Are Placebo Treatments Permissible In Clinical Practice?
This essay discusses placebo treatments, which, in rough terms, are medical treatments that do not involve pharmacologic substances. Placebo treatments may induce placebo effects: physical effects that bring about clinical improvement independent of pharmacologic properties. There is evidence to suggest that placebo effects are real, at least in some contexts. We do not know as much about them as we could because classic randomised controlled trials are not suitably designed to study them. To investigate placebo treatments and their effects would require a three-arm RCT that leaves two-thirds participants with inferior treatments. Such research, though in conflict with the traditional standard of permissibility, equipoise, might be permissible when accounting for informed consent and the proper goal of placebo treatment research – to understand placebo effects and how best to generate them, not to show the superiority of placebo treatments over conventional therapies. Equipoise is a key characteristic of the critical approach adopted by clinical practice. Evidence must show clinical treatments to be the best, most appropriate for the illness at hand. The lack of evidence supportive of placebo treatments is troublesome under this approach. This essay argues that placebo treatments are permissible in clinical practice if two conditions hold: (1) there is initial plausibility of the efficacy of placebo treatments and (2) there is an appropriate balance between individual and collective good. I argue these two conditions may hold, at least in some contexts. Further, if placebo treatments are permissible in clinical practice, then clinical research into placebo effects is called for so as to construct the best, most appropriate placebo treatments.
The essay will proceed as follows. Section 1 defines placebo treatments and placebo effects. Section 2 discusses the kind of clinical trials necessary to study them. Section 3 introduces the traditional notion of equipoise as the standard for the permissibility of clinical trials, and indicates why placebo trials are unlikely to satisfy this standard. Section 4 argues that there is initial plausibility for the existence of placebo effects in certain contexts. Section 5 puts forward a case for placebo treatments in clinical practice being in the collective good. Section 6 argues that placebo treatments are permissible in clinical practice in some contexts because the conditions of initial plausibility and collective good have been satisfied.
§1. This section defines ‘placebo treatment’ and ‘placebo effect’.
It is important to distinguish ‘placebo’ from ‘placebo treatment’. Placebos are inert substances containing no inherent pharmacologic properties that elicit effects. Sugar pills or saline solution are example placebos. A placebo treatment is the positive context of a medical treatment. Part of a placebo treatment might be the administration of a placebo. For example, a doctor may tell their patient that they have administered pain relief treatment when actually the patient has been given saline solution. The doctor’s suggestion of pain relief and the dramatics of the injection are the placebo treatment. The inert saline solution, used to make the placebo treatment more convincing, is a placebo. If the doctor suggested pain relief and injected their patient with morphine, then the doctor’s suggestion and the dramatics of the injection are the placebo treatment (the positive context) and the morphine, though again bolstering the placebo treatment, is an orthodox treatment. In this scenario the placebo treatment does not incorporate a placebo. So, placebo treatments are the positive context of medicine and might or might not involve placebos.
Placebo treatments try to induce a relaxed state of optimism in the patient because positive expectation is thought to elicit placebo effects. Placebo effects are neurobiological changes that aid healing but are not the effects of an orthodox treatment with inherent pharmacologic properties (Brody, 2002; Goldacre, 2009).
Prima facie evidence suggests placebo effects are real and that placebo treatments can induce placebo effects. For example, one study subjected 57 college students to a very boring hour-long lecture. Each student was given either one or two pills, which were either pink or blue. The students were told to expect to receive a stimulant or a sedative, but each student actually received sugar pills. After the lecture, researchers measured each student’s alertness. The study found that two pills were better than one for keeping students alert and that pink pills was better than blue for maintaining the students’ concentration. Since colours in themselves have no pharmacological properties, the difference in elicited effects must be due to the cultural meanings attached to the colours: pink implying alertness and blue implying cool and calm (Goldacre, 2009). This study further suggests that placebo effects are meaning responses to the patient’s experience of the treatment. Similar studies support this assertion: pain relief elicits stronger effects if the patient is told it costs $2.50 than if told it cost just 10 cents; newer forms of treatment administration tend to be more effective than conventional forms; brand name treatment elicits greater effects; the more dramatic the administration process, for instance injection rather than pill, the more effective the treatment (Goldacre, 2009; Benedetti, 2009).
