Parisod, Eric MD, FMH, FFPMANZCA; Murray, Robin F. PhD; Cousins, Michael J. AM, MD, FANZCA, FRCA, FFPMANZCA, FAChPM
Spinal cord stimulation has been widely used and reported to be effective in complex regional pain syndrome (CRPS) (1,2). However, careful patient selection and adequate evaluation of the patient’s psychosocial history are key to the success of this therapy because psychological factors can limit the effectiveness of this procedure (1,3,4). Thus , all patients should have careful assessment of the relative role of physical, psychological, and environmental factors including evaluation of any psychopathology.
Although the rate of complications of spinal cord stimulation is approximately 50% in some studies, most of these complications have been benign, related to minor displacement of electrodes, and requiring only noninvasive manipulations (5–7). Complications should be described to patients before performing the procedure.
Conversion disorder, classed as a somatoform disorder, and once known as “conversion hysteria,” occurs infrequently. The cardinal feature of a conversion disorder is the loss or alteration of a function. However, the occurrence of the disorder seems related to psychological rather than physical mechanisms. Conversion disorders consist of a single motor or neurological symptom (e.g., blindness or paralysis). Strong psychological evidence is required for this diagnosis. Often the conversion is medically impossible. As an example, hemibody paralysis, in which the paralysis extends from one side of the face to the same side of the body, is biologically impossible, but it conforms to the layperson’s perception of the nervous system. Criteria for conversion disorder described in the Diagnostic and Statistical Manual for Mental Disorders (DSM IV) (8) also include that the symptom or deficit causes clinically significant distress or impairments in function. The exclusion of symptoms that are deliberately feigned, produced by a substance or a culturally sanctioned symptom, is also required. In addition, the symptom or deficit is not limited to pain or sexual dysfunction and does not occur during the course of a somatization disorder.
A computer-generated search for case reports and literature reviews (Medline) found no reports of conversion disorder after implantation of a spinal cord stimulator. This report describes the management of a case in which psychopathology interfered with the outcome of a frequently used procedure for an otherwise well-defined chronic pain condition.
This left-handed, 43-yr-old man was referred to our clinic in 1999 5 yr after a work-related injury in which he hit the ulnar side of his left wrist against the sharp edge of a chair. His presenting complaint was left forearm and hand pain.
The patient had experienced multiple injuries to his left wrist. In 1976, he fractured both wrists in a motorbike accident. During the next 2 yr, he required two operations for arthrodesis of his left wrist and additional surgery to release the ulnar nerve. His arthrodesis did not initially result in either pain or neurological deficit. In 1984, he fell while roller-skating, fracturing the lower end of the left ulnar bone, which required surgical excision. Surgery left him with increased mobility and no further symptoms. In 1994, the patient, who was a biomedical engineer, suffered a blow to the left wrist when using a spanner at work. From the time of his fourth injury (1994), his problems with pain began. Numerous investigations, including nerve conduction study, computed tomography scan, and ultrasound studies, revealed no specific cause for his continuing pain.
The patient perceived pain described as constant, sharp, stabbing, and burning on the ulnar side of the left wrist consistent with an ulnar sensory distribution in the hand and medial aspect of the left arm, as well as the left shoulder. Pain disturbed his sleep, was exacerbated when he used his hand, and improved with rest, transcutaneous electrical nerve stimulation, and medication. There were also some dysesthesias over the ulnar distribution of the left hand and occasional swelling over his left hand and forearm. At the time of first assessment, the patient was taking dextropropoxyphene napsylate 300 mg, valproic acid 500 mg, and gabapentin 1200 mg/d.
On examination, he was unable to flex or extend his left wrist and had decreased range of pronation and supination of the left forearm without significant wasting of the intrinsic musculature of his left hand. Power in the left arm was essentially normal apart from slight weakness of the flexion/extension/abduction of his left little finger. Upper limb reflexes were symmetrical. Poor peripheral circulation in his left hand with clinically colder temperature was noted; however, there was no edema. Other features included a thoracic scoliosis concave to the left, depressed posture of the left scapula and shoulder and poor muscle tone around the left deltoid and supraspinatus muscles.
