Bioethics 11:3/4 (July/October 1997) Oxford/Boston: Blackwell
Do doctors undertreat pain?
At graduation, some North American medical students repeat the Prayer
of Maimonides "never to forget that the patient is a fellow creature
in pain, not a mere vessel of disease." How could a physician ever forget
that a patient is in pain? Don't physicians confront constant remindersmoans,
groans, winces, and other obvious manifestations of pain? Yes, but it is
those very "reminders," as I shall explain, that provoke at least
two kinds of forgetting common among physiciansone, psychological
and the other, conceptual. The psychological kind of forgetting is
primarily self-protective, but the conceptual kind has deeper roots in the
very definition of modern Medicine as curative and life-preserving. If my
analysis is right, more lecture time on pain and pain relief in medical
schools will do little to correct this "forgetting" of pain. But
there may be better remedies for pain-forgetting, some already at work in
North American medical practices.
Physicians and nurses have various familiar ways to forestall or discount patients' pain reports and requests for more pain relief. In advance of a painful procedure, they minimize the pain that a patient may experience. ('This may sting a bit,' 'You may have some headache for a short time after the lumbar puncture.') Their clinical rationale is to reduce a patient's fear and resistance, as well as the pain both may enhance. But these understatements also teach patients the acceptable linguistic parameters for their subsequent complaints. In general, when patients try to speak of pain, physicians and nurses routinely translate their reports into talk of discomfort and distress or even tenderness, the mere possibility of "distress."
Given clinicians' control of language and general authority, patients may begin to question their own pain-reports. They may begin to wonder, "Am I really in as much pain as I think?"a seeming exception to what philosophers call the "incorrigibility" of pain, that is, the impossibility of being wrong about the existence or severity of one's own pain.
Whatever the effects of clinician understatement on patients, this linguistic discounting helps clinicians forget how much or how often their patients are suffering, and thereby it helps clinicians distance themselves from the pain they continually encounter and often produce in the course of diagnosis and therapy. For such self-protective distancing to succeed, however, there must be clinical rationales that disguise this functionand they abound. Most common are such routine saws as i) Patients with a history of drug abuse are exaggerating their pain in the hope of getting enough drugs for a hospital high; ii) Patients identified by their ethnic affiliations are engaging in "typical Italian histrionics" or "the usual Jewish kvetching;" and, more sweepingly, iii) All patients tend to be "cry babies," regressing toward childhood under the strain of illness and hospital routine.
Such generalizatons and stereotypes may be based on little evidence,
or none at all in the case of particular patients. But given their value
for rationalizing clinicians' self-protective underdescription of patient
pain, evidence is not at issue.
Dismissive stereotypes and subtle self-protection aside, patient pain is subject to a different kind of forgetting due to medical training and professional self-definition of Modern Medicine that inspires it. From the outset students are trained to regard pain relief as a secondary concern. They are taught to regard pain as useful symptom for diagnosing disease and, accordingly, to respond not by relieving but by observing and exploring the pain, even if that involves enhancing it through palpation of soft tissues and manipulation of joints. Hence, immediate or pain relief would be at the sacrifice of clinical information.
Likewise, students learn how helpful pain can be in following the course of a disease, stages of healing, or the efficacy of drug therapy. More seriously, they learn the many ways in which analgesics, especially morphine and opioids, complicate therapeutic protocols. In addition to causing nausea, vomiting, sedation, or constipation, analgesics may depress vital functions (cardiac, respiratory, renal or hepatic) that may already be compromised by disease or other medication. Developing tolerance only makes matters worse, as increasing doses are needed to achieve the same level of relief, with increasing risk of physiological and psychological dependencies. Critics charge that the risk and severity of these side-effects are greatly exaggerated, in our Puritanical, anti-drug culture. Even so, a young, cautious physician need not be a Puritan to try to minimize these complications and risks by limiting analgesics.
In short, from the outset physicians learn to think of pain relief as a complication or hindrance to their diagnostic and therapeutic efforts, not as an integral part of therapy. They early adopt the policy First diagnose and treat; then relieve within limits.
