Placebo and Deception: A Commentary

Abstract

In a recent article in this Journal, Shlomo Cohen and Haim Shapiro (2013) introduce the concept of “comparable placebo treatments” (CPTs)—placebo treatments with biological effects similar to the drugs they replace—and argue that doctors are not being deceptive when they prescribe or administer CPTs without revealing that they are placebos. We critique two of Cohen and Shapiro’s primary arguments. First, Cohen and Shapiro argue that offering undisclosed placebos is not lying to the patient, but rather is making a self-fulfilling prophecy—telling a “lie” that, ideally, will become true. We argue that offering undisclosed placebos is not a “lie” but is a straightforward case of deceptively misleading the patient. Second, Cohen and Shapiro argue that offering undisclosed CPTs is not equivocation. We argue that it typically is equivocation or deception of another sort. If justifiable, undisclosed placebo use will have to be justified as a practice that is deceptive in most instances.

I. INTRODUCTION

In a recent article, Cohen and Shapiro (2013) have presented a novel and potentially important position on the ethics of placebo treatments: namely, that doctors are not being deceptive when they prescribe or administer some placebo treatments without revealing that these treatments are placebos. Although doctors are not being transparent in this situation, this lack of transparency does not amount to deception, according to Cohen and Shapiro.

Cohen and Shapiro offer two arguments against the deceptiveness of undisclosed placebos. First, they argue that offering undisclosed placebos is not lying to the patient, but rather intends to promote a self-fulfilling prophecy—telling a “lie” that, ideally, will become true. Cohen and Shapiro argue that this “lie” can be morally permissible, but they recognize that others will not agree. Thus, they make an additional argument to bolster the case for undisclosed placebos: for a specific subset of placebos, what they call “comparable placebo treatments” (CPTs), undisclosed placebo use is not deceptive at all.

“Comparable placebo treatments” are those placebo treatments “with objective biological effects similar to or virtually indistinguishable from the ‘real’ drugs they are supposed to replace” (Cohen and Shapiro, 2013, 700). For example, “various findings show that placebo analgesia can under certain conditions mimic the actions of opiates on the opioidergic system, the most potent pain-reducing system of the body” (699). Cohen and Shapiro are implicitly contrasting CPTs, which have objective biological effects similar to drugs, with other placebo treatments that have the same subjective effects as drugs (e.g., pain relief) without also having the same objective biological effects. Cohen and Shapiro argue that offering undisclosed CPTs need not be deceptive. In particular, they argue that when physicians communicate with patients about treatments being offered, such as saying “I am going to give you a treatment that should alleviate your pain,” the topic of this communication is the objective physiological effects of these treatments. Thus, when patients receive placebo treatments that have the same objective physiological effects as “real” drugs, they have not been deceived. The worst that can be said about undisclosed use of CPT, according to Cohen and Shapiro, is that it is manipulative, but a form of manipulation that can be ethically acceptable.

Cohen and Shapiro’s argument is important because if they are right that undisclosed use of CPTs is not deceptive, and that the lack of transparency involved in undisclosed use of CPTs is an acceptable form of manipulation, then this changes the landscape of ethical thinking concerning placebo treatments. However, their analysis is seriously flawed. As we will argue, their argument assumes, at root, that placebo deception, when it occurs, is deception about whether the treatment “works on the tissue level” or not. They conclude that because CPTs do work on the tissue level and not just psychologically, doctors do not deceive patients when they offer a CPT without revealing that it is a placebo. But Cohen and Shapiro are not entitled to assume that all that patients care about, and all patients can be deceived about, is whether their treatment has effects at the tissue level. As we argue below, offering a CPT with the disclosure, “I am going to give you a treatment that should alleviate your pain,” misleads the patient into concluding he is receiving a drug, or a “normal” treatment, when he is not.