The patient had diet-controlled type 2 diabetes mellitus and gastroesophageal reflux. He was a nonsmoker and nondrinker and the adopted son of scientist parents whom he described as kind. His parents later had a biological son. He felt that he was strikingly different in appearance from his parents and brother and that he did not quite belong. He described the important status of his adopted parents in the scientific community and, when questioned, frequently used medical and scientific terminology, sometimes inappropriately. As an example, he noted that his parents had decided not to conceive children of their own because both were Rh negative (an unlikely reason that scientists would use not to conceive). It seemed important to him to impress the interviewer with the status of his father, noting that when he had died (in a manner similar to the death of Marie Curie), the event had been front-page news. However, even when describing aspects of his life that might have produced emotion, his affect was colorless and bland. He was divorced and had remained out of touch with his former wife and his son, who was now 10 yr old, and spoke of them without emotion. He did not express emotion over difficulties in his work as an electrical engineer, although he noted that there had been a great deal of unhappiness in the work setting during the past 6 mo because of recent staff cutbacks. He had been living with a new partner for less than 1 yr. His partner had “three different kinds of epilepsy,” and was experiencing daily uncontrollable epileptic seizures. The patient brightened somewhat as he spoke of his partner’s illness, yet noted that because of the medications he was taking for his pain, he was unable to engage in sexual intercourse and that this was troubling his partner considerably. Thus, during the interview, his affect was generally inconsistent with the themes he was discussing.
After our initial multidisciplinary assessment, the patient was given the diagnosis of CRPS type II of the left arm (Code 203.X1h) (9). We recommended cognitive behavioral therapy and a trial of the opioid Physeptone (methadone) to assist in tapering his use of opioids. We also discussed the possibility of a trial of a dorsal column stimulator (DCS).
At 6-mo follow-up, the patient had failed to obtain significant improvement in his pain and his ability to function with medication. He had refused to attend the cognitive behavioral pain management program, citing his inability to take time away from work without jeopardizing his job. We considered carefully the patient’s unwillingness to attend a cognitive behavioral pain management program, and after weighing the advantages and disadvantages of proceeding with the implant of a DCS, we decided that the patient’s psychological history and preference not to undergo a cognitive behavioral program should not preclude the surgical procedure. Although we believed that psychological therapy preceding the implant would have been optimal, we felt that his lack of willingness to pursue this option did not justify denial of surgery. Thus, as a result of his lack of pain control, we undertook a trial of DCS. A single array of four electrodes was positioned spanning the C5 vertebral body. Stimulation was continued for 3 days, resulting in a marked reduction in pain and improved circulation to the left hand and permitting a significant improvement of left upper limb function. However, his shoulder pain persisted. It was decided to use a dual four-channel system to provide adequate pain relief for the whole area. The patient was readmitted for a 5-day period 3 mo after the initial trial (the delay was because of the lack of an available bed) for implant of a dual dorsal column electrode system at the C2-5 level. This provided good pain relief over all the painful areas and improved function. Postoperatively, he remained afebrile, and both abdominal and cervical wounds seemed to be healing well. An area of allodynia around the abdominal pocket was noted early on that was thought to be an exaggerated response to postoperative wound pain (possibly neuropathic). The patient was discharged on an antineuropathic drug regimen (tricyclic antidepressant and anticonvulsant) for the abdominal pocket symptoms, with slow release formulation of oxycodone as well as some oxycodone (short-acting formulation) for breakthrough postoperative pain.
We reviewed the patient on the 10th day after the implant for suture removal and started him on clindamycin (known allergy to penicillin and cephalosporin) for an inflamed and tender cervical spine wound. He was afebrile and had no new neurological signs at that time. He presented at the hospital emergency room on the same night feeling unwell and complaining of headache and photophobia. Heart rate and blood pressure were increased. A low-grade temperature of 37.7°C was noted. Neurological examination revealed a moderate weakness involving his entire left lower limb, with decreased sensation to pinprick over the left leg. Reflexes were normal. An urgent brain and cervical spine computed tomography scan and microbiology consult were obtained. No signs of an epidural collection or of meningitis were found. However, IV antibiotic therapy was instituted as a precautionary measure. The next morning, the patient reported some funny feelings in his left leg, some difficulties in passing urine, and tenderness over his cervical wound. On examination, he was found to be afebrile with normal blood pressure and slightly increased heart rate (107 bpm). A staff specialist in neurology who was asked to examine the patient found neither a consistent neurological syndrome nor any deficit in power or reflexes.
