and State-Trait Anger Scale^41–44 ; Beck Depression Inventory (BDI)^45–47; Locus of Control Scale^48–49; Absorption Scale⁵⁰; McGill Pain Questionnaire (MPQ)^51–52; Social Support Questionnaire⁵³; Sickness Impact Profile (SIP)⁵⁴; Oswestry Disability Index (ODI)⁵⁵; Roland Morris Questionnaire⁵⁶; and Fear-Avoidance Beliefs Questionnaire⁵⁷(Table 4-1). Interestingly, the authors noted Conversion Disorder to be a condition that could escape detection.

Table 4-1 Advantages and Disadvantages of Various Psychological Tests

From Raj PP, editor: Practical management of pain, ed 3, Mosby, 2000, p 414.

An unresolved major psychiatric co-morbidity; unresolved possibility of secondary gain; an active and untreated substance abuse disorder; inconsistency among the patient’s history, pain description, physical examination, and diagnostic studies; abnormal or inconsistent pain ratings; and/or a predominance of nonorganic signs (e.g., Waddell signs) were listed as psychological factors that should cause the clinician to defer, delay, or modify the screening trial.^23(pS238) The inability to control the device was considered an absolute contraindication. A successful screening trial, which tended to range from 3 to 8 days, was defined by (a) 50% or greater reduction in pain, (b) decreased pain despite provocative physical activity, (c) stable or reduced analgesic consumption, and (d) patient satisfaction. A 7- to 14-day postimplantation follow-up visit with monthly visits fading to annual visits was recommended.

The specifics and parameters of the evaluation process and what it means to psychologically clear a patient for SCS trial or internalization remains ill defined. One main reason may be the tendency for authors to merely state that their patients had been cleared psychologically without revealing, or being required to reveal, the screening process and outcomes. This observation becomes particularly poignant when considered in light of the fact that some 25% to 50% of implanted SCS patients report the loss of analgesia 12 to 24 months after implant, despite a functional SCS unit and continued concordant paresthesias.^58–60 Nevertheless, Long and associates⁶¹ reported a 70% “success” rate in patients who were screened and only 33% in those who were not.

In 2004 an Expert Panel Report incorporating input from clinicians in Europe and the United States⁶² addressed the issue of psychological assessment for SCS therapy in managing chronic pain. The pretrial assessment was to have two objectives: (1) to determine the presence of psychological and social characteristics that could increase the probability of benefit; and (2) to help the physician identify the small number of patients in whom this treatment would result in uncertainty, failure, or medicolegal consequences.^62(p214) The panel recommended evaluating (a) the present status and knowledge of the pain and it sensation, (b) painful behaviors and moods of the patients, (c) the patient’s premorbid personality structure, (d) environmental factors affecting the pain, and (e) the patient’s personal strengths and internal resources. The evaluation should include a clinical interview, structured inventory for pain, and psychometric testing. Specific measures of depression, anxiety, personality, and coping skills were reviewed. The panel did not provide a list of inclusion or exclusion criteria but rather a review of the literature. In their recent review of the intrathecal therapy literature from 1990 to 2005, Raffaeli and colleagues⁶³ concluded that the psychological evaluation should explore: (a) patient expectations, (b) quality and meaning of the patient’s pain, (c) psychological disease, and (d) barriers to patients and family compliance with treatment. The European Federation of the International Association for the Study of Pain Chapters (EFIC) goes a step further. In its 1998 Consensus Document on Neuromodulation Treatment of Pain, the panel states that, given the intricate relationship between psychological and physiological factors in chronic pain, neuromodulatory procedures “are not to be considered as standalone treatments.”^64(p208)