The authors would have been on more solid ground had they argued that undisclosed use of CPTs is deceptive, but deceptive placebo treatments can be ethically justified in some circumstances on paternalistic grounds. In fact, Cohen and Shapiro do sketch such an argument: because deceptive placebo treatment can have good intentions and good results and can minimize the bad (i.e., medical side effects), they can be ethical (Cohen and Shapiro, 2013, 700–701). ¹ However, their primary line of argument is that undisclosed placebo treatments are not clear instances of deception. In any case, we contend that there is no easy way to escape the conundrum of “the lie that heals,” to use the language of Howard Brody’s classic paper on placebo use (Brody, 1982). If justifiable, undisclosed placebo use will have to be justified as a practice that is deceptive in most instances. ² Given the reasonable expectations patients have about medical treatment, undisclosed use of placebos, even when it qualifies as CPT, deceives many patients.

In Section II, we deconstruct Cohen and Shapiro’s claim that undisclosed placebo use is a self-fulfilling prophecy, and a “lie” that becomes true. In Section III, we object to their argument that offering a CPT with the disclosure, “I am going to give you a treatment that should alleviate your pain,” is not deceptive. In Section IV, we discuss empirical evidence on patients’ attitudes toward placebos which undercuts the authors’ argument that undisclosed use of CPT, though manipulative, is an acceptable form of manipulation.

II. LIES, “LIES,” AND SELF-FULFILLING PROPHECIES

Cohen and Shapiro claim that when a physician offers a patient a placebo and says, “I am going to give you a treatment that should alleviate your pain,” this statement is a “lie” that can be transformed into truth, a self-fulfilling prophecy. And as such, Cohen and Shapiro remark, it is unclear—and unlikely—that such statements should be considered deception. They write:

the placebo “lie” (that the doctor is administering a helpful substance) is (ideally) transformed into truth once it is believed by the patient. This difference between the prospective and retrospective accounts of the act significantly blurs the meaning of its very designation as deception. We are familiar with this dynamic as the logic of self-fulfilling prophecies, which in many well-known cases we view favorably. Think, for example, of the praise we would give to the chairman of a central bank who, in the midst of economic crisis, knows how to boost confidence in the robustness of the nation’s economy, thereby consolidating the very robustness to which he was alluding. Or think of the ethical praiseworthiness of the Pygmalion effect, where a teacher’s encouragement of the pupil to believe in his ability to succeed creates the extra element needed for that very success. (Cohen and Shapiro, 2013, 701–2)

In this passage, Cohen and Shapiro characterize statements such as “I am going to give you a treatment that should alleviate your pain” as “lies” that can be transformed into truths. Although these statements are deceptive, as we argue below, we think it is a mischaracterization to call them lies or “lies.” Cohen and Shapiro also characterize these statements as self-fulfilling prophecies; as we explain below, we agree with this characterization. They then analogize the doctor’s self-fulfilling prophecy to other self-fulfilling prophecies that we find ethically praiseworthy, such as a teacher who transforms a poor student into a good student by encouraging him to believe in his ability. This analogy is meant to leave us with the impression that the apparent deception involved in undisclosed placebo use—the “lie”—is not really deception and is ethically on a par with the teacher’s praiseworthy self-fulfilling prophecy. But, the deception involved in the doctor’s self-fulfilling statement “I am going to give you a treatment that should alleviate your pain” is logically and ethically different from the teacher’s self-fulfilling encouragement of the student.

We think there is a confusion in the above passage, facilitated by the use of scare quotes. The apparent deception in question—the placebo “lie”—is never clearly identified, leaving the reader unable to judge whether it is on a par with praiseworthy self-fulfilling prophecies. For the sake of clarity, we will spell out this critique in some detail.

Suppose that a doctor says, “I am going to give you a treatment that should alleviate your pain” or “I am giving you a substance that I believe will help your condition,” as she gives placebo pills to a patient. Has the doctor lied to the patient? That is, has the doctor said something that she believes false with the intention of making the listener believe it is true? If the doctor does not believe that the placebo will alleviate the patient’s pain, or help his condition, but merely prescribes a placebo to placate a patient who is insisting on getting a treatment, then her statement is a lie. But this first scenario, in which the physician lies in a clearly unethical way, is not the scenario that Cohen and Shapiro have in mind.

