More than 45 years ago, Shealy and colleagues introduced the concept of spinal cord stimulation (SCS) as a means of electrically inhibiting pain that was consonant with Melzack and Wall’s gate-control theory of pain.^1–3 SCS, or neuromodulation, is now a widely used technique that delivers pulsed electrical signals, principally to the dorsal column of the spinal cord, for indications that include failed back surgery syndrome, traumatic nerve injury, postherpetic neuralgia, complex regional pain syndrome, refractory angina pectoris, peripheral vascular disease, neuropathic pain, and visceral pain.^4,5 The technique is minimally invasive and reversible; electrodes can frequently be placed percutaneously under local anesthesia during outpatient surgery; and, unlike more invasive surgical approaches, it does not ablate pain pathways or alter anatomy.^6,7 Although the exact mechanism of action is variously described, it is generally held that pain reduction is achieved by inhibiting the conduction of primary neural pathways through the stimulation of large nerve fibers that override the transmission of smaller nerve fibers more directly involved in pain sensation.^6,8–10

When indicated, SCS is generally a safe means of ameliorating chronic and otherwise poorly tractable pain and, when successful, can reduce patients’ dependence on medication—including opioid medication—while returning a fair degree of mobility and quality of life. It is no wonder, then, that the technique has proliferated in recent years to annual estimates in the 10,000 to 20,000 range worldwide for permanent implantation.^4,6,10,11 In general, a trial of SCS is considered successful if the patient reports at least a 50% reduction in pain in the affected area,¹² with initial success rates for achieving this threshold varying widely, from as low as 20% to as high as 80%, depending on the group studied, the particular pain syndrome targeted, the hospital or clinic reporting, and the patient selection criteria used.^4,13,14

Despite ongoing improvements in the technology of SCS and improved discrimination regarding the diagnoses most likely to respond favorably to the therapy, some studies suggest a significant loss of analgesia in 25% to 50% of patients within 12 to 24 months of implantation.^15–17 Although some of the variation can be attributed to operational factors—lead migration and erosion, electrical complications and malfunctions, and clinical misjudgment^10,18—an increasing amount of attention has been directed toward psychological variables, especially as these may be revealed and addressed through screening procedures during the patient selection process.^{4,6,8,10,11,14,19–22}

There is an ever-accumulating body of evidence that psychosocial variables are among the most predictive factors in the outcome of medical interventions and especially the outcome of invasive procedures, such as spinal surgeries.²³ Given the high variability in long-term treatment outcome for SCS, it certainly makes sense—from the perspectives of patients, providers, health care carriers, and industry—to try to narrow the field as accurately as possible to candidates for implantation who are best suited and most likely to benefit from the therapy. For this reason, many carriers, as well as many providers, independently, require psychological evaluation for their patients before proceeding with a trial of SCS.^4,8

At the Arnold-Warfield Pain Center at Beth Israel Deaconess Medical Center in Boston, Massachusetts, psychological evaluation of the candidate for spinal cord stimulator implantation involves assessment of the patient across three broad dimensions: (1) will his or her intellectual and cognitive capacity prove an obstacle to mastering the device and its ongoing operation; (2) is the patient motivated to collaborate in the management of his or her own pain; and, (3) is there any psychopathology that may inhibit or impair the patient’s postimplant adjustment to the device and its ongoing performance? We will look at the process through which these dimensions are examined in further detail, but, first, some consideration of the development and evolution of psychological selection criteria may prove illuminating.

BACKGROUND OF SELECTION CRITERIA

Shealy, who pioneered electrical neuromodulation for pain management, was also the first to suggest that psychological factors may affect the outcome of a patient’s experience.²⁴ He observed that personality variables appeared to be at least as salient as technical and physiologic issues, where outcome was concerned, and began administering the Minnesota Multiphasic Personality Inventory (MMPI-2) to his patients.^19,20,24 He noted that elevations on the first three clinical scales of the MMPI-2—Hypochondriasis (Hs), Depression (D), and Hysteria (Hy)—are common to many patients with chronic pain and did not appear to dispose them toward a poor outcome but that more florid elevations on additional scales may suggest poor stability. His short list of selection criteria included (1) emotional stability, while recognizing that problems with adjustment are nevertheless understandable among patients with chronic pain; (2) elevation of the depression scale (D) within an appropriate range, indicative of an expectable depressive reaction to the development of chronic pain and subsequent changes in circumstance; and (3) cooperation with a rehabilitation program, suggesting that motivation and collaboration are indicative of emotional stability and maturity.^19,20,24

Shealy was well aware that most patients with chronic pain are likely to present with symptoms of depression and anxiety, a certain preoccupation with their physical problems and pain, and difficulties adjusting to the consequences of their pain. From the outset of our application of psychological selection criteria to choosing candidates for neuromodulation, then, it was less a question of who is an ideal fit than who can potentially benefit and how can we prepare them to succeed as well as possible. Most candidates will exhibit a combination of depression and anxiety, somatic preoccupation, and reactivity to stress. As Doleys has rightly pointed out, psychological factors can be mediators, modulators, or maintainers of pain, and they can certainly vary in magnitude and the extent to which they may complicate a particular patient’s pain picture.^19,25,26 The question is this: What is the degree and complexity of the psychopathology? Just as important, can we address and remediate these problems psychotherapeutically, psychopharmacologically, and through other psychologically indicated multidisciplinary interventions?