If these studies correctly establish placebo effects as meaning responses to medical treatment, then it would follow that the whole atmosphere surrounding the treatment could affect the magnitude of the healing effect. The effects of an empathetic doctor-patient relationship and of a patient’s belief that a treatment is powerful, expensive or has a ‘good’ brand name are now important elements. Moreover, the effects of colour, method of administration, and content become significant (Howick, 2009). Further, these elements could be manipulated to produce the optimal healing effect for the patient (Benedetti, 2009). For instance, the doctor could overstate his confidence in the administered treatment to try and boost the positive expectancy his patient has in his prescribed treatment, which in turn might elicit placebo effects that add to the whole healing effect.
In short, placebo effects are induced when a medical treatment is surrounded by what the patient perceives as a positive context and neurobiological changes occur resulting in clinical improvement. Though, because placebo effects are meaning responses, it follows that they are not inevitable consequences of placebo treatments. Placebo treatments might or might not induce placebo effects all the time, or in every patient universally. The patient’s past experience, understanding and motivation will affect the psychosocial context of their medical treatment.
It has been suggested that placebo effects are merely false-positive error (Allan and Siegel, 2002). That is, the patient erroneously detects clinical improvement despite no supporting physical evidence. To illustrate, a doctor may administer placebo treatment to a patient experiencing pain. Part of which is telling the patient he is receiving pain relief. Following placebo treatment, the level of pain experienced by the patient is neither so high as to be unambiguously a signal the pain relief had failed nor so low as to be unambiguously background noise. The patient reports the intensity of the pain is variable, but the average intensity decreased after treatment (Allan and Siegel, 2002). A cost-benefit analysis was conducted in judging whether the medication was effective and a liberal criterion applied. In the example the patient perceives the cost of missing a signal that the treatment worked (false-rejection) to be greater than the cost of erroneously reporting the presence of such a signal (false-positive). Supporters of this argument contend it is likely a patient given a placebo treatment would rather make the mistake that pleases (Allan and Siegel, 2002). Then placebo effects are no more than reporting bias.
However, placebo effects are physical phenomena. Brain-imaging studies show the subjective experience of pain reduction following placebo treatment is accompanied by objective changes in several brain regions where pain transmission is inhibited (Benedetti, 2009). Patient-reporting bias might explain the effects of psychosocial factors when the effect is not accompanied by neurobiological change. But the mounting neurobiological evidence means there is a tangible and, at least, potentially measurable difference between patient-reporting bias and placebo effects (Benedetti, 2009). Patient-reporting bias is localised in the mind, placebo effects go beyond the mind and affect the state of the body. The changes in the body can be evidenced and measured, for instance, an increased level of endorphins (euphoria-inducing chemicals) in the brain (Collin and Pinch, 2005). Placebo effects cannot be dismissed as reporting bias.
§2. This section discusses the kind of clinical trial necessary to study placebo effects.
Classic randomised controlled trials (RCT), as used routinely in clinical research, are not good settings for understanding placebo effects. Studies show that magnitudes of pain relief for placebo treatments that incorporate placebos are higher in laboratory experiments investigating placebo pain relief mechanisms than clinical trials that used a placebo control (Benedetti, 2009). In the laboratory experiments the researchers are investigating placebo effects and want to find the best ways of inducing them. Elements like empathetic doctor-patient relationship are cultivated and become the subject of the test. In contrast, the RCT, conducted in accordance with certain extra conditions, for example, double blinding, tries to find an objective treatment effect and so eliminates elements that will elicit subjective effects. Double blinding a trial serves to eradicate treatment bias but when investigating placebo effects it is the treatment bias that is of interest. Ciphers, such as oral suggestions of improvement, which indicate medical treatment, are kept to a minimum. Restricting the patient’s experience of treatment detracts from their potential to derive a meaning response, namely, placebo effect.
Constructing an RCT that compares the best available treatment for a particular illness with placebo treatment, that is, incorporate ciphers that cause the patient to reach a state of positive expectancy of clinical improvement from the treatment administered, against a group taking the best available treatment without placebo treatment does not solve the problem. It is impossible to administer any treatment without placebo treatment creeping in. It is impossible to separate a medicine from its context. Simply being treated by a doctor could cause the patient to hope for clinical improvement and thus, potentially, induce placebo effects. So, this set-up is not a good way to isolate and measure the healing effects of any given placebo treatment because both groups mix pharmacologic based treatments and placebo treatments.
