This statement is published as:

Integration of Behavioral and Relaxation Approaches into the Treatment of Chronic Pain and Insomnia. NIH Technol Assess Statement 1995 Oct 16-18:1-34

For making bibliographic reference to technology assessment conference statement no. 17 in electronic form displayed here, it is recommended that the following format be used: Integration of Behavioral and Relaxation Approaches into the Treatment of Chronic Pain and Insomnia. NIH Technol Statement Online 1995 Oct 16-18 [cited year month day], 1-34.

---

• Abstract
• Introduction
• What Behavioral and Relaxation Approaches Are Used for Conditions Such as Chronic Pain and Insomnia?
• How Successful Are These Approaches?
• How Do These Approaches Work?
• Are There Barriers To the Appropriate Integration of These Approaches Into Health Care?
• What Are the Significant Issues for Future Research and Applications?
• Technology Assessment Panel
• Speakers
• Planning Committee
• Conference Sponsors
• Conference Cosponsors
• Bibliography
• About the NIH Consensus Development Program
• Statement Availability

How Successful Are These Approaches?

>

Pain

A plethora of studies using a range of behavioral and relaxation approaches to treat chronic pain is reported in the literature. The measures of success reported in these studies depend on the rigor of the research design, the population studied, the length of followup, and the outcome measures identified. As the number of well-designed studies using a variety of behavioral and relaxation techniques grows, the use of meta-analysis as a means of demonstrating overall effectiveness will increase.

One carefully analyzed review of studies on chronic pain, including cancer pain, was prepared under the auspices of the U.S. Agency for Health Care Policy and Research (AHCPR) in 1990. A great strength of the report was the careful categorization of the evidential basis of each intervention. The categorization was based on design of the studies and consistency of findings among the studies. These properties led to the development of a 4-point scale that ranked the evidence as strong, moderate, fair, or weak; this scale was used by the panel to evaluate the AHCPR studies.

Evaluation of behavioral and relaxation interventions for chronic pain reduction in adults found the following:

• Relaxation: The evidence is strong for the effectiveness of this class of techniques in reducing chronic pain in a variety of medical conditions.
• Hypnosis: The evidence supporting the effectiveness of hypnosis in alleviating chronic pain associated with cancer seems strong. In addition, the panel was presented with other data suggesting the effectiveness of hypnosis in other chronic pain conditions, which include irritable bowel syndrome, oral mucositis, temporomandibular disorders, and tension headaches.
• CBT: The evidence was moderate for the usefulness of CBT in chronic pain. In addition, a series of eight well-designed studies found CBT superior to placebo and to routine care for alleviating low back pain and both rheumatoid arthritis and osteoarthritis-associated pain, but inferior to hypnosis for oral mucositis and to EMG BF for tension headache.
• BF: The evidence is moderate for the effectiveness of BF in relieving many types of chronic pain. Data were also reviewed showing EMG BF to be more effective than psychological placebo for tension headache but equivalent in results to relaxation. For migraine headache, BF is better than relaxation therapy and better than no treatment, but superiority to psychological placebo is less clear.
• Multimodal Treatment: Several meta-analyses examined the effectiveness of multimodal treatments in clinical settings. The results of these studies indicate a consistent positive effect of these programs on several categories of regional pain. Back and neck pain, dental or facial pain, joint pain, and migraine headaches have all been treated effectively.

Although relatively good evidence exists for the efficacy of several behavioral and relaxation interventions in the treatment of chronic pain, the data are insufficient to conclude that one technique is usually more effective than another for a given condition. For any given individual patient, however, one approach may indeed be more appropriate than another.

Insomnia

Behavioral treatments produce improvements in some aspects of sleep, the most pronounced of which are for sleep latency and time awake after sleep onset. Relaxation and BF were both found to be effective in alleviating insomnia. Cognitive forms of relaxation such as meditation were slightly better than somatic forms of relaxation such as PMR. Sleep restriction, stimulus control, and multimodal treatment were the three most effective treatments in reducing insomnia. No data were presented or reviewed on the effectiveness of CBT or hypnosis. Improvements seen at treatment completion were maintained at followups averaging 6 months in duration. Although these effects are statistically significant, it is questionable whether the magnitude of the improvements in sleep onset and total sleep time are clinically meaningful. It is possible that a patient-by- patient analysis might show that the effects were clinically valuable for a special set of patients, as some studies suggest that patients who are readily hypnotized benefited much more from certain treatments than other patients did. No data were available on the effects of these improvements on patient self- assessment of quality of life.