Presumably, they have in mind the ethically controversial scenario in which the doctor sincerely believes that the placebo might relieve the patient’s pain. In this scenario, the doctor’s statement “I am going to give you a treatment that should alleviate your pain” is not a lie, because the doctor believes it to be true. Rather, it is a sincere prediction that the physician expects, or at least hopes, will come true. Like many predictions, it might not come true; but this does not make it a lie. For example, our sincere prediction “We are going to mail this letter, which should arrive in two days” might or might not come true, but this does not make it a lie. Predictions do not linger in purgatory as provisional lies, waiting to be rendered retrospectively true or retrospectively false by outcomes. If predictions are sincere, then they are not lies—even if they are not guaranteed to come true. ³

Of course, the doctor’s statement is a special kind of prediction, one made with the intent that it will be a self-fulfilling prophecy. The statement “I am going to give you a treatment that should alleviate your pain” can bring about its own fulfillment, when believed by the patient. The doctor makes this statement with the hope that the patient will believe it is true and the hope that this belief will render the statement true. So when the doctor’s statement is sincere, it is not a lie. Nor is it a provisional lie, waiting to be rendered true or false as events unfold. Rather, it is a prediction of a special sort—a self-fulfilling prophecy—that might or might not come true. ⁴

Although it is not a lie, the doctor’s sincere statement is deceptive when it is not accompanied by a disclosure that the treatment is a placebo—that is, it is a pill that contains no medication. Cohen and Shapiro helpfully identify two distinct kinds of deception: making false statements with the intention of getting the listener to believe they are true (lies) and making true statements with the intention of misleading the listener. The doctor’s sincere statement “I am going to give you a treatment that should alleviate your pain” is not a lie, but it does intentionally mislead the patient into concluding that he is receiving a medication rather than a placebo. The doctor intends for the patient to conclude that he is receiving a medication; the doctor omits to tell the patient “This is a pill that contains no medication” precisely so that the patient will reach the false conclusion that he is receiving a medication. In this way, the doctor engages in deception by omission.

Is the doctor’s sincere statement a “lie”? This depends, of course, on what “lies” are. We can be sure that “lies” are not the same as lies; if they were, the scare quotes would be unnecessary. Beyond this, it is unclear what “lies” are. But here is a suggestion. If lies are false statements made with the intention of getting the listener to believe something false, perhaps “lies” are false statements made with another ultimate intention. One kind of “lie” is a self-fulfilling “lie”—a false statement made with the proximate intention that the listener believe it to be true, and the ultimate intention thereby to change what is true and false, so that the false belief instilled in the listener turns out true. Fiction, and art more generally, could be another kind of “lie:” works of fiction contain statements that are not strictly speaking true, with the intention of helping us see a deeper truth about life. This might be what Picasso meant when he said, “Art is the lie that enables us to realize the truth”—though he should have said, “Art is the ‘lie’ that enables us to realize the truth.”

On this understanding of self-fulfilling “lies,” some self-fulfilling prophecies are self-fulfilling “lies” but some self-fulfilling prophecies are not lies or “lies” at all. ⁵ Suppose that a student has a lot of innate ability but is a very weak student. His teacher, wise to the Pygmalion effect, hopes to turn him into a good student by encouraging him. She might tell him: “You have a lot of ability and by the end of the year, if you apply yourself, you could be one of the better students in the class.” When the teacher says this sincerely, she does not lie or “lie.” She makes a sincere prediction that is meant as a self-fulfilling prophecy; in this respect, the teacher’s statement is analogous to the placebo-prescribing doctor’s statement, “I am going to give you a treatment that should alleviate your pain.” But there is also a crucial disanalogy between them: the teacher’s sincere statement involves no deception whatsoever, but the doctor’s sincere statement intentionally misleads the patient about the treatment (e.g., making him conclude it is a drug, not a placebo) and is therefore deceptive. Thus, analogizing the doctor’s deceptive statement to the teacher’s non-deceptive statement should not make us feel better about the doctor’s deception. All this analogy can do is to help clarify the self-fulfilling logic of the doctor’s statement.

