Institute of the Pennsylvania Hospital and University of Pennsylvania, Philadelphia, Pennsylvania 19139
The effect of psychological factors in both the response to and the recognition of perception of pain has long been recognized. Thus, soldiers in the heat of battle have been unaware of injuries, prize fighters have been known to continue boxing using fractured hands, and many reports are available of individuals able to function despite debilitating injuries until the crisis is safely past. In such instances, a person becomes aware of extreme pain only when he is in a situation where he can afford to do so. Conversely, patients in severe pain often become calm and relatively comfortable when a trusted physician takes charge. A wide variety of factors from placebo to faith healing have proven effective in reducing and even eliminating pain in many instances.
Severe pain is a dramatic and extremely noxious experience of central interest to medicine. Not surprisingly, there is a tendency to dichotomize methods of pain alleviation into those based on physiological mechanisms as opposed to those involving psychological processes.
While such a distinction is useful in evaluating various analgesic techniques, there is a tendency to group all psychological factors together and consider them as a unitary phenomenon. However, it is necessary to differentiate among various psychological mechanisms that may be involved in the potentiation and control of the pain experience much in the same way that it is necessary to differentiate among the vastly different mechanisms that may be involved in the control of pain by physical means. The same level of physical stimulation leads to different pain experience depending upon the psychological significance of the situation. The effect of prior experience has been vividly demonstrated by how different individuals react to dental procedures. Although few individuals enjoy having a cavity repaired, many accept the situation with equanimity, but others experience extreme anxiety with accompanying reports of severe pain. Many factors such as inherent differences in pain perception may be implicated, but much of the variance is a function of specific prior experience.
This chapter will focus on the use of hypnosis in the suppression of pain.
564 PAIN SUPPRESSION BY HYPNOSIS
Here, too, however, it is well to keep in mind that hypnosis itself is a complex phenomenon and consequently it may affect the organism in a variety of different ways. Prior to the discovery of chemical anesthesia, hypnosis had been used to carry out a broad range of surgical procedures. Given modern methods of anesthesia, it is rare that some cannot be employed with safety. Thus hypnosis now has little practical significance as a sole anesthetic. Nevertheless, the fact that hypnosis can serve as the only anesthesia in major surgery has important theoretical implications.
There are some colleagues who try to explain away hypnoanesthesia by stating that hypnosis does not really block pain perception -- the patients really are suffering. These individuals continue silently to tolerate the painful experience in order to please the hypnotist. While such mechanisms are important to consider, they cannot account for the phenomenon of hypnotic analgesia. Thus, unhypnotizable individuals are hardly eager to undergo major surgery without anesthesia even if offered very substantial monetary inducements. Nor is it possible to explain hypnotic analgesia by arguing that the hypnotizable individuals are selected because they have a high pain tolerance. Indeed, the opposite is often the case. Shor (1) carried out a study on hypnotically induced analgesia which involved permitting subjects to set their own tolerance level of electric shock. Interestingly, the highly hypnotizable subjects were far less inclined to tolerate pain when challenged to do so than those individuals who were unable to enter trance. These experimental laboratory findings closely parallel anecdotal reports of dentists which indicate that many of the patients who undergo dental procedures with hypnosis as the sole anesthetic had previously shown a particularly low pain threshold.
There are, of course, a large number of questions about the nature of hypnotically induced anesthesia, many of which are as yet unsolved. Clearly we are not dealing with a process similar to other forms of central anesthesia. One of the aspects which often seems most troublesome relates to the nature of pain itself. Because the stimulus for pain during surgery involves overt tissue damage, it seems to be difficult to conceive of a psychological mechanism preventing the appreciation of the pain sensation. It is well to remember, however, that pain, like other perceptions, remains a subjective event regardless of the physiological mechanism which may be involved. Similarly, many of the psychological responses to a painful stimulus depend upon the psychological meaning that pain has for the patient.
EXPERIMENTAL STUDIES OF HYPNOTIC ANALGESIA
In a major series of studies, Hilgard and his associates (2) have demonstrated lawful effects relating hypnotic depth and the presence or absence of suggested anesthesia to the subject's threshold and tolerance level for
565 PAIN SUPPRESSION BY HYPNOSIS
pain. They have employed both immersion in a water bath at 0°C and ischemic muscle pain.