Thirty-six hours after admission, the patient felt generally better with less pain in the cervical wound. He was able to walk almost normally. He was afebrile, the wound looked less inflamed, and there was only a minor degree of weakness in the left lower limb. He was continued on the same regimen of antibiotics. Two days after admission, the patient reported a sudden onset of lightheadedness and right arm and leg weakness with right facial numbness but without alteration of consciousness. On neurological examination, he was oriented and normally responsive with normal cranial nerve function, although there was some decrease of sensation to pinprick over the V2 and V3 divisions of his right trigeminal nerve. Tone was symmetrically normal and power was 0 of 5 on the right upper and lower limb and 5 of 5 on the left hemibody. Sensation to pinprick was reduced over the right hemibody with normal sensation on the left. Proprioception was normal on both sides. Right lower limb reflexes were absent with normal reflexes on the left. Urine retention required catheterization. Taken together, these symptoms and signs were not consistent with a spreading CRPS. Cardiopulmonary examination revealed no abnormality. We requested a magnetic resonance image (MRI) of the brain and the spinal cord after turning off the spinal cord stimulator. The MRI showed that spinal cord stimulator leads were well positioned. No epidural collection, thecal enhancement, nor abnormal cord signals were noted. A further neurological review was requested. A second examination performed 2 h later by the neurologist revealed a rapid recovery of normal function of the right arm with some remaining weakness of the right leg and bilateral decrease of knee and ankle reflexes. The neurologist could not explain this presentation on anatomical or pathophysiological grounds and noted some behavioral inconsistencies. He recommended a bolus of steroids (500 mg IV of methylprednisolone) as a precaution to deal with possible implantation evoked edema in the central neuraxis and requested somatosensory evoked potentials (SEPs).
One day later, there was an almost complete resolution of the patient’s lower limb symptoms. He was able to void completely after catheter removal. An initial SEP showed normal symmetrical response in both upper and lower limbs. The study was repeated 4 days later and confirmed the absence of neurological abnormalities. All cultures over the next days remained sterile. Full blood count remained normal with a C reactive protein in the normal range. Blood sugar level was only transiently increased, likely related to the bolus of steroid. A second bolus dose of methylprednisolone 500 mg was given IV. IV antibiotics were halted and he was placed on oral antibiotics for a total of 4 wk. The treating physiotherapist began a graduated stretching and strengthening program. As this program progressed, he reported improved confidence and tolerance of increase in movement.
Psychological review was repeated after the occurrence of the patient’s hemibody paralysis and after medical and surgical specialists found no discernible reason for his symptoms. We noted the patient’s apparent indifference to his symptoms and his statement that even if one side of his body was paralyzed, he could manage with the other. The neurologist also remarked on his apparent insouciance. The psychologist noted that a significant number of stressors had recently occurred in the patient’s life. Within the previous month he had found his biological mother, who was living in another country, and had arranged to travel to meet her, his partner’s seizures remained uncontrolled by medication, and his work situation remained tenuous. He continued to describe these events and his recent paralysis with indifference. He also noted with pride “I am the first person to have had an MRI with an implant in place.”
Several factors were notable in the patient’s presentation. First, he seemed to want to be medically special. He seemed to have little capacity for empathy and to have a preoccupation with medical symptoms to the exclusion of any recognition of feelings or emotional states. His smile while describing symptoms that might normally be expected to inspire anxiety or fear was particularly significant. Thus, his presentation suggested a personality structure characterized by narcissistic traits, which included a lack of empathy, a desire to be special, and a sense of entitlement (10). Second, his acute presentation was suggestive of a conversion reaction. The possibility of a factitious disorder was considered. We were convinced that symptoms were not intentionally produced. External incentives for his behavioral symptoms were thought to be absent because the patient had already obtained financial compensation for his injury. We therefore believed that emotional or psychological factors were likely to underlie his symptoms and were outside his awareness. Conversion symptoms afford a patient’s either primary gain (protection from experiencing painful underlying feelings) or secondary gain (the gratification of receiving concern and support from others) (11). This patient’s apparent inability to experience his own emotional distress, a characteristic of those whose personality is characterized by narcissistic traits, seemed to be the underlying cause for the development of a conversion disorder. Probable primary gains included the attention he received as a special case; secondary gains included the avoidance of his partner’s sexual demands and of a difficult work situation.