The doctor’s sincere statement “I’m going to give you a treatment that should alleviate your pain” is neither a lie nor a “lie”; it is a straightforward case of deception—saying something vague but true with the intention to mislead. Cohen and Shapiro’s invocation of a mysterious kind of apparent deception—the “lie”—makes this basic fact harder to see.

III. WHY IT IS DECEPTIVE TO OFFER UNDISCLOSED COMPARABLE PLACEBOS TREATMENTS

Cohen and Shapiro are rather skimpy in describing the context of physicians prescribing or administering CPTs. They begin by noting

When the doctor administering the placebo tells the patient, ‘I am giving you a substance that I believe will help your condition,’ the crux of the deception according to the traditional understanding is that the doctor expects the patient to assume the substance works on the tissue level, while intending his utterance to mean it just works psychologically. Deconstruction of that distinction could go a long way toward ethical rehabilitation of PT [placebo treatment]. (Cohen and Shapiro, 2013, 698)

It goes a long way according to the authors—indeed, all the way to obviating deception in their opinion—because recent scientific experimentation has shown that many placebo treatments work according to neurobiological mechanisms that are very similar to, if not indistinguishable from, that of drugs (Finniss et al., 2010).

But there is a basic problem in the way that the authors characterize CPTs from a scientific perspective: “Recent scientific research has produced growing evidence that objective parameters of the mechanisms of action of the placebo can be virtually indistinguishable from those of the ‘real’ drugs they are supposed to simulate” (Cohen and Shapiro, 2013, 698). They proceed to review briefly some of this research. Their account, however, glosses over the psychological dimension of placebo effects and the initial pathway within the brain by virtue of which some placebo treatments, such as those intended to relieve pain, turn out to have neurobiological mechanisms that are similar to drugs such as analgesic agents. One way in which placebo treatments are effective is that they tap an expectation pathway by means of a “top-down” process in the brain. They work via the patient’s belief that he is taking an effective treatment. In the case of placebo analgesia, the expectation pathway may activate the release of endogenous opioids, which produce pain relief (Benedetti, 2009, 39–42, 251). By contrast, analgesic drugs are taken by mouth and enter the blood stream, whereby they produce analgesic effects in the brain. One obvious marker of the difference is that analgesic drugs can relieve pain for unconscious patients or those who have no knowledge that they are receiving the drug (Colloca et al., 2004). This is not possible for placebo treatments.

Analgesics have a direct physiological effect on the patient’s body, whereas placebos, including CPTs, have a psychological-cum-physiological effect. It’s true that analgesics also have this psychological-cum-physiological effect—the normal way in which analgesic drugs do their work combines a characteristic physiological mechanism relating to the biochemical properties of the drug plus a placebo effect. But in the case of analgesics, the “top-down” psychological-cum-physiological process accompanies their direct physiological effects, whereas in the case of placebos, they work entirely via the “top-down” process.

The fact that there is something similar about analgesics and CPTs does not mean that we can gloss over the differences between them, namely, the psychological dimension of placebos. Part of the psychological dimension of placebos is the beliefs of patients who receive placebo treatments. Patients who are prescribed or administered placebo treatments without being informed that they are placebos are highly likely to have false beliefs about the treatments they take. Moreover, physicians administering these treatments in the way that Cohen and Shapiro suggest are highly likely to be intending to produce these false beliefs or at least foreseeing that patients will harbor such false beliefs. In other words, they are being deceptive. But there is a prior problem with Cohen and Shapiro’s discussion, which is that they make implausible assumptions about how doctors will administer placebos. Given the context in which placebo treatments can be prescribed or administered in contemporary clinical practice, it is not plausible that physicians can typically administer placebos non-transparently—that is, without revealing that the treatments contain no medication—without engaging in deception.