The simplest way to test placebo effects with an orthodox methodology is to conduct a three-arm RCT. One group would receive the conventional best available treatment (the treatment effect of which is already known), the second would receive a placebo treatment involving a placebo, and the final group would receive no treatment whatsoever. A study of this type would be able to measure the efficacy of placebo treatments. Altering the combination and proportion of elements making up the placebo treatment would show the efficacy of the placebo treatment as a whole, and researchers could tinker with the formula to find the most efficacious combination.
§3. This section discusses equipoise as the standard for permissibility of clinical research and explains why placebo treatments are unlikely to satisfy this standard.
A doctor is required to give his patient the treatment he believes the best and most appropriate, and at least do no harm. Clinical research has shied away from using placebo-controlled trials. Guidelines state placebo-controlled trials are only permissible if there is no pre-existing conventional treatment. A three-arm RCT as specified above would leave two-thirds of participants without orthodox treatment. Such research risks patient safety and would therefore go against the doctor’s ethical obligation.
Using placebo treatments would not present a problem if the doctor believes them to be the best, most appropriate treatment. If the doctor is unsure whether placebo treatments or some other treatment would be the best and most appropriate, a trail design as above would still be permissible. Equipoise, to be precise individual equipoise, is the term used to describe the doctor’s uncertainty as to the comparative efficacy of treatments. Equipoise would call for an RCT to identify the superior treatment. However it may be that though an individual doctor is confident as to which treatment he considers the best, most appropriate for the given patient, this opinion is not the consensus of the medical community. Then the doctor may be forced to prescribe a treatment he considers inferior, which again conflicts with his ethical obligation.
In recognition of this, a new kind of equipoise is appealed to. The institution of medicine takes the decision of what is the best, most appropriate treatment out of the hands of the individual doctor and places the onus on the beliefs of the medical community. Adopting a critical approach where a treatment must show itself superior in RCT also protects the institution of medicine from extreme, improper and possibly dangerous beliefs of individual doctors. This clinical equipoise does not allow the doctor to favour any particular treatment unless it has been shown superior in RCT, only when statistical evidence supports this contention does the treatment become the best, most appropriate for the illness at hand (Truog, 1999). Clinical equipoise would require placebo treatments to undergo RCT against best pre-existing treatment and be shown superior.
But when the alternative to the conventional treatment is no treatment at all or placebo, the position of clinical equipoise is difficult to justify. Doctors and researchers are administering treatments (or no treatment) known to be inferior, which seems indefensible both in the case of individual doctors and the abstract collective construct of the medical community.
In the context of clinical research into placebo treatments and placebo effects, patient participants are informed that there is two-thirds chance they will not receive the conventional best available treatment, a third chance they receive the placebo treatment and a third chance they receive no treatment. Further, they are informed as to the risks of receiving placebo treatment and no treatment rather than conventional treatment for their given illness. Any patient participants have given informed consent and so the research is not ethically questionable. Of course, if fully informed there may not be very many patients who give their consent, which would make clinical research into placebo treatments and placebo effects difficult. If it remains a requirement for clinical practice that treatments have shown themselves to be superior in RCT then placebo treatments are unlikely to satisfy this standard.
Yet the point of conducting the clinical research into placebo treatments and placebo effects is not to establish them as superior treatments, but to understand how to induce placebo effects and measure the healing effects of placebo treatments. The idea being that since the context of medical treatment could have its own healing effects, how best to harness these healing effects to maximise the likelihood of clinical improvement. When considering whether placebo treatments and placebo effects are permissible areas of clinical research and consequently permissible in clinical practice, clinical equipoise does not seem an appropriate framework to evaluate.
Below I argue that, firstly, there is initial plausibility that placebo effects are real and, secondly, that placebo treatments in clinical practice, in specific contexts, is in the collective good. It is my contention that in satisfying these two conditions is enough to make the use of placebo treatments in some contexts permissible in clinical practice. Further, I suggest these conditions call for clinical research into placebo treatments and placebo effects. The greater the doctor’s understanding of placebo effects the more effective placebo treatments can be.
§4. This section argues that there is initial plausibility for the existence of placebo effects in certain contexts.
This argument requires doctors to take a utilitarian stance, modifying the doctor’s ethical obligation to account for future patients rather than just his immediate patients. This is a position similar to one needed to justify clinical equipoise. If a new experimental treatment is being tested in RCT to show its superiority against the conventional treatment, then the new treatment has reached this stage because it is initially plausible that it will be shown superior. The RCT risks giving some patients an inferior treatment (either the experimental or the conventional treatment is likely to be inferior), but is permissible because future patients may be given a treatment that is superior to the current conventional treatment.