Diagnosis according to DSM IV criteria was as follows:
Axis I: 300.11 Conversion Disorder
Axis II: Narcissistic Traits
Axis III: CRPS
Axis IV: Occupational problems, marital discord, and planned first contact with biological mother
Axis V: Global Assessment of Functioning 55 (moderately serious difficulties at work and in marital relationship).
The patient was discharged a week later with no objective neurological abnormalities and a good resolution of the skin rash over his cervical spine wound. His medications included clindamycin 450 mg four times a day, gabapentin 800 mg three times a day, tramadol 100 mg two times a day, amitriptyline 50 mg at night, omeprazole 20 mg two times a day, and Diamicron 80 mg three times a day.
Given this patient’s inability to reduce his pain medication and evidence of psychological factors probably underlying this difficulty, our multidisciplinary team felt that the patient would benefit from an intensive cognitive behavioral pain management program. Psychometric scores generally used to indicate the patient’s psychological status before the program commenced indicated that he was not depressed, that he felt significantly disabled, and that his confidence in his ability to manage his life with chronic pain was low. His focus was intensely somatic, and he continued to describe his transitory right-sided hemibody paralysis as a stroke. During the program, he continued to describe himself as disabled. Both he and his partner were seen for psychological therapy during the cognitive behavioral program. His partner reported considerable relationship distress, which she believed caused more health problems for both. She described the patient as unable to maintain an intimate sexual relationship. Toward the end of the program, the patient had learned skills in relaxation, pacing of activities, and other applied cognitive techniques. He had improved the management of his diabetes, lost weight, and was less somatically focused. All pain medications were ceased, and he had improved on a number of physical measures. Psychometric scale scores showed improvement in depressive symptoms, a marked increase in self-efficacy, and a significant decrease in perceived disability and pain-related distress. We reviewed the patient on three different occasions over the next 6 mo. He consistently increased his level of activity and eventually resumed full-time work. He continued to use psychological strategies to manage his pain and reduced the use of his spinal cord stimulator from 6 h to 2 h per day with no additional pain medication. One year after treatment, the patient remained in full-time work, was taking no medication, and was using the stimulator on average only 1 h/d.
Psychometric assessment was completed before, after, and 1 yr after the cognitive behavioral pain management program. Clinic norms were developed by Nicholas (12) from a database of 3215 patients attending the Royal North Shore Hospital’s Pain Management Department. The following measures were used.
Depression, Anxiety, and Stress Scales (DASS)
The DASS is a 42 item Australian instrument developed with the aim of maximizing differentiation between the constructs of depression, anxiety, and stress while still covering the full range of core symptoms of each of these syndromes. It was developed by Lovibond and Lovibond (13) at the University of New South Wales and consists of 3 14-item self-report scales designed to measure the negative emotional states of depression, anxiety, and stress in a single self-report questionnaire. The Depression scale is similar to the Beck Depression Inventory and to DSM IV Major Depression criteria but with the nondiscriminating items removed. The Anxiety scale is similar to the Beck Anxiety Inventory but is shorter. The content corresponds to all DSM IV Anxiety Disorders, except General Anxiety Disorder, and of these most closely matches Panic Disorder. The internal consistency of the instrument as a whole and for each scale of the DASS was calculated in 1996. For the Depression, Anxiety, and Stress scales, Cronbach’s α equaled 0.96, 0.89, and 0.93, respectively (14). The Stress scale was not used in scores reported in Table 1.
Pain Self-Efficacy Questionnaire (PSEQ)
The PSEQ is a 10-item self-report inventory that measures both the strength and generality of a person’s beliefs regarding his/her ability to accomplish a range of activities despite continuing pain. It was developed by Nicholas in 1989 (15) and is based on Bandura’s (16) concept of self-efficacy, which emphasized persistence in the face of obstacles and aversive experiences. Patients are asked to rate how confident they are that they can perform each of the 10 activities described despite pain by selecting a number on a 7-point scale where zero equals not at all confident and 6 equals completely confident. Scores may range from 0 to 60, with higher scores indicating greater self-efficacy beliefs. In 1996, Gibson and Strong (17), in a study of patients with lower back pain, assessed the PSEQ’s internal reliability (Cronbach’s α) as 0.94.