A number of studies have been carried out seeking to document the physiological differences between the response to stimulation with and without hypnotically induced anesthesia. Although some differences have been reported, these were observed under conditions where subjects were probably relatively anxious when tested without hypnotically induced analgesia. Presumably, hypnosis serves not only to block the response to pain but also to reduce anxiety. In a study (3) where every effort was made to reduce anxiety, painful stimuli produced remarkably little physiological response; consequently, no physiological difference was seen between the physiological response to stimuli which the subject reported as very uncomfortable and those which, thanks to hypnotic anesthesia, he reported as not painful. Although subjects vastly preferred the latter circumstance, this preference was not reflected in a decreased physiological response to the stimulus itself.
Recognizing that hypnosis affects pain perception, we were interested in determining the kind of mechanisms that were involved; in particular, McGlashan, Evans, and Orne (4) sought to differentiate between the effect of hypnotic suggestion on deeply hypnotized subjects and the placebo response. We argued that just as the administration of an active analgesic pill might produce a placebo response to pain perception, the induction of hypnosis might well lead to a similar kind of placebo-like response which is independent of any hypnotic effects as such. Because of the tendency to group all psychological effects together, no previous study had addressed this issue. In an elaborate design we compared the response to ischemic muscle pain between a group of subjects who were able to enter deep trance and a comparison group who, despite repeated efforts to do so, had failed to respond to hypnosis. Both groups were to be treated in an identical manner by an investigator who was unaware of the subjects' past experience with hypnosis, analogous to the double-blind procedures used in psychopharmacology.
This basic approach tends to have one significant drawback: subjects who fail to respond to hypnosis are themselves aware of their own lack of hypnotizability. Consequently, these individuals when used as a comparison group tend to be placed in a psychological situation that has an entirely different meaning from the one in which hypnotizable subjects find themselves. To examine potential effects of a placebo, the patient must first believe in its effectiveness. We sought to correct this design difficulty by a special session where all subjects in the unhypnotizable group were seen by yet another investigator who spent a considerable period of time with the subjects. These subjects were given further experience with hypnotic techniques using a relaxation induction procedure which was different from the subjects' past experience. Care was taken not to test
566 PAIN SUPPRESSION BY HYPNOSIS
the subject's hypnotic responses in any way which would indicate his lack of responsivity. Each subject was treated as though he were responding extremely well to the particular procedures being employed. Anesthesia of the hand was then suggested, and time was taken in making the suggestion vivid and explicit. The anesthesia was tested by giving electric shocks from an inductorium and asking the individual to compare his pain experience in the analgesic hand with that in the normal hand. All subjects experienced a considerable degree of analgesia because, unknown to the subjects, the setting on the transformer was altered appropriately; they, in fact, received significantly less shock to the hand where analgesia had been suggested compared with the levels administered to their normal hand. It should be emphasized that this manipulation was done very carefully and within plausible limits. Consequently, this unsusceptible group of subjects expressed considerable fascination at the success of the analgesia suggestion. A careful postexperimental discussion with our subjects reinforced our opinion that these individuals actually believed that they had been able to respond successfully to the analgesia suggestion. The purpose of this procedure was to create a group of unsusceptible subjects who would have faith in their own ability to respond to hypnotically induced analgesia, but who would not in fact be responsive to hypnosis. This subtle distinction is a very real one which helped to clarify the difference between hypnosis as an active phenomenon and to hypnosis as a placebo phenomenon.
All subjects had a first session during which baseline ischemic pain tolerance was measured. They were required to perform work which was quantified by the length of time pain was tolerated, as well as the amount of water they pumped, while the arterial supply to the arm was occluded by means of a blood pressure cuff inflated well above systolic pressure. Both time and the amount of water pumped were meaningful measures of ischemic muscle pain. Subjects were urged to pump as long as they could and to report pain threshold (where the sensation in the arm was first noticeably painful) and pain tolerance (when they reached the point where the pain was so unbearable that they were unable to go on any longer). After this initial baseline session, all subjects returned and were tested by a blind experimenter in a hypnotic session where, again, relaxation induction was employed and analgesia was induced for the appropriate arm.