In the natural sciences it is generally thought permissible to carry out research based on intuitions despite more often than not the failure of those intuitions to generate useful scientific knowledge. It is the initial plausibility that gives license for further investigation.
Although placebo effects are real neurobiological changes that can be witnessed, it is impractical for researchers to brain scan every patient participating in a study to determine whether they are reporting bias or placebo effects. So it is important placebo effects be plausible. If placebo effects are plausible it goes part way towards justifying clinical research into placebo effects and placebo treatments.
Intuitively it seems that for so called ‘subjective’ conditions, such as pain or depression, if the patient believes they feel better then they are actually feeling better (Collins and Pinch, 2005). There is initial plausibility to think that for conditions with subjective outcomes placebo effects exist. For example, one orthodox treatment for depression is psychotherapy. Psychotherapy relies heavily on a positive and empathetic doctor-patient relationship and verbal suggestions, which would point to psychotherapy being a placebo treatment (albeit without a placebo). An explanation for the efficacy of psychotherapy might be rooted in the trust, belief, expectation, hope and motivation sparked by this human interaction. If this were so, then it would imply there are healing effects (placebo effects) in a benign human relationship (Benedetti, 2009). Different forms of psychotherapy work because they contain similar elements, such as rituals to reinforce the doctor-patient relationship and the presence of a thoughtful listener. It is difficult to disentangle placebo effects from psychotherapy both methodologically and conceptually. If the most important ingredient of psychotherapy is faith and expectation, we should consider it a placebo effect because an objectively effective psychotherapy should work through mechanisms other than faith, belief and expectation (Benedetti, 2009).
Similarly, Alternative Medicine uses an empathetic doctor-patient relationship to elicit placebo effects. Homeopathy is an exemplary case. Homeopathic remedies are substances that mimic the symptoms the patient is suffering (for example, caffeine for insomnia). The remedies are diluted to a point where the solution finally handed to the patient contains no molecules of the original substance. According to orthodox medicine, molecules of the remedy need to be present in the body but in the case of homeopathy this is not so. By any laws known to science the remedy cannot interact with the biochemistry of the patient’s body (Brooks, 2010). Yet homeopathy remains popular, particularly with patients suffering chronic, quality of life affecting conditions. Further, there is evidence to suggest that homeopathy induces healing effects. A study found that histamine solutions, both at pharmacological concentrations and diluted out of existence, lead to statistically significant inhibition of patients feeling they need to go to hospital (Brooks, 2010). The ‘ghost’ presence of histamine in the homeopathic solution was enough to prevent hospital activation. Homeopathy has an effect, but it must be a placebo effect (Brooks, 2010).
Psychotherapy and homeopathy have been shown to elicit healing effects sometimes in clinical practice. If these treatments are in fact placebo treatments (as I think the above suggests) then these healing effects go to the efficacy of placebo treatments, and therefore that placebo effects are intuitively plausible for subjective conditions. In which case clinical research into placebo treatments and placebo effects should not be dismissed outright.
Importantly, in limiting the initial plausibility of placebo effects to certain illness, a set of conditions can be identified as likely to induce placebo effects. The illness at hand should be subjective, such as pain or depression, and the context surrounding the treatment (placebo treatment) should reinforce a positive message that the patient will experience clinical improvement, for example the expression of confidence in the administered treatment from a trusted doctor. There seems to be a rule of thumb for inducing placebo effects, which is a stronger claim than just an initial plausibility they exist. Doctors can set up placebo treatment in line with the rule of thumb conditions and plausibly expect to induce placebo effects. The rule of thumb plausibility of beneficial placebo effects would seem to make the idea of placebo treatments in clinical practice defensible and, consequently, justifies further clinical research.
§5. This section argues that the use of placebo treatments in clinical practice is in the collective good, which in turn means clinical research into placebo treatments and placebo effects is in the collective good.
Utilitarianism provides a framework suitable for healthcare. Given that health is a public good it is appropriate to talk in terms of collective good. A popular view for publicly funded national health services is to operate on utilitarian principles. Placebo treatments would be permissible in clinical practice if and only if they produced as great a balance of benefit (pleasure over pain) as any alternative action. Similarly, clinical research into placebo effects would be permissible if and only if it produced as great a balance of benefit as any alternative action. If it can be shown that placebo treatments are in the collective good then it would follow that clinical research into placebo effects would be in the collective good because it would be in the interest of the doctors to have as good an understanding of them as possible and so optimise the benefits of placebo treatments. That the utilitarian framework exists is enough for the purposes of this essay to continue this argument.