Roland and Morris Disability Questionnaire (RMDQ)
The RMDQ was originally developed and validated as a self-report instrument to measure physical disability in patients with lower back pain (18). The self-report questionnaire requires patients to check a statement if they think it applies to them today. Scoring involves adding positive responses to the 24 items. Total scores range from 0 (no pain related disability) to 24 (extreme pain related disability). Twenty-five items were initially selected from the Sickness Impact Profile (19) covering a range of normal daily tasks. After modifications were made to differentiate disability caused by pain from disability caused by other forces, the final version of the RMDQ included 24 items. Because this study used a heterogeneous sample of chronic pain patients, modification was made to the wording of statements to reflect pain as a general condition rather than back pain only. This modification of the RMDQ has been previously reported and found valid (20). The psychometric properties of the RMDQ have been well established. It has been found to be reliable, valid, and sensitive to change among patients with chronic lower back pain (21) and among various other chronic pain conditions (20). When compared with other measures of function or disability, it has been judged favorably and is generally considered a useful and valid tool (22).
The gate control theory (23) elaborated in the mid-sixties provided the theoretical basis for spinal cord stimulation therapy. The theory supported the notion that pain is multidimensional and responsive to both internal and external stimuli. Later theoreticians (3,24) also described pain as multidimensional. Thus, comprehensive assessment must take into consideration more than the presence or absence of nociception. The examination of one dimension in isolation will lead to an incomplete assessment (25). An understanding of the concept of suffering and its relation to pain is also important. Suffering is a negative response induced not only by pain but also by other psychological states such as fear, anxiety, stress, and loss of loved objects (26). As stated by Cassel (27), “suffering occurs when the physical and psychological integrity of the person is threatened.”
This patient had a well-defined pathoanatomical cause for his pain (repetitive nerve injury) and observable physical signs supporting the diagnosis of CRPS. Initial multidisciplinary assessment revealed some degree of musculoskeletal deconditioning and, from a psychosocial perspective, an inadequate approach to pacing his activities. He preferred a pharmacological approach to the management of his pain and disability. In retrospect, had we given more emphasis to his psychological needs earlier and continued to involve one of our psychologists or psychiatrists in his continuing treatment, we might have handled this case in a different manner, perhaps preventing, or at least diagnosing more promptly, the basis of his acute neurological presentation.
Although most reviews describe complications of spinal cord stimulation therapy, most complications seem benign and nonpermanent and are related either to misplacement of electrodes or technical failure of percutaneously placed electrodes (1,2,6,7). Burton (28) found neurological deficits related directly to spinal cord stimulator implantation in a series of research studies in which electrode arrays placed by laminectomy caused spinal cord compression. Pineda (29) also found complications clearly related to the surgical technique such as seroma, pseudomeningocele, and pain at the site of surgery.
Turner et al. (6) performed a meta-analysis of research studies on long-term risks and benefits of spinal cord stimulation for patients with failed back surgery syndrome covering the period from January 1966 to June 1994. Thirty-nine case studies were analyzed. Mean follow-up time for these cases was 16 months (range, one to 45 months), with 59% of patients reviewed reporting 50% pain relief. Complications occurred in 42% of patients (range, 20%–75%). These were generally minor with few, if any, complications leading to permanent new neurological loss or death. The infection rate ranged from 0% to 12% (mean, 5%). Most complications were electromechanical, with approximately 30% of patients having one or more complications (range, 0%–75%). When compared with the outcomes and complications reported in the studies referred to, this patient received a satisfactory outcome with good pain relief. His increased temperature after the insertion of the DCS suggested that, as had occurred in a small minority of reported cases, postoperative infection was likely, and it is possible that the patient’s response to infection was related to the development of conversion symptoms.
Only two cases of permanent major neurological complications after spinal cord stimulation, such as permanent paraplegia (5,30), are reported in the medical literature. Both patients had spinal cord stimulator and computerized pump systems implanted simultaneously. Neurological symptoms developed over the following couple of days and remained permanent. In both cases, signs during exploratory surgery were suggestive of adhesive arachnoiditis; these changes were found at the level of the tip of the catheter and below the electrode of the neurostimulator. Microbiology revealed neither infection nor histology suggestive of arachnoidal fibrosis. Recently, there is evidence in some patients of formation of a fibrous tissue mass at the tip of implanted spinal systems of both epidural and intrathecal type (31). To date, these seem to be related to drug infusion not the device material.