Consider how placebo treatments might be prescribed or administered in contemporary medical practice. Physicians generally are not in a position to write out a prescription for an “inert” placebo pill that will be sold to the patient in a pharmacy. Presumably, then, they will be handing the patient a bottle of pills. This is not a usual practice; however, sometimes physicians may give patients drugs that they have obtained from pharmaceutical representatives as free samples. Physicians might obtain placebo pills and give them to patients with instructions on how they are to be taken. If so, the bottle of pills will need to be labeled in some way. Surely, it won’t be labeled “comparable placebo treatments”; nor will it do to label it “obecalp.” Instead, it is likely to be given a name that sounds like that of a drug to which the placebo treatment is physiologically comparable. In this context, doctors do not maintain the strategic vagueness that Cohen and Shapiro imagine: they are not offering an unlabeled treatment with the vague statement “I am going to give you a treatment that should alleviate your pain.” Rather, they are giving the patient a pseudo-medication presented as a medication. It is natural to suppose that many patients receiving such a bottle of pills to take, say, to relieve pain will (falsely) conclude that the pills contain a medication that is pharmacologically active against pain. Some of these patients will Google the name of the pseudo-medication and realize that it is a placebo or they will ask the physician to clarify whether it is a placebo, requiring the physician to either engage in outright lies or reveal that it is a placebo.

But let’s grant, for the sake of argument, the scenario that Cohen and Shapiro (presumably) have in mind: a doctor offers a CPT, in an unmarked bottle, with the vague statement, “I am going to give you a treatment that should alleviate your pain.” Cohen and Shapiro argue that in this case, the doctor has not deceived the patient. Cohen and Shapiro acknowledge that this disclosure is manipulative and that after being given the disclosure, the patient remains ignorant of something about his treatment. However, they maintain that the patient has not been deceived. Recall that Cohen and Shapiro identify two kinds of deception: lying, and making true statements with the intention of misleading the listener. In the scenario under consideration here—a doctor offers a CPT with the sincere statement “I am going to give you a treatment that should alleviate your pain” —the doctor does not lie. Cohen and Shapiro’s remaining argumentative task, then, is to argue that the doctor does not intentionally mislead the patient. They do so by arguing that the physician’s statement is not equivocation—or at least, it is not equivocation that rises to the level of deception.

Equivocation is when a speaker says something that is open to multiple interpretations, some of which are true and some of which are false, with the intention of getting the listener to believe one of the false interpretations. Cohen and Shapiro thus set about to argue that when a doctor offers a CPT with the statement “I am going to give you a treatment that should alleviate your pain,” she has not said something that is open to multiple specific interpretations, some of which are true and some of which are false. The doctor’s statement, they argue, is not open to multiple interpretations—or at least, it is not open to multiple interpretations that are significantly different enough from one another to amount to meaningful equivocation. They write: “The CPT dynamic involves no hiding of information regarding the crucial element of the effect of treatment on the patient” (Cohen and Shapiro, 2013, 703). And:

[w]hen the doctor is, say, giving the patient a drug, the patient normally does not care about the chemical composition of the pill he is getting, but rather about what it does to him, and this, normally, is also the mental content that forms the doctor’s intention when addressing the patient about the treatment; this therefore is the plane on which equivocation could be morally consequential. (703)

Although the patient has been manipulated in a sense, Cohen and Shapiro think that this “does not involve deception (in the sense of concealment of information regarding the topic of communication between doctor and patient, i.e., regarding the effect of the intervention on the body of the patient)” (703–4).

Cohen and Shapiro contend that when doctors give patients disclosures about recommended treatments, the topic of communication between doctor and patient is the effect of the treatment on the patient’s body, meaning the ultimate physiological effect. The topic of communication does not include the mechanism whereby this effect occurs—that is, that the treatment’s mechanism of action is just the psychological-cum-physiological effect known as the placebo effect and does not include a direct physiological effect. Since Cohen and Shapiro think that the topic of disclosure communication includes only the ultimate physiological effect on the drug, the sole interpretation of the doctor’s statement “I am going to give you a treatment that should alleviate your pain” is, presumably, something like “I am going to give you a treatment that should alleviate your pain by causing the biological effects typical of pain-relieving treatments.” As we interpret them, Cohen and Shapiro maintain that other interpretations of the doctor’s statement, such as “I am going to give you an active drug that should alleviate your pain” or “I am going to give you a treatment that should alleviate your pain in the normal way that drugs do, by having a direct pharmacological/physiological effect,” are not feasible interpretations.