Approach Through Exclusion Criteria

In the years following Shealy’s initial suggestion of psychological selection criteria, there was considerable disagreement regarding the utility—in some cases, discussion regarding the complicated ethics—of evaluating patients for neuromodulation based on psychological factors.^20,27 Nelson and colleagues²⁰ were among the first to propose a functional list of exclusionary screening criteria based on a synthesis of a careful historical review and their own experience of evaluating candidates for spinal cord stimulator implantation. The list, including nine items, may be summarized as follows:

1. 
   

   Active psychosis

   
   
2. 
   

   Active suicidality

   
   
3. 
   

   Active homicidality

   
   
4. 
   

   Untreated or poorly treated major mood disturbance

   
   
5. 
   

   Somatization disorder or other somatoform disorder

   
   
6. 
   

   Alcohol or other drug dependency, excessive drug-seeking behavior, or uncontrolled escalation of either prescribed or nonprescribed substance use

   
   
7. 
   

   Compensation of litigation resolution, including long-term disability determination, dependent on SCS outcome

   
   
8. 
   

   Lack of appropriate social support

   
   
9. 
   

   Neurobehavioral cognitive deficits sufficient to severely compromise reasoning, judgment, and memory

The first six of these criteria are clearly descriptive of symptoms or clusters of symptoms that would greatly compromise an individual’s chances of making a good psychological adjustment postimplant. The seventh suggests that an individual’s decisions and even his or her subjective experience—at least, in the short run—may be unduly influenced by considerations of livelihood and economics. The eighth recognizes that even the most self-reliant of patients cannot do everything for themselves postsurgically and may not be the best monitors of their own mood and level of functioning. The ninth simply acknowledges that a sophisticated piece of equipment requires a certain degree of cognitive acuity to operate well and that memory and good judgment are essential components of collaborative health care. Individuals with intellectual limitations, such as patients with dementia, for example, pose special challenges to the successful use of a spinal cord stimulator.

Nelson and colleagues²⁰ did recommend that information regarding these criteria be obtained through observation, interviews, and psychological testing, with clinicians from appropriate disciplines being involved, and they suggested a flexible approach to integrating and weighting the data to arrive at a particular judgment regarding each candidate’s fitness to proceed. The screening criteria, although useful as a starting point, however, were in no way standardized or empirically validated, and no particular weights or scores on any measures were suggested as thresholds.

Approach Through Inclusion Criteria

Doleys¹⁹ later took a different approach to assembling a list of screening criteria, outlining 15 characteristics of patients that he believed were associated with positive outcome:

1. 
   

   General psychological stability

   
   
2. 
   

   Effective defensiveness

   
   
3. 
   

   Moderate levels of self-confidence and self-efficacy

   
   
4. 
   

   Realistic concern regarding illness and proposed therapy

   
   
5. 
   

   Mild depression appropriate to the situation

   
   
6. 
   

   General optimism regarding outcome

   
   
7. 
   

   Ability to cope with flare-ups, complications, and side effects appropriately

   
   
8. 
   

   Appropriately educated regarding the procedure and the device

   
   
9. 
   

   Supportive and educated family

   
   
10. 
    

    History of compliance or cooperation

    
    
11. 
    

    Behavior and symptoms consistent with identifiable pathological condition

    
    
12. 
    

    Behavioral or psychological evaluation consistent with symptoms and reported psychosocial status

    
    
13. 
    

    Comprehension of instruction

    
    
14. 
    

    Appropriate expectation by patient and significant other

    
    
15. 
    

    Ability and willingness to tolerate paresthesias

Doleys acknowledged that, as with Nelson’s and others’ attempts to identify salient psychological selection criteria to include or exclude candidates from proceeding with SCS implantation, his own efforts represented a general apprehension of what has emerged from clinical observation and research and did not represent a clear, evidence-based attempt to delineate a template for selection. Rather, as with many other researchers in the field, he was attempting to map the terrain of consideration: What exactly are the psychological traits, states, features, and symptoms we need to be thinking about, as we consider candidates for SCS?

Arguably, Doleys was laying down a broad, general set of guidelines, allowing psychiatric clinicians—and, here, we will take this to mean any specially trained and licensed clinician, including psychiatrists, clinical psychologists, clinical social workers, and psychiatric nurse practitioners—to observe, interview, and take a history; possibly integrate the results of a battery of psychological tests and inventories; and develop a set of conclusions and recommendations. Because no clear algorithm for psychological selection has yet emerged⁴ and no clear approach to psychological screening has yet to demonstrate consistent efficacy in the selection of candidates for neuromodulation,^10,11,18,21 much is left to the clinical acuity and judgment of the individual psychiatric clinician.

Most of Doleys’s inclusion criteria are general enough to allow the psychiatric clinician broad scope in his or her consideration of the patient, and this is likely intended, given that each criterion may encompass a multitude of moderating factors. The suggestion, for example, that a certain degree of depression is expectable, appropriate to the situation of chronic pain, gives the clinician appropriate latitude in the assessment of the candidate’s mood. It is certainly accurate to say that moderate to severe depression has been identified in a number of studies as reducing the efficacy of neuromodulation, but mild to moderate depression may also be reactive to pain and not a significant factor in a patient’s premorbid personality. Additionally, it would be unusual to encounter an individual who does not react with some degree of depressive symptomatology, given the onset of chronic pain and accompanying deterioration in circumstances in his or her life.^4,10,11