As discussed above, clinical equipoise is inappropriate for the question of this essay. In §4 I established that there was rule of thumb plausibility for placebo effects, which, in the particular circumstances, should take away a doctor’s individual equipoise as to whether placebo treatments would be efficacious. Although some individual patients may not benefit from their placebo treatment, there is no human cost to using placebo treatments in clinical practice in the sense that the average result is benefit so long as the rule of thumb plausibility transpires. Clinical research requires informed consent meaning clinical research should not have a human cost because participants volunteer in knowledge of the risk.
The collective good derived from the use of placebo treatments in clinical practice and clinical research into placebo effects may be limited. Every medical treatment has an inescapable and inseparable context and the ‘orthodox’ context probably elicits its own placebo effects. Perhaps orthodox medicine’s effort to eliminate placebo treatments itself produces placebo effects (Kaptchuk, 2002). Patients, especially emergency patients, might elicit placebo effects in response to a calm, orderly, high-tech hospital environment and a kind but focused doctor who does not stop long to chat but instead brings his full attention to the pressing business at hand (Spiegel and Harrington, 2008). The world of clinical practice may be full of placebo effects that either bolster efficacious treatments, or perhaps cover up the fact that these treatments are not efficacious. The rigorous methodology imposed by orthodox medicine, while trying to merit the success of medicine to science, may paradoxically have instilled an increased belief in the power of medicine that has made this success possible. Orthodox medicine systematically undermines the idea of the power of the mind over the body, yet the patient’s knowledge that orthodox treatments have proven their worth in RCT means that the patient expects the treatment to work for them, the expectancy can generate placebo effects, which are an example of the mind manipulating the body. Orthodox medicine may be like the Red Queen running to stand still, deliberately denying the power of the mind because it requires placebo effects to maintain its success rate (Brooks, 2010). Eroding the trust the patient has that they are receiving an orthodox treatment, could stop the placebo effects the patient enjoys from the hope they associate with orthodox, scientifically ‘proven’ treatments.
Collective good may also be limited because, so far, it is only plausible that placebo effects are elicited in patients with subjective illnesses. Placebo effects will not bring about clinical improvement in patients with illnesses like malaria or cancer (though placebo effects may help to alleviate some of the subjective symptoms of these illnesses).
Collective good may be further limited in the case of placebo treatments in clinical practice because of the need to deceive patients. If patients know they are taking a placebo treatment they will derive no benefit (Brody, 2002). As soon as you tell the patient they are taking a placebo treatment it ceases to be a placebo treatment and becomes ‘no treatment’, or merely a placebo. And vice versa, if the doctor fools the patient into thinking they are being treated when they are not, then they are receiving a placebo treatment (Collins and Pinch, 2005). The key here is that placebo treatments are never merely placebos but a positive context. Granted, some patients will see through the positive context and remain unhopeful. It is not one placebo treatment fits all. The art of medicine is to fit a particular placebo treatment to a particular patient. A good doctor will want to offer help to patients suffering illnesses for which there is no scientifically accepted means of alleviation. In this case offering a placebo treatment while concealing from the patient that there is no orthodox treatment may be acceptable (Collins and Pinch, 2005). Such a situation does not feel intrinsically bad or immoral providing the patient derives benefit from the lie, or even if the intention of the doctor was to benefit the patient. The problem enters when the patient discovers the deception. Even when doctors use placebo treatments for good reason and obtain good results, if patients (or their relatives) discover the deception, this may contribute to an erosion of trust and damages the institution of medicine (Benedetti, 2009). Doctors seem wary of this. Published surveys show that many doctors are aware of the benefit their patients may experience from placebo effects, but are concerned about how to generate placebo effects in a non-deceptive manner (Fässler et al., 2011). For instance, doctors are hesitant to use placebos and would rather use treatments that have known pharmacological action but that cannot be expected to have any direct therapeutic effects for the respective illness or in the chosen dosage (Fässler et al., 2011).