The same technique was then employed to evaluate the effect of hypnotically induced analgesia. All subjects were finally asked to return for a third day in the context of a drug study which, it was explained to them, would be used to help establish how effective hypnotic analgesia really was. It was pointed out that a powerful drug would be used to determine the effect of true, physiologically induced analgesia so that we could determine how effective hypnosis was compared to the active drug. It was implicit that we knew how well the drug would work, and we wanted to see how close their hypnotic analgesia could come to this pain-killing drug. The
567 PAIN SUPPRESSION BY HYPNOSIS
experimenter believed himself to be comparing Darvon ® and placebo in a double-blind manner, whereas, in fact, he was comparing placebo with placebo. This particular intervention was important because it made certain that the experimenter thought himself to be giving a potent drug while still allowing us to study the effect of placebo on the total sample.
Figure 1 summarizes the main findings. It should be clear that, in the group that is not hypnotizable, the hypnotic intervention can be seen as measuring the placebo response to hypnosis. This group's response to hypnosis shows a significant increase in pain tolerance and an even greater increase in pain threshold. However, the level of magnitude of this increase is virtually the same as that achieved by placebo in the form of a pill. In contrast, the highly hypnotizable group, although responding to placebo in much the same way as the insusceptibles, performs far better in the hypnotic amnesia condition. Not surprisingly, there is no meaningful association between the likelihood of being hypnotized and the placebo response to medication.
These differences are seen even more clearly in the subjective pain re-
568 PAIN SUPPRESSION BY HYPNOSIS
ports which were obtained during all of these conditions. Among the insusceptible subjects the response to placebo correlates highly with the response to hypnotic amnesia (r = 0.76, p < 0.01). For the highly hypnotizable group, however, the correlation is totally insignificant (0.06). These data support the view that there is a true placebo component associated with hypnosis, and that superimposed upon this component is an additional pain reduction associated with the specific effect of suggestion, provided the subject is responsive to suggestion. Hypnotizable subjects respond to suggestions that they cannot experience pain in much the same way as they respond to any other suggestion of a negative hallucination. In addition, a less substantive relief of pain can be brought about, particularly in insusceptible subjects, by mechanisms analogous to a placebo response. Presumably this takes effect as a function of the expectations arising from the social context of experiencing hypnotic procedures.
SOME CLINICAL IMPLICATIONS FROM THESE OBSERVATIONS
An interesting discrepancy can be seen in the literature on pain suppression by hypnosis between the reports by dentists that more than 90% of their patients obtain significant pain relief with hypnosis as opposed to the relatively rare individual who is able to undergo major surgery with hypnosis as the sole anesthetic. The phenomenon reported in the dental situation seems analogous to glove anesthesia in the laboratory which can be induced even in individuals who have only a moderate degree of hypnotizability. It would appear that this phenomenon involves the anxiety-induction component of the hypnotic induction procedure as well as a congruent change in the subjective criteria of pain. When glove anesthesia is induced, it is generally tested by a pinprick, a pinch, or a mild electric shock. Such stimuli are not particularly painful, but a large proportion of the population describe differences between the affected hand and the normal hand. This relatively large percentage of individuals who respond to suggestions such as glove anesthesia must be contrasted to the much smaller percentage of individuals who are able to respond to hypnotic suggestion with the profound anesthesia required for major surgery. In these latter cases, we seem to be dealing with individuals who are able to experience a negative hallucination for the pain experience itself, and it is well known that negative visual or auditory hallucinations are very difficult hypnotic phenomena, being experienced by about 10 to 15% of the population (5). It should not be surprising, therefore, that the percentage of individuals who can experience negative somesthetic hallucinations allowing them to block even intense painful stimuli is approximately the same as that of individuals able to experience convincing auditory and visual hallucinations.
We would propose that there are two different kinds of mechanisms by which hypnosis affects pain perception, one of which is closely related to
569 PAIN SUPPRESSION BY HYPNOSIS
other forms of negative hallucinations and seems necessary to obtain profound surgical anesthesia. In a relatively small proportion of the population, this can be evoked by hypnotic induction procedures and by appropriate suggestions even under conditions of low stress. Under conditions of intense motivation, on the other hand, many more individuals may be capable of experiencing such negative somesthetic hallucinations. Under some circumstances, this may occur spontaneously (for example, the soldier who is unaware of being shot until after the battle is over).