In the United States a psychological evaluation is mandated by the Centers for Medicare and Medicare Services (CMS: Medicare Insurance). The policy states, “Patients must undergo a careful screening, evaluation and diagnosis, by a multidisciplinary team prior to implantation. (Such screening must include psychological as well as physical evaluation)” (Medicare Guidelines: SCS. Psychological evaluation. Medicare Coverage Issues Manual: Section 65-08: Electrical Nerve Stimulators). The American Psychological Association (APA)^64a has outlined ethical guidelines regarding psychological assessment. Section 9.01b states, “Psychologists provide opinions of the psychological characteristics of individuals only after they have conducted an examination of the individual adequate to support their statement or conclusions”; and Section 9.06 states, “When interpreting assessment results, including automated interpretations, psychologists take into account the purpose of the assessment and various factors, test-taking abilities, and other characteristics of the person being assessed such as situational, personal, linguistic, and cultural differences that might affect the psychologists’ judgment or reduce the accuracy of their interpretation.”^64a

Deyo and associates⁶⁵ have recommended assessing pain, mood, personality, and function when treating chronic pain conditions. Consistent with their suggestion, the authors have relied on the MPQ, BDI, MMPI, ODI, and clinical interviews in the evaluation of SCS candidates. The total patient/family time required for the evaluation is approximately 3 to 4 hours at a cost of about $500.00 U.S. dollars or less, depending on insurance coverage. One or more preparation/educational sessions⁶⁶ may be involved, depending on the particular patient’s needs. In recent years increased attention has been given to psychological factors vs. diagnosis/states (e.g., depression, anxiety) and statistically derived cutoff scores. The latter represent a sterile, deterministic, and statistical approach to the chronic pain patient for which there is little clinical and theoretical support. Do those espousing this approach really believe that patients are incapable of change? Their readiness-for-change is obviously an issue to be considered. In addition, we are also focusing more on patient expectations (functional pain level), functionally oriented trials (patient specific goals), conscientiousness (committing to and accepting obligations), and activity engagement (active despite pain).

As recommended by North and Shipley,²³ the evaluation should be conducted before the final determination is made about a temporary SCS trial. Whenever possible, the clinical interview should include a significant other, given the reported discrepancy between the patient and significant-other interpretation of treatment effectiveness.^67,68 The patient’s goal(s), expectations, current coping strategies, level of readiness for change, degree of acceptance of the realities of his or her situation and projected outcome, and potential complications and side effects can be measured informally via the interview or by the use of questionnaires designed for that particular purpose.⁶⁹ Information should be sought regarding (a) any untreated or undertreated major affective disorder; (b) presence of any personality disorder (PD) and how it may affect the patient’s perception of pain, compliance, or cooperation; (c) any undertreated or untreated drug or alcohol abuse problems, present or past; (d) the contribution of nonphysical factors to the patient’s pain perception and behavior, (e.g., spousal reinforcement, secondary gain); and (e) the type and degree of social support.

Block and associates⁷⁰ have created an algorithm containing a number of risk factors obtained from the clinical interview, psychological testing, and the patient’s medical background. These factors include personality types, mood states, and insurance status. Each factor is assigned a value of 0, 1, or 2 on the basis of how strongly it has been associated with outcomes, predominantly in the spine surgery literature. The total score is combined with the number of adverse clinical features, (e.g., deception, PD, medication seeking). Patients are then classified as having a good, fair, or poor prognosis.

This presurgical behavioral medicine evaluation (PBME) has only recently been applied in the field of neuromodulation.⁷¹ The usefulness of this scheme might be enhanced by adding variables derived from the preimplantation trial, an advantage not available with spinal surgery, creating a pretrial and a posttrial score.

Areas of patient responsibility such as exercise and increased functional activity accompanying a reduction in pain should be emphasized to minimize an overdependence on the SCS technology as the sole source of clinical efficacy. Patients should be encouraged to understand that there may be a hierarchy of symptom improvement. For example, improved functioning and quality of life do not always accompany a reduction in pain but may need to be treated as a separate targeted problem. Similarly, I have recently encountered several patients with sacral stimulators wherein improvement in bladder functioning preceded significant reduction in pelvic pain. As noted previously, appreciating pain as an experiential, multifactorial symptom, the components (sensory, affective, cognitive) of which interact in a dynamic fashion, at times rendering pain intensity and unpleasantness (affect) somewhat independent but related aspects, should influence the assessment. Instruments emphasizing this multidimensional approach should be favored.