That multiple interpretations are not feasible is, we take it, what Cohen and Shapiro mean when they write “it will be virtually impossible to articulate two different senses of the doctor’s statement, which would establish an equivocation and thereby raise the charge of deception (Cohen and Shapiro, 2013, 705). To round out Cohen and Shapiro’s argument: because the patient’s only possible interpretation of the disclosure—“I am going to give you a treatment that should alleviate your pain by causing the biological effects typical of pain-relieving treatments”—is true, no deception has occurred.

On what basis do Cohen and Shapiro claim that the topic of disclosure communication is circumscribed in this way, to include only the effect of the treatment on the patient’s body? Cohen and Shapiro point out that patients typically do not care about the chemical composition of the pill they are getting. But even if patients may have no interest in the chemical composition of pills, they could still be interested in general facts about how the pills work. For example, patients might care whether the pills work by having a direct pharmacological/physiological effect on the patient’s body or just by having a psychological effect that causes a physiological effect. Moreover, they might care about whether the pill is a drug or not, precisely because they have a legitimate expectation that when prescribed a treatment in pill form, the pill contains medication known to be helpful in treating their condition. Whether or not patients do care about these distinctions is, of course, an empirical question. Interestingly, the authors ignore published studies reporting the attitudes of patients to placebo treatments. But even if patients do not care about these distinctions, they still might reasonably assume that these distinctions are part of the topic of the disclosure conversation. And if these distinctions are within the purview of disclosure communication, then patients will sometimes interpret vague placebo disclosures like “I am going to give you a treatment that should alleviate your pain” as making specific and false claims about the treatment being given.

Furthermore, even if Cohen and Shapiro were correct that the only salient interpretation of the disclosure “I am going to give you a treatment that should alleviate your pain” is “I am going to give you a treatment that should alleviate your pain by causing the biological effects typical of pain relieving treatments,” this does not show that the disclosure is not deceptive; it merely shows that the disclosure is not equivocation. But equivocation is only one form of deception. Another form of deception is making a true statement with the intention of causing someone to make inferences, on the basis of that statement, that are incorrect. If the doctor says, “I am going to give you a treatment that should alleviate your pain,” with the intention that the patient interpret this statement as “I am going to give you a treatment that should alleviate your pain by causing the biological effects typical of pain-relieving treatments” (which is correct) and the intention that the patient infer “This treatment is an active drug” (which is incorrect), then the doctor deceives the patient.

At this point, Cohen and Shapiro might object, arguing that the doctor who makes the intentionally vague statement “I am going to give you a treatment that should alleviate your pain” does not intend for the patient to reach false conclusions; she intends for the patient to reach the vague, true conclusion “I am going to give you a treatment that should alleviate your pain by causing the biological effects typical of pain-relieving treatments.” We think a more accurate description is that the doctor intends for the patient either to reach this vague, true conclusion or to reach a specific, false conclusion along the lines of “I am going to give you a drug that should alleviate your pain.” The doctor expressly wants the patient not to believe what is specific and true, namely, “I am going to give you a placebo that should alleviate your pain” or “I am going to give you a treatment that should alleviate your pain by having a psychological-cum-physiological effect known as the placebo effect.” The doctor intends for the patient either to remain ignorant of certain specifics or to form a false belief about them. The doctor engages in a communicative act that intends to produce one of two results, ignorance or false belief. ⁶ Although this lack of transparency is not the simplest form of deception, it is deception.

Perhaps some patients do care only about the biological effect of treatments on their bodies and not about the psychological, physiological, and pharmacological mechanisms underlying these effects. Perhaps some patients do not care if the pills they take are placebos, but only care if they are effective. These patients might not care if they are deceived about treatment mechanisms; and we might therefore question whether it is unethical to deceive these patients about placebos. But the fact that some patients do not care about being deceived should not be confused with the claim that the patients are not deceived.