To adequately evaluate the relative success of different treatment modalities for insomnia, two major issues need to be addressed. First, valid objective measures of insomnia are needed. Some investigators rely on self-reports by patients, whereas others believe that insomnia must be documented electrophysiologically. Second, what constitutes a therapeutic outcome should be determined. Some investigators use time until sleep onset, number of awakenings, and total sleep time as outcome measures, whereas others believe that impairment in daytime functioning is perhaps another important outcome measure. Both of these issues require resolution so that research in the field can move forward.

Critique

Several cautions must be considered threats to the internal and external validity of the study results. The following problems pertain to internal validity: (1) full and adequate comparability among treatment contrast groups may be absent; (2) the sample sizes are sometimes small, lessening the ability to detect differences in efficacy; (3) complete blinding, which would be ideal, is compromised by patient and clinician awareness of the treatment; (4) the treatments may not be well described, and adequate procedures for standardization such as therapy manuals, therapist training, and reliable competency and integrity assessments have not always been carried out; and (5) a potential publication bias, in which authors exclude studies with small effects and negative results, is of concern in a field characterized by studies with small numbers of patients.

With regard to the ability to generalize the findings of these investigations, the following considerations are important:

• The patients participating in these studies are usually not cognitively impaired. They must be capable not only of participating in the study treatments but also of fulfilling all the requirements of participating in the study protocol.
• The therapists must be adequately trained to competently conduct the therapy.
• The cultural context in which the treatment is conducted may alter its acceptability and effectiveness.

In summary, this literature offers substantial promise and suggests a need for prompt translation into programs of health care delivery. At the same time, the state of the art of the methodology in the field of behavioral and relaxation interventions indicates a need for thoughtful interpretation of these findings. It should be noted that similar criticisms can be made of many conventional medical procedures.

How Do These Approaches Work?

The mechanism of action of behavioral and relaxation approaches can be considered at two levels: (1) determining how the procedure works to reduce cognitive and physiological arousal and to promote the most appropriate behavioral response and (2) identifying effects at more basic levels of functional anatomy, neurotransmitter and other biochemical activity, and circadian rhythms. The exact biological actions are generally unknown.

Pain

There appear to be two pain transmission circuits. Some data suggest that a spinal cord-thalamic-frontal cortex-anterior cingulate pathway plays a role in the subjective psychological and physiological responses to pain, whereas a spinal cord- thalamic-somatosensory cortex pathway plays a role in pain sensation. A descending pathway involving the periaqueductal gray region modulates pain signals (pain modulation circuit). This system can augment or inhibit pain transmission at the level of the dorsal spinal cord. Endogenous opioids are particularly concentrated in this pathway. At the level of the spinal cord, serotonin and norepinephrine appear to play important roles.

Relaxation techniques as a group generally alter sympathetic activity as indicated by decreases in oxygen consumption, respiratory and heart rate, and blood pressure. Increased electroencephalographic slow wave activity has also been reported. Although the mechanism for the decrease in sympathetic activity is unclear, one may infer that decreased arousal (due to alterations in catecholamines or other neurochemical systems) plays a key role.

Hypnosis, in part because of its capacity for evoking intense relaxation, has been reported to reduce several types of pain (e.g., lower back and burn pain). Hypnosis does not appear to influence endorphin production, and its role in the production of catecholamines is not known.

Hypnosis has been hypothesized to block pain from entering consciousness by activating the frontal-limbic attention system to inhibit pain impulse transmission from thalamic to cortical structures. Similarly, other CBT may decrease transmission through this pathway. Moreover, the overlap in brain regions involved in pain modulation and anxiety suggests a possible role for CBT approaches affecting this area of function, although data are still evolving.

CBT also appears to exert a number of other effects that could alter pain intensity. Depression and anxiety increase subjective complaints of pain, and cognitive-behavioral approaches are well documented for decreasing these affective states. In addition, these types of techniques may alter expectation, which also plays a key role in subjective experiences of pain intensity. They also may augment analgesic responses through behavioral conditioning. Finally, these techniques help patients enhance their sense of self control over their illness enabling them to be less helpless and better able to deal with pain sensations.

Insomnia

A cognitive-behavioral model for insomnia (see Figure 1) elucidates the interaction of insomnia with emotional, cognitive, and physiologic arousal; dysfunctional conditions, such as worry over sleep; maladaptive habits (e.g., excessive time in bed and daytime napping); and the consequences of insomnia (e.g., fatigue and impairment in performance of activities).