Allowing the patient to examine and manipulate the hardware to be used may be beneficial, as can be the viewing of audiovisual materials. Assessing as best one can the patient’s acceptance and willingness to cope with persistent concordant paresthesias and the potential for electrode migration and positional sensitivity is a priority. Although these matters are often reviewed by the implanter, the repetition can be useful.

The information presented to the patient, level of understanding and comprehension of the patient and significant other, degree of discussion/agreement, and acknowledgement of awareness of complications and side effects should be documented in the chart note. The ill-prepared or uncertain patient and significant other may benefit from additional educational/orientation sessions. Projected outcomes should be based on evidence-based literature.

Levels of patient anxiety and depression can be assessed via the BDI and STAI. These tests can provide a measure of baseline depression and anxiety. Those patients exhibiting a significant amount of apprehension may benefit from cognitive behavioral therapy, relaxation, or other stress management pretrial. A certain amount of anxiety and depression is to be expected and should be appreciated as such. The MMPI and Symptom Checklist 90 (SCL 90) can yield a measure of the patient’s general psychological status and personality tendencies. Patients manifesting manipulative or strongly maladaptive PDs should be identified and approached with caution.

The role of PDs is sometimes overshadowed by the emphasis on the mood disorders. PDs make up a large part of the Axis II diagnoses in DSM-IV scheme and are marked by “…behavior that deviates markedly from the expectations of the individual’s culture” and leads to distress and impairment.^72(p287) A personality trait represents a pattern of perceiving or relating to one’s environment and presents as less pathological compared to PDs. The incidence of PDs in chronic pain has been estimated to be as high as 50%.^73,74 The PDs have been conveniently ordered into three clusters. Cluster A (i.e., paranoid, schizotypal) are characterized by individuals with odd or eccentric behavior. Cluster B includes the more dramatic, emotional, and manipulative individuals such as borderline, histrionic, narcissistic, and antisocial PDs. The anxious, fearful, and depressive PDs (i.e., dependent, avoidant, and obsessive-compulsive) make up cluster C.

As discussed by Doleys,²⁴ cluster A patients are more prone to unusual somatic experiences. They may perceive the hardware or stimulation as producing some psychological or somatic distortions. In extreme cases there may be associated hallucinations or somatic delusions. Cluster B–type patients often pose the greatest management problem. They tend to be noncompliant, challenging of authority, and demanding. They may have hidden agendas and pose significant management problems. Cluster C patients may benefit from behavioral therapies to address their fears, anxieties, and depression, any of which can influence their perception of pain and degree of disability. In our experience these patients are more likely to have a better short-term vs. long-term result and pose a risk of a false-positive trial (i.e., they have good response to the preimplant trial but report decreased effect over time).

The use of psychophysiological scaling might prove useful during the screening trial. This would involve several, perhaps four or five, trials wherein the SCS intensity is gradually increased (ascending trials) to establish detection threshold, analgesia, and tolerance. Descending trails (i.e., starting at tolerance and decreasing the SCS intensity to the level of stimulus detect) can be interspersed among the ascending trials. An overall average for stimulus threshold, analgesia, and tolerance can then be calculated. This information would yield a usable range, which is the difference between the average threshold intensity and the tolerance intensity of the SCS stimulation. The characteristics of this usable range and the location of the analgesic intensity within the usable range could be evaluated every 6 months after implant. Comparing changes in these characteristics with pain relief (i.e., magnitude of the useable range) may give additional information to use during the screening trial. In addition, certain psychological states or personality characteristics such as hysteria and hypochondriasis may be associated with certain patterns. For example, highly anxious and emotionally reactive patients may have a much more narrow usable range and thus be more prone to loss of analgesia after implant despite a positive trial.

A significant percentage of patients report the negative impact of pain and related psychological distresses on their quality of life. Although this technically defines a somatoform disorder, it does not necessarily exclude the patient as an appropriate candidate. The greater concern is for those patients meeting the DSM-IV-TR¹³ criteria for somatization disorder, indicating a lifelong pattern likely to be unaltered by an intervention. The evaluation process should also focus on identifying the presence of any substance abuse or addiction problem. Most feel that an active, especially untreated, drug abuse or addiction disorder to be an absolute contraindication SCS therapy.