There are, however, exceptions to this implicit policymost notably in the case of patients who are terminally ill or in chronic pain with no discoverable organic causes. But these are "exceptions that prove the rule"the very patients of whom physicians say or think: "I'm afraid there is nothing more we can do for you." There is, of course, something more they could do, namely, to provide palliative care. But, significantly, once they are certain of these diagnoses or prognoses, physicians often relegate that task to hypnotists, acupuncturists, biofeedback specialists, hospice nurses, or other non-physicians. The modern physician's proper work is curing or at least arresting disease, not providing comfort.
Not all physicians, however, transfer patients once they are judged to be terminally ill or suffering from pain without a discoverable organic cause. Some are willing to shift from curative efforts to pain relief, including opioids in high and increasing dosages as needed for full relief. In so doing, however, these physicians often provoke their colleagues' charges that they are overdosing or, even, engaging in unprofessional conduct bordering on homicide.
In their self-defense, physicians so charged may invoke one of several familiar ethical principlesfor example, the Principle of Double Effect ("Even if we foresee that death may result, it is relief of pain not death that we intend"), the Principle of Patient Autonomy ("I providing the care that my patient has competently and freely chosen"), or the Principle of Humane Aid ("I am relieving intolerable pain"). But these replies will not persuade critics whose standards of pain relief derive from a conception of Medicine as essentially curative and life-preserving. For them, knowingly to cause, or even risk the death of a patient for the sake of patient comfort is to forsake the defining goals of modern Medicine. By so doing, physicians forfeit the right to call themselves "doctor," not unlike Jack Kevorkian.
I'll come back shortly to such heated claims about "the Goals of
Medicine." There is, I think, an underlying, more subtle issue, namely,
the appropriate concept of pain. What, I suggest, physicians' training
produces is a new, clinical concept of pain that tends to replace
their prior lay concept of pain. As a result, what physicians in
their training and practice come to forget is this prior, ordinary
concept that most of their patients continue to hold. As much or more than
the psychological "forgetting" of patient pain, it is this forgetting
of patients' concept of pain that sets physicians apart from their "fellow
creatures in pain." .
To appreciate the conceptual character of this change it is necessary first to challenge a common, but simplistic analysis of pain as solely a private sensation. On this view, we know directly only our own pain and must infer the pains of others from their "outward" behavior. Accordingly, what physicians acquire is a capacity for sophisticated inferences. From their observations and explorations, they develop a wider range data on which to base inferences to a detailed descriptions of a patient's pain (its location, severity, pattern), with due allowance for the ways patients may mislead less trained inferers.
It might be said that physicians thereby have a more precise concept of pain than the rest of us, but not a different concept. An analogy might be the logician's more precise conception of validity, sharpened by formal techniques for testing validity of a wider range of argument than the untrained arguer can manage. But I think that the difference between physician and lay concept of pain is more than degrees of precision and sophisticated inference. And it lies not in physicians' better inferences but in their peculiar trained responses to a patient's pain.
Normally we do not infer someone's pain from their behavior, rather, we respond to people's pain the pain manifest in their facial, vocal, and bodily expressions. Pain is indeed a sensation but a sensation that is expressed in these various ways, subject to our respondents. In infants, pain manifestations are initially nonvoluntary. Crying is as natural as the suckling that relieves hunger, and so, too, within a culture, are parental responses. With time, a child's pain-manifestations become more selective. Even before speech, infants modulate their crying, accentuating or suppressing it in the light of the appearance or absence of recognizable relief-givers (and pain-causers). We early learn who will and will not respond, and the circumstances in which no one will respond, and cry accordingly. Our crying becomes largely limited to those situations in which relief is expectableincluding, of course, the relief of crying in private.