The failure to perceive stimuli which would otherwise be highly painful can be reliably evoked in suitable individuals with deep hypnosis, and thus can be brought into the laboratory for systematic study. The process in these individuals is likely to be the same as that seen in a much larger proportion of individuals under circumstances of intense motivational stress. In more than 90% of a normal population, however, hypnosis may also serve to alter the pain experience in other ways. These involve another individual who is seen as competent and capable taking responsibility for the situation, the expectation and anticipation of relief, a redefinition of a sensory experience as a strong stimulus rather than pain, and so on. It is possible to conceptualize many of these mechanisms as a response to the placebo component of hypnosis insofar as it does not require the subject to be capable of successfully responding to other hypnotic suggestions (6).
It seems clear that we are grouping together several psychological processes, the effects of which can and should be separated in future research. From the point of view of this discussion, however, it emphasizes the fact that the hypnotic procedure has some effects on the pain experience which have little to do with the subject's ability to respond to hypnotic suggestion. These mechanisms are activated in many other circumstances where, a trusted healer is seen by the patient as performing some activity which he anticipates will bring surcease from suffering.
Once it is recognized that hypnosis may involve at least two clearly distinguishable sets of mechanisms by which pain can be reduced, it becomes possible to understand how, depending upon the kind of pain reduction one is seeking to achieve with hypnosis, it will be effective in either a small or large percentage of the population.
CLINICAL APPLICATION OF HYPNOSIS
In my own experience, hypnosis has been quite effective in the treatment of pain, especially the severe chronic pain associated with terminal cancer. In these instances, the anxiety reduction aspect of hypnosis, the hope of pain relief which a physician provides, and the fact that some positive action is being taken to aid the sufferer have in themselves been major factors. Further, hypnosis helps those patients who choose to deny their illness, an aspect which can be very helpful to these individuals; other patients, how-
570 PAIN SUPPRESSION BY HYPNOSIS
ever, can and do utilize hypnosis without denial. Because of these factors, related more to the situation of hypnosis than to the ability to respond by being hypnotized, even those patients who fail to have profound trance experiences may still gain considerable relief.
When hypnosis is used in the treatment of severe chronic pain, especially in the case of terminal cancer, it should be incorporated in an appropriate overall treatment program. In such a context, every effort should be made to meet the patient's psychological needs. By allowing the patient and those close to him to focus on pain relief as a goal, activities may be initiated which will lead to realistically attainable improvement in the patient's condition. Usually by the time a patient of this kind is seen, his physician has become understandably discouraged with the prognosis and feels powerless and uncomfortable in dealing with either the patient or his family. Similarly, the family finds it increasingly difficult to interact with the patient. In this type of context one can still focus on attainable objectives such as increased appetite with some weight gain, a decreased need for narcotics, an increased ability to function, and so on. The previously bedridden patient who is successfully able to wash himself in the bathroom sees it as significant improvement. The ability to enjoy meals, read a book, watch a television program or again carry out any other previously abandoned activity should be encouraged and accepted as important in the patient's coping with his physical problems.
In the context of trying to help the patient deal with terminal cancer, formal hypnotic induction as well as self-hypnosis using autogenic training exercises are useful. I have also successfully used massive doses of phenothiazines, initially, to potentiate narcotics; the patient then is in a position to permit reduction of the narcotics and eventually to do without them. Patients with severe pain can often tolerate remarkably high doses of phenothiazines -- up to 2000 mg a day -- and are able to function quite well at these levels. However, the patient must be followed with great care because if the cause of the pain stimuli is somehow eliminated -- by radiation or chemotherapy -- the tolerance for phenothiazines drops precipitously, and unless the dosage is immediately reduced, the patient becomes somnolent.
Following this general treatment strategy, if the patient is on narcotics and is able to reduce the amount gradually and finally eliminate his dependence on these drugs, it is dealt with as a mastery experience, thus permitting the patient the feeling of accomplishment. In the management of these situations, continuing and frequent contact is essential; brief telephone conversations can play an important role in reducing the number of required visits.
Although the approach can only be sketched in the context of this paper, it should be clear that the hypnotherapeutic management of the pain of terminal cancer involves a great many aspects, going well beyond any specific hypnotic effects. Thus, it is probably more important to break down the feeling of helplessness of both the patient and those around him and allow
571 PAIN SUPPRESSION BY HYPNOSIS
them to focus on positive improvements in functioning rather than the inevitable physical decline. It is, of course, important not to raise false hopes but rather to accentuate what the patient is able to achieve and to provide those around him with a role in ameliorating the discomforts and suffering.