One frequently asked question is how to find a pain-oriented psychologist to perform these evaluations. There are several ways: (1) contact universities granting a doctoral degree in psychology; (2) obtain the membership rosters of the American Pain Society (APS) and International Society for the Study of Pain (IASP) as they list members by degree and geographical location; (3) contact the state board of examiners in psychology; and (4) discuss with your local technology representative. Not all psychologists, even those already practicing in the area of pain psychology, are familiar with the nuances of SCS therapy. However, a growing amount of reading material, conferences/workshop, and proctorships is available. A discussion with the prospective evaluator should take place to emphasize that the evaluation for SCS therapy be objective and not biased by concerns over potential referrals.

Some pain physicians have chosen to bypass this process by ordering and administering computer-scored tests to their patients and relying on the interpretation of the computer. No matter how much effort has gone into the development of such questionnaires (and I have no doubt as to their validity), to use them in the absence of a clinical interview/examination is tantamount to performing surgery on the basis of radiologist interpretation of magnetic resonance imaging without examination of the patient. Although more expedient, this approach may be doing a disservice to the patient, and I question its potential benefit and ethics.

Brief Review of the Literature

Although obviously finite, there are an ever growing, and perhaps inestimable, number of psychosocial factors capable of influencing a patient’s experience of pain and pain-related disability on quality of life. These psychological factors may be mediators, moderators, or maintainers of the outcomes. Indeed, their influence may vary over time in accordance with the patient’s circumstances (e.g., resolution of insurance claim, effectively treated depression or anxiety, degree or pain acceptance, progression through the stages of change). Gender, sex, ethnicity, culture, and type of pain interact with psychological factors in ways that are only beginning to be understood. Our speculations (and that is about all that we can call them at this time) about these relationships, interactions, and/or associations are based on the assumption that we are measuring the right things in the right way.⁷⁵

The following is a brief summary of some of the studies examining psychological variables in SCS. The information presented should be interpreted with the preceding in mind. Most studies have used preimplantation or preimplantation and postimplantation psychological evaluations in an effort to determine which psychological variables tend to be associated with or predict the outcomes. More detailed reviews can be found elsewhere.^76–79

Several studies have used the MMPI. Patients with elevations on hypochondriasis (scale 1), depression (scale 2), and hysteria (scale 3), up to two standard deviations (SDs), were thought to be appropriate for SCS trial; but those with elevations on four scales or more not.⁷⁸ Long and Erickson⁸⁰ recommended including patients with evidence of depression (scale 2) and anxiety (scale 7). However, elevation of hypochondriasis (scale 1) and depression (scale 2) have been associated with negative outcomes.^81,82 North and colleagues⁸³ found that patients with higher scores on hypochondriasis(scale 1) and hysteria (scale 3) tended to have successful trials resulting in internalization, but those with high scores on hysteria (scale 3) reported diminished effect at 3-month follow-up. The appearance of a conversion V pattern, as represented by elevations in hypochondriasis (scale 1) and hysteria (scale 3) relative to depression (scale 2), is thought to suggest the presence of a conversion disorder. Somewhat surprisingly, and perhaps contrary to conventional wisdom, patients with similar profiles have demonstrated successful preimplantation trials and/or long-term outcomes.^82–86 Lower scores on depression (scale 2) and mania (scale 9) correlated with a positive outcome in the Olson and associates’ study.⁸⁷ Meilman, Leibrock, and Leong⁸⁸ failed to find any correlation between MMPI scores and SCS outcomes. However, they noted the accuracy of prediction to be greater for simple mononeuropathies (71%) compared to the more complex arachnoiditis (32%). This observation suggests a possible interaction between the complexity or the underlying pathology and psychological factors.