Some theorists have hypothesized that the conversion symptom has symbolic meaning to the patient. Although these patients are sometimes found to have histrionic personalities, many do not. Conversion symptoms may begin at any age but usually surface during adolescence or early adulthood. They are generally triggered by an acute psychosocial stressor and arise suddenly, last for several days or weeks, and then stop abruptly. Symptoms occur more frequently in women than in men, with reported ratios varying from 2:1 to 10:1 (8). Symptoms may occur more frequently in people from lower socioeconomic strata and may be more common in subcultures that consider these symptoms expectable rather than bizarre.
Therapy aims to temporarily remove the patients from the situation that has overwhelmed their usual ability to cope, to reassure them that they will soon recover, to minimize any secondary gain that may prolong recovery, and to reinforce alternative coping strategies. For our patient, all these approaches were used acutely, including performing MRI scan and SEPs to demonstrate absent underlying pathology. Alternative forms of therapy include hypnosis, psychodynamic therapy, or psychoanalysis. Given this patient’s rapid recovery when no medical reason for his symptoms could be found, we considered these forms of therapy inappropriate. Indeed the conditions many clinicians would consider important as a precondition for psychodynamic treatment were not present. These conditions are: an awareness of psychic suffering; an interest in the search for self-knowledge; the ability to continue with this type of therapy, which is generally without the immediate gratification of a rapid result; and the willingness of the patient to undergo psychodynamic therapy (32). A group cognitive behavioral program for treatment of chronic pain was available. We believed that this program, which was aimed at teaching behavior that would give this patient the skills he lacked, would have been beneficial to him. At a later time for this patient, the group cognitive behavioral program was also designed to achieve these aims.
Although our patient did not meet DSM-IV criteria for Narcissistic Personality Disorder, narcissistic traits were evident. This patient required a great deal of attention from the medical staff and seemed to enjoy being the professor’s special case. Although we could not assign the conversion reaction to a single definite stressor, there were numerous stressors identified. He fulfilled the criteria for conversion disorder as follows:
1. Presented with motor dysfunction suggesting a neurological condition. The neurological picture was compatible with a hemiplegia; however, inconsistent symptoms were present such as sensory changes including his face on the same side as those in his lower body.
2. Stressors were identified related to his relationship with his partner, her health and his work, his adoption history, and the discovery of his biological mother, although it was difficult to pinpoint a time relationship.
3. Although a factitious disorder was part of the differential diagnosis, we felt that his symptoms were genuinely involuntary and beyond his conscious awareness.
4. The patient was extensively investigated with repeated imaging of his spine and brain as well as nerve conduction studies. No objective findings explained his symptoms.
5. His presentation definitely warranted continuing medical evaluation and would have caused major impairment in social and occupational areas of function if they had continued.
6. The deficit was not limited to dysfunction caused by pain.
We managed this patient suffering from CRPS with a multidisciplinary approach aimed at recognizing and addressing the different dimensions of his pain experience. Although much attention had been afforded to the care of this patient, we were confronted with an unexpected acute presentation thought to be of psychological origin. The knowledge of possible complications resulting from spinal cord stimulation, as well as the invaluable help of other specialists, allowed us to manage this case safely without removal of the implanted DCS. We believe that the patient was determined to have a mechanical fix for his symptoms. With the relief of his pain and the strong recommendation of the medical pain specialists who were treating him, the patient reluctantly began an intensive cognitive behavioral pain management program. Despite his initial reluctance, he completed the program successfully. Without the surgical intervention, he may not have been willing to enter into group therapy. Thus, it is probable that for this patient, both modes of therapy contributed to his improvement. This case provides a good example of the need to assess patients from a biopsychosocial perspective to provide the best possible outcome.
1. Oakley JC, Weiner RL. Spinal cord stimulation for complex regional pain syndrome: a prospective study of 19 patients in two centres. Neuromodulation 1999; 1: 47–50.
2. Kumar K, Nath RK, Toth C. Spinal cord stimulation is effective in the management of reflex sympathetic dystrophy. Neurosurgery 1997; 40: 503–8.
3. Doleys DM, Murray JB, Klapow J, Coleton MI. Behavioural medicine/psychological in the pain patient. In: Ashburn M, Rice L, eds. The management of pain. New York: Churchill Livingstone, 1997.