COMPLICATIONS IN TREATMENT
Thus far the use of hypnosis in the amelioration and suppression of pain with clear organic causation has been discussed. Understandably there is a tendency to assume that because hypnosis involves psychological mechanisms it would be particularly useful in the treatment of functional pain.
However, there is a real paradox: not only is hypnosis frequently ineffective in ameliorating functional pain, but, worse yet, it can lead to serious complications. Conversely, when the use of hypnosis is limited to the suppression of pain with a clear organic etiology, it is remarkably safe.
In general, the complications seen with the use of hypnosis can best be understood by determining the functional significance of symptoms or relationships threatened by therapy. The suppression of functional pain -- particularly low back pain -- should be approached with considerable caution since it has often become an important coping mechanism. To the extent that functional pain serves an important need, it is necessary for the patient to develop alternative mechanisms or alternative symptoms. Management by psychological methods other than hypnosis is generally desirable when it is clear that the pain is used in the service of significant interpersonal needs.
On the other hand, pain with a clear organic basis, especially when it is time limited or caused by a treatment procedure, can usually safely be suppressed, even chronic pain associated with malignancies, shingles, tic douloureux, and so on; that is, where the pain itself has not begun to serve powerful psychological needs, it can be appropriately treated by hypnotic suggestion. Under circumstances such as these, even when the patient uses his pain in an interpersonal fashion or is rewarded for hurting as in the form of receiving insurance payments, the risk is relatively small provided the etiology is reasonably clear.
One final point deserves emphasis. Regardless of the degree of benefit which a patient may receive from the hypnotic suppression of organic pain, it is important to keep in mind that the pain experience continues at some deeper level of awareness. Thus, hypnoanesthesia should never be confused with the effects of a chemical anesthesia. Even though the patient reports feeling comfortable, may look comfortable, and will subsequently insist that he was comfortable at the time of surgery, there remains an appreciation of pain at another level. Thus, if he is asked to engage in automatic writing, or any other dissociative procedures are employed, the patient who asserts he is comfortable on the one hand will write "It hurts!" on the other. This, as
572 PAIN SUPPRESSION BY HYPNOSIS
well as many of the other paradoxes associated with hypnosis, ought not to lead us to dismiss the phenomenon either as trivial or as unscientific. Rather, it should cause us to question many of our more general assumptions about the nature of pain and what constitutes a normal response to pain.
Hopefully, some of the methods that have begun to explicate the nature of hypnosis may also prove useful in understanding such diverse phenomena as the control of pain in religious rites, in yoga, in acupuncture, in extreme stress, and so on. In each instance, a variety of physiological and psychological mechanisms are involved, and science will be served neither by denying the existence of puzzling events nor by offering glib or simplistic explanations. It is not that these phenomena are the same as hypnosis but rather that some aspects involve related mechanisms. Whatever the differences between these diverse events, it is likely that the kinds of research needed to explain them will be similar to that necessary to understand hypnosis. However, as interesting and important as these esoteric phenomena may be, in the final analysis their greatest significance will probably be in helping to clarify important aspects of everyday events in man's continuing struggle to cope with the debilitation, the fear, and the anguish associated with pain.
The substantive studies upon which this report is based were supported in part by U.S. Public Health Service grant #MH 19156-02 from the National Institute of Mental Health.
I particularly wish to thank Frederick J. Evans and Emily Carota Orne for their helpful comments and suggestions during the preparation of this report.
1. Shor, R. E., Int. J. Clin. Exp. Hypnosis, 12, 258 (1964).
2. Hilgard, E. R., Amer. Psychol., 24, 103 (1969).
3. Shor, R. E., Int. J. Clin. Exp. Hypnosis, 10, 183 (1962).
4. McGlashan, T. H., Evans, F. J., and Orne, M. T., Psychosom. Med., 31, 227 (1969).
5. Orne, M. T., Hallucinations, p. 211. Grune & Stratton, New York (1962).
6. Evans, F. J. Proc. 77th Ann. Con v. APA, 889 (1969).
Figure 1 (p. 567) (from McGlashan, T.H., Evans, F.J., & Orne, M.T. The naure of hypnotic analgesia and placebo response to experimental pain. Psychosomatic Medicine, 1969, 31, 227-246.). © 1969 by American Psychosomatic Society, Inc. is reproduced here with the kind permission of Lippincott Williams & Wilkins ©.