Depression and anxiety are among the more common psychological co-morbidities found in chronic pain patients. Using the BDI (scoring range 0 to 63), Olson and colleagues⁸⁷ found that patients having a successful outcome had an average score of 12.4 (SD 6.9), whereas those who were determined to have a poor outcome averaged 16.4 (SD 6.5). Although the average BDI score decreased after implant, the change was not statistically significant. Daniel and associates⁸⁹ claimed an 80% accuracy rate when combining the results of the MMPI with the BDI. Burchiel and colleagues^85,86 noted that the BDI emerged as predictor, but their average score was only 13, considered to be in the mild range of depression. The average prereduction/postreduction was 2 points. This relatively small effect raises the question as to whether the finding is of any clinical significance or merely a statistical anomaly.

The STAI (scoring range 0 to 80) was used by Olson and colleagues⁸⁷ as a measure of anxiety. Patients proceeding to SCS trial scored an average of 23.5 and 19.3 on State and Trait anxiety respectively, whereas those who did not go on to a trial scored 25.1 and 21.1. The average STAI scores in Long, BenDebba, and Tirgerson⁹⁰ were 39 SAI and 37 TAI. A sample of patients in the author’s day treatment program had average SAI of 51 and TAI of 47. These different STAI scores could represent three different chronic pain populations. Therefore the relationship of STAI scores in the lower range, as in the Olson and associates’ study,⁸⁷ with SCS outcomes might not hold for groups scoring in the higher range. Thus the correlation between the degree of anxiety and long-term outcome with SCS remains unclear. The relationship between an anxiety disorder and SCS also remains uncertain. However, Verdolin, Stedje-Larasen, and Hickey⁹¹ reported on the treatment of neuropathic pain by SCS in war veterans also diagnosed with posttraumatic stress disorder (PTSD). The presence of PTSD did not affect, nor was it affected by, successful SCS therapy.

One area often overlooked is that of the consequence of SCS therapy on psychological conditions. Panic attacks beginning after SCS implantation were noted by Sheu and colleagues,⁹² which resolved only when the unit was explanted. Han and associates⁹³ sighted the brief occurrence of “locked-in syndrome” (i.e., defined as quadriplegia and anarthria), thought to indicate a conversion disorder following implantation SCS. Parisod, Murray, and Cousins⁹⁴ also reported the appearance of a conversion disorder in a complex regional pain syndrome patient following internalization, which required additional treatment. This latter case highlighted the potential advantages of combining a more comprehensive cognitive behavioral functional restoration program with SCS therapy. Indeed, Molloy and colleagues⁹⁵ demonstrated improved outcomes when both approaches were used. In addition, they did not find an order effect. Of interest was the association of improvement in functional and psychological variables with multidisciplinary therapy and pain ratings with SCS treatment.

A recent review by Celetin, Edwards, and Jamison⁹⁶ of 753 articles relating to pretreatment psychological screening for lumbar surgery and SCS found that only four SCS-related studies (1981, 1995, 1995, 1996) met the criteria for analysis. Psychological factors (e.g., depression, coping, anxiety, somatization, hypochondriasis) were more predictive than other variables but mostly for short-term vs. long-term outcomes. Kumar and Wilson⁹⁷ suggested that patients manifesting psychological factors such as fear avoidance, depression, secondary gain, refusal to be weaned off narcotics, or currently under workers’ compensation should be avoided. Although a number of psychological factors have been identified as potentially associated with outcome of SCS therapy, there is yet to be any consensus as to which have the most reliable and predicable impact (i.e., may be considered predictors).²⁴

One of the more recent tools for examining the patient’s response during the screening trial is neuroimaging. One report by Nagamachi and associates⁹⁸ used SPECT technology before and after SCS. They compared patients reporting a four point or greater reduction in pain on a 0 to 10 scale (good response, GR) to those reporting a decrease of three points or less (poor response, PR). The authors noted a distinctive pattern of activity before and after SCS for the GR and PR groups involving thalamic structures and the anterior cingulate gyrus. Others have documented activation of pain and emotional processing cortical structures in response to occipital nerve stimulation in normal volunteers. As yet there have been no attempts to correlate data collected from the psychological evaluation with that of neuroimaging. Whether or not neuroimaging becomes a reliable and practical tool for evaluating patient response to SCS trail screening remains to be seen. The possibilities at least are intriguing.