Simply put, it is deceptive when a doctor offers a patient a CPT with a statement such as “I’m going to give you a treatment that should alleviate your pain.” For if the patient believes that the placebo treatment is a medication, as he likely will, then the patient has a false belief that the physician has intentionally created. The fact that the treatment works by some neurobiological mechanism comparable to an analgesic drug does not undo the deceptive communication that sets in motion the psychological process leading to effective treatment. This point is lost because the authors suggest that, in communication between doctors and patients about a treatment, the only issue that matters to patients is the effects of the treatment. Since CPT has essentially the same downstream effects as comparable drugs, there is nothing (or at least nothing of any significance) about which the patient is being deceived, according to the authors. But this assumption regarding what patients care about is belied by data on the attitudes of patients to placebo treatments, discussed below.

IV. IS UNDISCLOSED PLACEBO USE AN ACCEPTABLE FORM OF MANIPULATION?

While Cohen and Shapiro argue that use of CPT is not deceptive, they note that it does involve manipulation—indeed, a form of manipulation that is not ethically problematic. Additionally, Cohen and Shapiro claim that “manipulation (of the specific, restricted sort of temporary ignorance) is a necessary element of cooperation (between doctor and patient, to achieve the therapeutic goals the patient seeks)” (Cohen and Shapiro, 2013, 704). But this begs the question whether concealing from the patient that the placebo treatment is a pill lacking in medication is necessary to promote a therapeutic placebo response. In an endnote, the authors make the point that “placebo can sometimes work despite the awareness of manipulation” (707), citing two studies that support this. It is far from clear that keeping patients in the dark about receiving a placebo treatment is necessary to achieve the therapeutic goal of important symptom relief.

Cohen and Shapiro importantly assert that patients will (or should) be comfortable with physicians administering CPTs in a way that they acknowledge has an element of manipulation. Is that true? In a recently published interview survey of 853 US patients, respondents were asked a series of questions about “placebo treatments,” defined as follows: “A patient experiences a placebo effect when they [sic] get better after taking a treatment, not because of the treatment itself, but because the patient expects they [sic] will benefit from the treatment. For the rest of this survey, we will ask you to assume that a placebo treatment is a treatment that only works because it can produce a placebo effect. Placebo treatments can be sugar pills or other treatments used to create a placebo effect.” One question probed whether it is acceptable for doctors to recommend a placebo treatment described as “a real medicine” (Hull et al., 2013). Half of the respondents agreed and half disagreed. Now, this is not exactly the way in which Cohen and Shapiro describe CPTs being administered; however, it suggests that half of the surveyed patients would not endorse the use of placebo treatments when they are not described transparently, contrary to the assumption of the authors.

More specifically, respondents were presented with a scenario in which a doctor recommends a placebo treatment for a patient suffering from the common cold with the following disclosure, “I would like to recommend some medicine that might help you. It is safe and has no side effects. Some people get better when they take it, and I’d like you to try it.” When asked how the patient would feel if he later discovered that he had received a placebo treatment in this sort of scenario, 54% stated that it would have a negative effect on the future doctor–patient relationship. It is doubtful that the responses would have been different if the placebo treatment had been described simply as a “treatment” or a “substance” rather than a “medicine.” Thus, contrary to Cohen and Shapiro’s assumption that patients will be comfortable with physicians administering CPTs non-transparently, the available evidence suggests that patients are divided on this matter.

V. CONCLUSIONS

The argument of Cohen and Shapiro exemplifies the maxim that you can’t have your cake and eat it, too. Doctors cannot produce the paternalistic benefit of deceptive communication about placebo treatments without being deceptive. Because it is false that prescribing or administering placebo treatments without disclosing that they are placebos can be done without deception, we are left with pondering some thorny questions about the ethics of placebo treatments. Can the deception involved in nontransparent prescription or administration of placebo treatments be justified, all things considered? More specifically, can deceptive use of placebos be justified and effective if patients authorize it in advance by some process of disclosure of this practice and negotiation with physicians? Finally, can non-deceptive techniques of communication about placebo treatments be effective in promoting therapeutic placebo responses? More ethical inquiry and ethics-related empirical research will be needed to answer these questions.