In extreme cases, suppression may become virtually total and habitual as with the "warehoused" infants who live in cage-like cribs without responsive attendants. Just so, patients may learn to privatize their painto "suffer in silence," to "keep their complaints to themselves," to "put on a good face" or a "good act." This may be prompted by clinicians' routine verbal discounting of pain-reports mentioned above, or by their routine pseudo-inquiries, "How are we feeling today?"a perfunctory greeting, not a request for information. Or patients, like good soldiers, may not want to trouble their superiors. Or they may wish to avoid further painful investigations that honest revelation of pain would provoke. But, clearly, even such "privatized" pain is response-relative: patients suppress manifestations of pain in order to prevent impatient, or dismissive, or investigative responses of their caretakers.
The techniques that researchers, as well as clinicians, use in investigating patients' pains also reveal the public or social aspect of pain. For example, standard lists of descriptors in pain questionnaires characterize pain as stabbing, burning, pinching, stretching, pulling, wrenching, cutting, drilling, gnawing, scalding, pressing, crushing. That is, we distinguish pains from one another by way of the actions of external agents, human and otherwise, that characteristically cause our pain. Admittedly, some descriptors are more tied to sensation (hot, tingling, dull, throbbing, radiating, itchy), but they a smaller fraction than we would expect if pain were conceptually a solely private sensation.
The social and public aspects of pain are also reflected in the techniques that pain researches and clinicians use to "objectify" pediatric pain. Children are shown a series of cartoon sketches or photographs of children's faces ranked by the severity of the pain they supposedly represent (the so-called "Oucher Scale"). They are asked to match their pain to one of the faces an easy task for most of them. If pain is thought of as essentially private, it is hard to see how they come to be able to do this. Are they reading into the pictures sensations that they have previously correlated with their own facial expressions of pain?with the aid of a mirror? More plausibly, they may learn the representation of pain from the exaggerated faces that their parents and other caretakers make in sympathetic response to their crying, grimacing, and other naturally graded pain manifestations. Or, fascinated from an early age by other children crying, they may learn something about pain gradations by observing caretakers' efforts at comfort. In each of these possibilities, learning would run from public manifestations and social responses to the "inner" sensation, not in the reverse order as the private-sensation theory requires. In sum: pain is a public, social, response-relative concept.
If so, then in learning to substitute one kind of response for another
to patients' manifestations of pain, physicians are acquiring a different
concept of pain. In the presence of physicians who exhibit and subtly impart
their learned clinical concept of pain, some patients may themselves come
in time to take the same distanced curiosity in their pain that their physicians
show, coming to regard their own groans and winces, not as demands for immediate
relief, but as symptoms for assessment. To that extent, they will have themselves
taken on the physicians' clinical concept of pain, even in the midst of
their own pain. But the majority of patients are not so acculturated: their
expressions of pain continue in hope of sympathetic efforts at relief. Hence,
they see doctors and nurses who fail to respond appropriately, according
to this ordinary concept of pain, as insensitive or worse.
To patients who have not become medically acculturated, their physicians and nurses may seem sadistic or callous. Freud thought that medicine attracts people with relatively strong sadistic impulses, but not as a way of acting on these impulses, but as a way of suppressing them through "reaction formation." Admittedly, our advance rescue techniques (CPR, ventilators, open heart surgery, toxic chemotherapies) may provide "undefended" sadists with more opportunities than the physicians of Freud's day enjoyed. And, of course, to the extent that sadism is about power over weaker, dependent people, sadists have the simpler device of stinting on the pain-killers they control.
For the most part, however, I think that physicians' routine practices run counter to standard sadistic practices. As noted earlier, I take clinicians' standard understating of patients' pain reports to be self-protection, not against sadistic impulses but against the pain they witness or inflict in their diagnostic and therapeutic efforts. Sadists also engage in systematic misdescription, but in just the opposite direction: they overstate the pain they intend to inflict or are inflicting on their masochistic partners. Whips, chains, and studded leather hoods are ways of exaggerating in the imagination of both sadists and masochists the amount of pain they actually inflict and endure, respectively. This becomes necessary to avoid the serious injury or death that they might otherwise risk. Their only similarity with most doctors is their commitment to "above all, do no harm."