4. Engel GL. The need for a new medical model: a challenge to bio-medicine. Science 1977; 196: 129–36.
5. North RB, Cutchis PN, Epstein JA, Long DM. Spinal cord compression complicating subarachnoid infusion of morphine: case report and laboratory experience. Neurosurgery 1991; 29: 778–84.
6. Turner JA, Loeser JD, Bell KG. Spinal cord stimulation for chronic low back pain: a systematic literature synthesis. Neurosurgery 1995; 37: 1088–96.
7. Kemler MA, Barendse GAM, Van Cleef M, et al. Spinal cord stimulation in patients with chronic reflex sympathetic dystrophy. N Engl J Med 2000; 343: 618–24.
8. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. Washington, DC: American Psychiatric Association, 1994: 457.
9. Merskey H, Bogduck N. Classification of chronic pain. Seattle, WA: International Association for the Study of Pain, 1994.
10. Kernberg O. Borderline conditions and pathological narcissism. London: Jason Aronson, 1990: 3–47.
11. Maxmen JS, Ward NG. Essential psychopathology and its treatment. 2nd ed. New York: Norton 1995: 293–6.
12. Nicholas M. Development of normative statistics for the psychometric assessment of persisting non-cancer pain patients. Final Report to the Motor Accidents Authority (New South Wales), May 21, 2002.
13. Lovibond SH, Lovibond PF. Manual for the Depression Anxiety Stress Scales. 2nd ed. Sydney: Psychology Foundation, 1995.
14. Brown TA, Chorpita BF, Korotitsch W, Barlow DH. Psychometric properties of the Depression Anxiety Stress Scales (DASS) in clinical samples. Behav Res Ther 1997; 35: 79–89.
15. Asghari A, Nicholas MK. Pain self-efficacy beliefs and pain behaviour: a prospective study. Pain 2001; 94: 85–100.
16. Bandura A. Self-efficacy toward a unifying theory of behavioral change. Psychol Rev 1977; 84: 191–215.
17. Gibson L, Strong J. The reliability and validity of a measure of perceived functional capacity for work in chronic back pain. J Occup Rehabil 1996; 6: 159–75.
18. Roland M, Morris RA. Study of a natural history of back pain. I. Development of a reliable and sensitive measure of disability in low back pain. Spine 1983; 8: 141–4.
19. Bergner M, Bobbitt RA, Carter WB, Gibson BS. The sickness impact profile: development and final revision of a health status measure. Med Care 1981; 19: 787–805.
20. Jensen MP, Storm SE, Turner JA, Romano JM. Validity of Sickness Impact Profile Roland scale as a measure of dysfunction in chronic pain patients. Pain 1992; 50: 157–62.
21. Beurskens AJ, de Vet HC, Koke AJ. Responsiveness of functional status in low back pain: a comparison of different instruments. Pain 1996; 65: 71–6.
22. Beurskens AJ, deVet HC, Koke AJ, et al. Measuring the functional status of patients with low back pain: assessment of the quality of four disease-specific questionnaires. Spine 1995; 20: 1017–28.
23. Melzack R, Wall PD. Pain mechanisms: a new theory. Science 1965; 150: 971–9.
24. Loeser JD. Perspective on pain. In: Turner NP, ed. Proceedings of first world congress on clinical pharmacology and therapeutics. London: Macmillan, 1980: 316–26.
25. Klapow JC, Slatter MA, Patterson TL, et al. Psychological factors discriminate multidimensional clinical groups of chronic back patients. Pain 1995; 62: 349–55.
26. Loeser JD, Melzack R. Pain: an overview. Lancet 1999; 353: 1607–9.
27. Cassel EJ. The nature of suffering and the goals of medicine. N Engl J Med 1982; 306: 639–45.
28. Burton C. Safety and clinical efficacy of spinal cord stimulation. Neurosurgery 1997; 1: 214–5.
29. Pineda A. Complications of dorsal column stimulation. J Neurosurg 1978; 48: 64–8.
30. Aldrete JA, Vascello LA, Ghaly R, Tomlin D. Paraplegia in a patient with an intrathecal catheter and a spinal cord stimulator. Anesthesiology 1994; 81: 1542–5.
31. Coffey RJ, Burchiel K. Inflammatory mass lesions associated with intrathecal drug infusion catheters: report and observations on 41 patients. Neurosurgery 2002; 50: 1–9.
32. McDougall J. Theatres of the body. London: Free Association, 1989: 5–8.