Summary

The search for one or more optimal psychological variable(s) or test scores that will predict the outcome of SCS therapy continues. This more-or-less statistical approach assumes the existence of one or more psychological states or variables compatible or not with SCS therapy and which can be detected by the psychological evaluation process. The emergence of an interest in narrative analysis⁹⁹ and other qualitative approaches calls this assumption into question. Models that profess to predict the outcomes of SCS therapy on the basis of pretrial or preimplantation evaluation are, at least in part, predicated on the assumption that the circumstances present at the time of evaluation (a) will persist; or (b) if changed, will not alter the predicted relation; or (c) if altered, the effect will not be clinically or statistically meaningful (i.e., not strong enough to invalidate the prediction). A possible alternative to attempts to predict outcomes is to emphasize developing a detailed description of the particular patient, pain, and therapeutic variables thought to influence the outcome and create a somewhat dynamic therapeutic algorithm that is adjusted according to the patient circumstances and response to treatment.

Each clinician is likely to have his or her favorite tests. For the efforts to distill meaningful information from the evaluation and treatment process to advance, there needs to be some standardization of the process, if not the particulars. Doleys²⁴ proposed the following process suggestions. (1) Well-known and validated tests should be used. (2) Ideally the test(s) should have validity scale(s) or some mechanism for detecting dissimilation. (3) Tests should be used in the context of an overall evaluation, including clinical interview. (4) The assessment should be done by a knowledgeable and experienced, preferably doctoral level, provider. (5) The evaluator should have contact with the patient, or at least the outcome data, from the preimplant trial and follow up to determine the correlation between the evaluation and outcome. (6) Screening tests should be readministered on follow-up. (7) Both disease-specific and generic measures should be obtained.

Dr. Timothy Deer kindly invited me to give a talk entitled, “The psychology of implantable devices: avoiding mental land-mines and achieving great outcomes,” at the 2010 American Academy of Pain Medicine annual meeting. The following were presented as recommendations, albeit not necessarily experimentally verified but a result of 25 years’ experience with implantable devices. First, follow the 3 Ps: pain pattern, pain pathology, and psychological status (i.e., stick to the basics, especially new implanters or less experienced clinicians). There is significant evidence to support the use of SCS in patients with extremity pain, with or without back pain (pain pattern), having a neuropathic component (pain pathology), and presenting as psychologically intact and stable (psychological status). That is not to say that exploring new applications should be avoided, but it should be undertaken by those few with extraordinary skills and technical and clinical support. Second, at least moderately stringent criteria should be applied when evaluating the screening trial. Those outlined by North and Shipley²³ as a result of their review of the literature seem very appropriate, especially continued pain relief despite provocative physical activity. Finally, consistent with the growing emphasis of chronic pain as a disease, regular follow-up visits to reinforce/stimulate psychological/behavioral changes should be required. More often then not, patients are essentially ignored until there is a complication or mechanical failure.

It would be difficult to find a publication from the United States, Canada, Europe, or Australia relating to the use of SCS with chronic pain that does not call for its use in the context of a multidisciplinary framework.^17,23,97 Unfortunately, with the possible exception of very few works such as that of Molloy and colleagues,⁹⁵ which examined the effects of combining an intensive cognitive-behavioral program with SCS therapy, it has largely been more rhetoric then reality. Enormous amounts of time and money have been devoted, and justifiably so, to developing new and more advanced technology (i.e., building a better mouse trap). Ironically, especially when considered in the context of the relative lack of emphasis on regular postimplantation follow-up visits, these efforts encourage the patient to interpret success or failure as beyond their influence and solely dependent on the procedure and technology. This is precisely the message we do not want to convey. I wonder what would happen if the technology companies devoted even 1% of their research and development funds to examining how to make what works work better by supporting multidisciplinary research. A simple beginning point would be the random assignment of SCS patients to X number of postimplantation cognitive-behavioral therapy sessions or usual care. Hopefully the profession has not already arrived at a point where the fascination for gadgetry is betraying sound sense.