Callousness is a far more serious worry. If, as I suggest, physicians learn to forget their patients' concept of pain, they will find it easy to ignore their patients' expectations of pain relief. What counter-measures might be taken? Vivid films or stories about patient suffering and physician callousness may help; so, too, physician-patients accounts of their own suffering at the hands of other physicians. Another corrective for callousness might be to require graduating medical students to spend some time as hospital patients. Claiming to have vague symptoms, they would at least undergo some of the painful diagnostic tests that they will routinely impose on their own patients, as well as the hospital delays and indignities that increase patient suffering.
But if I am right about the causes of physicians' "forgetting that their patients are fellow creatures in pain," then the underlying therapy/palliation contrast must be challenged directly. Indeed, a variety of just such scientific and social challenges are underway. Thanks to new guidelines, large conferences, and publicity, physicians are beginning to see how much "information" about morphine and opioid toxicity, tolerance, addiction, and depression of vital functions is myth. Likewise, they are learning that lower doses are needed when patients are allowed to administer their own analgesics at will, especially before the onset of pain. Moreover, research is beginning to show that unrelieved pain has itself deleterious effects on vital functions, for example, on the immune system and hence on healing.
The social challenges to the therapy/palliation contrast are even greater. A new generation of physicians has been trained in midst of diseases for which there have been no available cures (AIDS, Alzheimer's) or for which therapies have proven less curative than believed (coronary artery bypasses, some cancer chemotherapies). Hence, symptomatic relief and palliation had to become the center of their work, not a secondary goal.
Moreover, even when there is "something more" physicians can do to try to cure, or at least arrest a debilitating or degenerative disease, patients or their insurers increasingly are unwilling to "fight to the end." In such cases, palliation or "comfort care" becomes a therapeutic option, or even "the treatment of choice"not an admission of clinical failure or fatigue. Relatedly, patients are asking physicians to collaborate with non-medical (acupuncturists, hypnotists, herbalists). Significantly, such non-medical therapies tend to be less painful than many medical and surgical therapies. Indeed, they may work primarily or solely by reducing the fear, anxiety, and other forms of suffering that enhance patients' pain, whatever the cause. Moreover, these "healers" tend to count pain-relief as a primary goal of therapy, not a secondary concern and, so, retain (by my account) their patients' concept of pain. Less dramatic, but with similar effect, physicians are collaborating more with nurses and nurse-practitioners for whom pain-relief has always been a defining goal of their profession, one that distinguishing them from doctors. .
It remains to be seen whether these social changes, along with more precise
knowledge of pain's harms and analgesia's manageable side effects, will
give pain-relief greater status in clinical training and practice. As the
current professional debate over physician-assisted death shows, the medical
profession's self-definition as curative and life-preserving is tenacious.
But there is reason to believe that the current contrast I have drawn between
physicians' concept of pain and patients' concept of pain will shrink. If
so, then the Prayer of Maimonides may become more than ceremonial and physicians
will more easily remember that their patients are "fellow creatures
in pain, not just vessels of disease". The two forms of conceptual
and psychological forgetting would diminish, together. Were physicians to
remember their patients' concept of pain as demanding relief and act accordingly,
then, of course, they would reduce the amount of pain they would need to
"forget" by self-protective misdescription and dismissive stereotypes.
Departments of Philosophy and Psychiatry, NYU
1. A revised version of a paper read at the panel, "Mismanaging Pain," III World Congress of Bioethics, San Francisco, California, on November 24, 1996.
2. Attributed to the 12th century physician-philosopher Maimonides (Rabbi Moses ben Maimon, or RamBam) but possibly of 18th century origin.
3. Other self-protective euphemisms: surgeons "lose" patients, oncologists detect "growths," infants are born "with problems." Even acronyms and eponyms may play a euphemistic role: 'ALS' and 'Lou Gehrig's disease' seem less dire than the fully descriptive 'amyotrophic lateral sclerosis.'
