Deception, harm, and expectations of pain

Introduction

In response to the national opioid epidemic – 42,000 Americans died from opioid overdoses in 2016 alone (Centers for Disease Control and Prevention, 2017) – clinicians have been urged to reduce unnecessary prescribing of opioids and other potentially addictive medications. One way to reduce such prescribing is to prime the patient to expect less pain, taking advantage of the cognitive modulation of pain to reduce its intensity. But while such priming may alleviate pain without pharmacological risks, it may also have significant moral and practical costs.

Gligorov suggests, without explicitly endorsing, the intriguing claim that it is not deceptive to tell a patient that a procedure “won’t be too painful” if that assurance will, or is reasonably expected to, make it less painful. She observes that such an assurance recruits the same or similar biological mechanisms as actual analgesics. We argue that (1) such an assurance would indeed be deceptive, and (2) claims of minimal pain that are justified primarily by their impact on the patient are not only deceptive but potentially harmful, as they may discourage prudent preparation and erode trust.

The conditionality of expectations

We agree with Gligorov that a clinician has considerable discretion in how she warns her patients about the potential for pain, and that she “ought to choose words that do not prime the patient to experience pain” (p. 12). In describing a procedure for which there is considerable variability in the pain patients report, the clinician would not deceive her patients in asserting that “although some people experience pain during the intervention, there is uncertainty about whether the procedure will have the same effect for each particular patient” (p. 13). But it is doubtful that such a cautious statement would have the desired analgesic effect. That effect is far more likely to be achieved by statements that predict pain at or near the lower end of the spectrum of patient reports, e.g., “this should only hurt a little” or “this shouldn’t hurt much more than a pinprick.” It is the acceptability of such statements that we wish to consider.

If the clinician were predicting a strictly physiological outcome, it would clearly be deceptive to forecast a result at or near one end of the spectrum if the probability of that outcome were low and she knew of no patient characteristic that made that outcome more likely. If a similarly optimistic prediction about pain is not deceptive, it is only because the clinician believes her statement is very likely to induce a belief in the patient that results in his feeling the low level of pain predicted. She would not deceive the patient if she made this explicit by conditioning her statement: “The level of pain you’ll feel will depend on your expectations. If you expect to feel only a little pain, that’s probably all you’ll feel.” But such a statement might not have the desired analgesic effect, instead inducing the same kind of debilitating uncertainty as the familiar warning, “The dog won’t bite you if you don’t show fear.”

Like the clinician, the dog owner could lower the odds of an adverse result by omitting the conditional, simply telling the passerby, “The dog won’t bite.” But if the passerby showed fear despite that assurance and was bitten as a result, he could surely complain that he had been misled. The owner failed to disclose a critical contingency – that he was safe only if he didn’t show fear. The clinician similarly fails to disclose a critical contingency – that the patient won’t feel much pain only if he believes he won’t. If the patient doesn’t believe an unconditional assurance and experiences more intense pain as a result, he, like the passerby, could complain that he had been misled. The clinician failed to mention something that the patient would very likely want to know, since it could affect whether he accepted the procedure or how he prepared for it. Her omission would be more objectionable than the dog owner’s, because she has a special duty to ensure that her patient’s consent is adequately informed. We suggest that to speak truthfully, the clinician must believe that her optimistic pain prediction is likely to be true, regardless of whether the patient believes it.

It is instructive to compare the closely related question of whether it is deceptive for a clinician to offer a patient a placebo with the explanation, “I am going to give you a treatment that should alleviate your pain.” Although Gligorov does not distinguish this kind of statement from an optimistic pain prediction, we see two differences – one making the clinician’s description of the placebo appear less deceptive, and the other more. The former concerns the fact that the clinician who describes a placebo “intentionally mislead[s] the patient into believing that he is receiving a treatment rather than a placebo” (Barnhill & Miller, 2015, p. 73). In contrast, an optimistic pain prediction does not rely on the patient’s belief in a false proposition for its fulfillment; it is merely a proposition whose truth is contingent on the patient’s believing it. On the other hand, as Gligorov points out, the analgesic affect of a placebo is not mediated entirely, or even primarily, by the patient’s belief that it is pharmacologically active. The very act of swallowing a pill that looks like a real drug may, through classical conditioning, achieve some analgesic effect, even if the patient doubts its efficacy or is told that it is a placebo. Its truth is less belief-dependent than an optimistic prediction, and in that respect, the clinician’s failure to reveal the source of the analgesia is less deceptive.

The importance of setting optimistic but realistic expectations

There is a practical upshot to this questionable means of shaping patient expectations. As Sweeny and Andrews note in a review of pre-surgical optimism and surgical outcomes, “As a whole, the evidence strongly supports an association between optimistic expectations and positive psychosocial outcomes following surgery if those expectations are based in reality [emphasis supplied]. If preoperative optimism ultimately turns out to be unrealistic, however, it is likely to be a postoperative liability” (2017, p. 374). In line with this finding, downplaying the potential for pain in an effort to establish optimistic expectations might result in the patient’s failure to take sensible precautions. For example, a patient might plan to return to work immediately or decide not to have someone accompany him to an outpatient procedure. He might not fill a prescription for pain medication pre-procedure, only to find that he needs the medication post-procedure but is no longer in a state to fill the prescription himself. Beyond simple inconvenience, it is not hard to see how such decisions might hinder recovery if they lead to ill-advised post-procedure overexertion or excessive worry that the procedure failed because the pain exceeds expectations. Even if the clinician had recommended taking precautions “out of an abundance of caution,” the patient might have dismissed her advice as defensive medicine, given her comment about the low level of expected pain. In seeking to establish optimistic but realistic expectations, there is a difficult but important balance to strike between choosing words that do not prime the patient for pain and ensuring post-procedure readiness in case of less desirable outcomes.

Furthermore, because patients differ in their credulity as well as in their pain tolerance and medical histories, such assurances may actually increase the variability in the levels of pain experienced in comparison to statements clarifying that patients experience varying levels of pain. For instance, a patient with complex regional pain syndrome may expect the worst if she is simply told an injection shouldn’t hurt more than a pinprick, given that her hyperalgesia has made pinpricks quite painful in the past. But if she were told that patients respond differently to the injection and an optimistic but realistic outlook might help mitigate injection-induced pain, she might be able to recalibrate her expectations to a lower level of pain. This example aligns with Gligorov’s discussion of the subjective standard for informed consent (p. 13), and this kind of nuance is an important safeguard for preserving the trust on which patient-clinician relationships are built. Ultimately, setting expectations for the pain associated with particular procedures has to be considered against the broader contours of helping patients maintain trust in their overall care plans; this is even more important for chronic pain patients, for whom pain management will likely remain an integral part of their lives.