In the treatment of insomnia, relaxation techniques have been used to reduce cognitive and physiological arousal and thus assist the induction of sleep as well as decrease awakenings during sleep.

Relaxation is also likely to influence decreased activity in the entire sympathetic system, permitting a more rapid and effective "deafferentation" at sleep onset at the level of the thalamus. Relaxation may also enhance parasympathetic activity, which in turn will further decrease autonomic tone. In addition, it has been suggested that alterations in cytokine activity (immune system) may play a role in insomnia or in response to treatment.

Cognitive approaches may decrease arousal and dysfunctional beliefs and thus improve sleep. Behavioral techniques including sleep restriction and stimulus control can be helpful in reducing physiologic arousal, reversing poor sleep habits, and shifting circadian rhythms. These effects appear to involve both cortical structures and deep nuclei (e.g., locus ceruleus and suprachiasmatic nucleus).

Knowing the mechanisms of action would reinforce and expand use of behavioral and relaxation techniques, but incorporation of these approaches into the treatment of chronic pain and insomnia can proceed on the basis of clinical efficacy, as has occurred with adoption of other practices and products before their mode of action was completely delineated.

Are There Barriers To the Appropriate Integration of These Approaches Into Health Care?

One barrier to the integration of behavioral and relaxation techniques in standard medical care has been the emphasis solely on the biomedical model as the basis of medical education. The biomedical model defines disease in anatomic and pathophysiologic terms. Expansion to a biopsychosocial model would increase emphasis on a patient's experience of disease and balance the anatomic/physiologic needs of patients with their psychosocial needs.

For example, of six factors identified to correlate with treatment failures of low back pain, all are psychosocial. Integration of behavioral and relaxation therapies with conventional medical procedures is necessary for the successful treatment of such conditions. Similarly, the importance of a comprehensive evaluation of a patient is emphasized in the field of insomnia where failure to identify a condition such as sleep apnea will result in inappropriate application of a behavioral therapy. Therapy should be matched to the illness and to the patient.

Integration of psychosocial issues with conventional medical approaches will necessitate the application of new methodologies to assess the success or failure of the interventions. Therefore, additional barriers to integration include lack of standardization of outcome measures, lack of standardization or agreement on what constitutes successful outcome, and lack of consensus on what constitutes appropriate followup. Methodologies appropriate for the evaluation of drugs may not be adequate for the evaluation of some psychosocial interventions, especially those involving patient experience and quality of life. Psychosocial research studies must maintain the high quality of those methods that have been painstakingly developed over the last few decades. Agreement needs to be reached for standards governing the demonstration of efficacy for psychosocial interventions.

Psychosocial interventions are often time intensive, creating potential blocks to provider and patient acceptance and compliance. Participation in BF training typically includes up to 10-12 sessions of approximately 45 minutes to 1 hour each. In addition, home practice of these techniques is usually required. Thus, patient compliance and both patient and provider willingness to participate in these therapies will have to be addressed. Physicians will have to be educated on the efficacy of these techniques. They must also be willing to educate their patients about the importance and potential benefits of these interventions and to provide encouragement for the patient through the training processes.

Insurance companies provide either a financial incentive or barrier to access of care depending on their willingness to provide reimbursement. Insurance companies have traditionally been reluctant to reimburse for some psychosocial interventions and reimburse others at rates below those for standard medical care. Psychosocial interventions for pain and insomnia should be reimbursed as part of comprehensive medical services at rates comparable to those for other medical care, particularly in view of data supporting their effectiveness and data detailing the costs of failed medical and surgical interventions.

The evidence suggests that sleep disorders are significantly underdiagnosed. The prevalence and possible consequences of insomnia have begun to be documented. There are substantial disparities between patient reports of insomnia and the number of insomnia diagnoses, as well as between the number of prescriptions written for sleep medications and the number of recorded diagnoses of insomnia. Data indicate that insomnia is widespread, but the morbidity and mortality of this condition are not well understood. Without this information, it remains difficult for physicians to gauge how aggressive their intervention should be in the treatment of this disorder. In addition, the efficacy of the behavioral approaches for treating this condition has not been adequately disseminated to the medical community.

Finally, who should be administering these therapies? Problems with credentialing and training have yet to be completely addressed in the field. Although the initial studies have been done by qualified and highly trained practitioners, the question remains as to how this will best translate into delivery of care in the community. Decisions will have to be made about which practitioners are best qualified and most cost-effective to provide these psychosocial interventions.

Next: What Are the Significant Issues for Future Research and Applications?