4. This diagnostic response to pain is caught by the old medical school joke:
Q. "What are the five classical signs of infection?"
A. "Rubor, calor, tumor, dolor -- and clamor." Pain (dolor) and its expression (clamor) are assimilated to redness, heat, and swelling-all signs or symptoms useful for diagnosis of their pathological causes.
5. Ludwig Wittgenstein:"....(Pity, one may say, is a form of conviction that someone else is in pain.)" Philosophical Investigations I, para.287. Readers of Wittgenstein will appreciate that my remarks are variations on his general attack on the view that psychological terms are to be thought of as names for private sensations, rather than as tools whose meaning is given by uses in what he called "forms of life," "the stream of life."
6. For us, parental comforting of a crying baby seems as natural, or spontaneous as the crying itself. Parents, especially mothers, who do not so respond are thought to be abnormally depressed, exhausted, or otherwise distracted. Observers of other cultures-and honest reporters of our "deviant" responses-show how culturally defined the interaction of sufferer and respondent may be.
7. Wittgenstein: "A child discovers that when he is in pain for instance, he will get treated kindly if he screams; then he screams, so as to get treated that way. This is not pretense. Merely one root of pretense." Last Writings, Volume I, para.867 (Blackwell 1982).
8. See Ronald Melzack, "The McGill Pain Questionnaire: major properties and scoring methods," Pain I, 1975, 277-299.
9. Is such reflexive crying an unlearned basis or occasion for developing empathy? If so, then the movement is from your expression of pain to my response, not from my pain through inference to your similar sensation.
10. These public and social aspects of the concept of pain (both lay and clinical) makes conceptual space for the radical cultural differences that anthropologists report in manifestations of pain (for example, in village ceremonies) and responses to pain (for example, in maternal responses to infant whimpering in chronic food-scarce conditions).
11. The conflict between the relief-response concept of pain and the clinician-response concept is especially acute in neonatal matters, partly because the relations between causes, manifestations, and effective relief of pain are too tenuous and variable for clear definition. Hence, the counter-charges between "heartless" surgeons and "sentimental" lay critics. See Nancy Cunningham Butler, "Infants, pain and what health care professionals should want to know -- an issue of epistemology and ethics," and Dr. Neil Campbell's response in Bioethics 3:3, 1989, 181-210.
12. Indeed, even in the wild fantasies of the Marquis de Sade (Justine) or Pauline Réage (The Story of O), the hapless victims who endure and then come to enjoy the worst of imagined abuses seem to recover quickly, ready for another orgy of whippings and penetrations.
13. For example, "Dax's Case," a film about a severely burned patient treated over months against his will. During his excruciatingly painful tubbings and debridements the paramedics keep their radio blaring. In commenting on his case, his physicians can seem almost as unhearing. Also, The Right to Die?: The Dax Cowart Case New York: Routledge CD ROM 1996.
14. In Ernest Hemingway's "Indian Camp," a physician tells his young son that he does not hear the screams of the Indian woman on whom he performs a Caesarian section without anesthesia. Nor does he hear the empathetic cries of the woman's husband in the bunk above her-a suicide by the end of the ordeal.
15.In a Leg to Stand On, Oliver Sachs recounts his often callous treatment for a painful leg injury (New York: Harper & Row 1984). .
16. See the film, "The Doctor" (dir. Randy Haines 1991) about a physician (William Hurt) who required hospitalization as part of his medical students training after the humiliations of hospital treatment he himself had recently suffered.
17. E.g. U.S. Department of Health and Human Services, Acute Pain Guidelines Panel. Acute Pain Management: Operative or Medical Procedures and Trauma. Rockville, MD, 1992.
18. Cf. David Joranson, et al. "Opioids for chronic Cancer and Non-Cancer Pain: A Survey of State Medical Board Members. Bulletin of the Federation of State Medical Boards of the United States, June 1992: 15-49.
19. I wish to thank Dr. Neil Campbell, James Dwyer, F.M. Kamm, and Dr. Ronald Miller for help in clarifying these thoughts, even though not fully sharing them.