However, evidence suggests doctors underestimate the openness of patients to placebo treatments and the healing power of placebo effects (Fässler et al., 2011). The American Psychological Association has suggested a criterion by which deception is acceptable in clinical trials. Slightly adapting this criterion, I suggest one detailing when it is acceptable to use deception in clinical practice. When it is expected to derive significant social and scientific value, when an equally effective non-deceptive approach is not available, and when patients are not deceived about any aspect of their treatment that would affect their willingness to take it. Placebo effects can be induced about any orthodox treatment with no deception. Possibly this approach, boosting orthodox treatments with placebo treatments, rather than placebo treatments involving placebos would be more appropriate for clinical practice as it would not damage the institution of medicine. The only deception would be exaggerating confidence in a positive treatment outcome in certain contexts. This is still reason enough to investigate placebo effects because there is reason to think this approach might elicit more positive treatment outcomes and thus maximise the healing power of orthodox medicine. Again, informed consent means that there is no deception involved in clinical research into placebo effects.
Moreover, using placebo treatments in association with orthodox treatments would allow any placebo effects induced from the patient’s knowledge that they are receiving orthodox treatment to continue.
So, collective good from placebo treatments in clinical practice is limited to subjective illness and to placebo treatments being administered in association with orthodox treatments. Below I outline some benefits to using placebo treatments in this way and argue that, despite the limitations, on balance placebo treatments in clinical practice are still in the collective good.
If placebo treatments are efficacious, as is plausible, then refusing to permit them in clinical practice could damage medicine as therapy even while improving it as science. The purpose of medicine is to heal patients, but medicine is also succour. Acknowledging placebo effects acknowledges medicines role as succour and can be done in a way to maximise long-term collective healing by incorporating placebo treatments alongside the administration of orthodox treatments. Treatments that heal mind and body can in a sense be tailored at the level of the individual because they take into account that the person of the patient can elicit healing effects. Just as the person of the doctor, by virtue of being a doctor, can elicit healing effects (Sullivan, 1993). With placebo treatments it may be possible for medicine as science and as succour to unite and for the interests of the individual and of the collective to converge (Collins and Pinch, 2005).
Placebo treatments can restore the autonomy loss a patient suffers as a result of their illness. Though the efficacy of placebo treatments and hence placebo effects turns on the patient being offered no genuine choice, as any attempt to offer a choice is self-defeating, placebo treatments can be justified by invoking the concepts of paternalism and benevolent deception. The doctor’s purpose is not to deceive but to cure. Sometimes the effect of the illness is an undermining of the patient’s autonomy and the doctor must restore this loss (Benedetti, 2009). To retain the trust in the institution of medicine rules are required as to when the use of placebo treatments is permissible. Benedetti (2009) suggests benevolent deception should never be employed for the convenience of the caregiver, it should only be used is cases where there is substantial evidence that it is necessary, the doctor should determine whether any physical or psychological illness for which other treatments are indicated would be mastered by placebo treatment.
Placebo treatments can be evidenced, measured and to a point made scientifically objective and therefore suitable for clinical practice. For instance, sugar pills administered with a positive message have been found to be associated with less fear, less anxiety, and less pain (Goldacre, 2009). One study defined the content of the doctor-patient relationship as including questions concerning the patient’s symptoms, how the illness related to the patient’s relationships and lifestyle, possible non-illness symptoms the patient might be suffering, and how the patient understood the cause or meaning of his illness. The doctor also incorporated five behaviours in meetings with the patient, including a warm and friendly manner, active listening (such as repeating the patient’s words back to them, or asking for clarifications), empathy (“I can understand how difficult your illness must be for you”), 20 seconds of thoughtful silence while feeling the patient’s pulse or pondering a treatment place, and communication of confidence and positive expectation for the treatment (Kaptchuk et al., 2008). Thus the doctor-patient relationship can to some extent be made scientific, that is, it can be defined, measured, and evidenced. As such there should be room for placebo treatments in clinical practice. The methods of psychotherapy might be an example of this.
Regardless of whether they are efficacious placebo treatments are already used in clinical practice, to let this use go unchecked risks patient safety. Placebo treatments are traditionally used in clinical practice to please or placate anxious and complaining patients. A survey of Israeli doctors determined that 60% had prescribed placebo treatments, and over half once a month or more. Of the doctors who prescribed placebo treatments, 94% said they found them to be an effective means of treatment (Nitzan and Lichtenberg, 2004). A survey of nearly 800 Danish doctors found that almost half prescribe placebo treatments 10 or more times a year (Brooks, 2010). Even researchers who question the efficacy of placebo treatments concede their impact for illnesses that are continuous and subjective (Kaptchuk, 2002). Despite not understanding the mechanisms and without orthodox evidence, doctors still take advantage of placebo effects routinely in clinical practice. Introducing some professional standards could help regulate such employment of placebo treatments, and clinical research could aid the medical community’s understanding of placebo effects so placebo treatments can be informed.
Similarly, the Alternative Medicine industry takes advantage of placebo effects, sometimes with good results. The market for alternative medicine suggests a market for pseudoscience (Goldacre, 2009). In particular, ‘subjective’ illnesses that diminish one’s quality of life, which evidence has shown are most likely to respond to placebo effects (Kaptchuk, 2002). Many patients may be drawn to alternative medicine because of the holistic concern for well being they are likely to experience, and many may also experience appreciable placebo effects. Doctors sometimes refer their patients to alternative or complementary therapists. The World Health Organisation calls Homeopathy an integral part of the national healthcare system of a number of countries including Germany, UK, India, Pakistan, Sri Lanka, and Mexico (Brooks, 2010). If orthodox medicine itself provided the placebo effect inducing atmosphere homeopathy does, then such referrals would be unnecessary. The public association of orthodox medicine with the fantastical pseudoscientific explanations of alternative medicine would not damage medicine as science. Effective placebo treatments might explain why so many patients are prepared to pay to be treated by alternative therapists despite the treatments being unevidenced (Spiegel and Harrington, 2008). If the patient’s illness is simply not one that is placebo effect responsive, even if some of its symptoms are, then leaving the patient in the hands of Alternative Medicine could be life threatening. Permitting placebo treatments in clinical practice may well save lives by keeping patients within the fold of efficacious, rational, orthodox medicine (Brooks, 2010).
Finally, the greater understanding of placebo treatments and placebo effects that clinical research may bring could have experimental benefits. Research into placebo effects will help medicine to understand how better to account for their impact on results. Hopeful expectancy can boost the direct effects of experimental treatments in RCT. Placebo effects can be additive effects. Their mechanisms can enhance the specific effect of the experimental treatment (Benedetti et al., 2005). Alternatively, the specific effects of the experimental treatment may only occur over a threshold level of placebo effect activity; placebo effects may potentate the specific effects of the experimental treatment (Howick, 2008). Placebo treatments can bring about placebo effects, but the actual administration of a placebo is not necessary. As soon as expectancy factors, even real neurobiological effects cannot be attributed to the experimental treatment in an RCT. Also, placebo effects can boost the positive outcome of the control group and as such make it harder for the experimental treatment to show its superiority. If a successful treatment is one that is superior to a placebo in a double blind RCT, then every occurrence of a placebo effect makes it less likely a promising new treatment will be shown to be useful (Brody, 2002). Placebo effects can obscure or exaggerate (or both) treatment effects in RCT, and can alter the efficacy of approved orthodox treatments in clinical practice. Double blinding is not a good control for expectancy, and moreover is difficult to maintain. If one group shows dramatic improvement then those involved in the trial may guess this group is the experimental group that has been administered the experimental treatment. Or if one group suffers side effects, such as a dry mouth or dizziness, those involved may guess they have been administered the experimental treatment. Certainty is not required for placebo effects. All that is required to elicit a placebo effect is suspicion. Suspicion generates expectancy, possibly optimism. Optimism induces placebo effects. Simply the patient’s knowledge they are participating in a clinical trial can elicit placebo effects (Hawthorne effects). Meta-analyses suggest that in clinical trials for major depression which demonstrated a positive outcome, a quarter is due to the specific action of the pharmacologic treatments, a quarter is due to other factors, like spontaneous remission, and half is due to a placebo effect (Benedetti, 2009). Research into placebo effects will help orthodox medicine to understand how better to account for their impact on results.
§6. Placebo treatments are permissible in clinical practice.
I have established in §4 that there is a strong initial plausibility for placebo effects within certain contexts. The rule of thumb plausibility for placebo effects is for subjective conditions when the context of treatment invokes a state of positive expectancy in the patient. Also, I have explicated a number of clinical and experimental benefits to permitting placebo treatments in clinical practice and clinical research into placebo effects. In fact the limitations necessary to maintain patient safety and trust in the institution of medicine may be an additional benefit. Combining placebo treatments with orthodox treatments allows medicine to fulfil its unique duel role of science and succour. I suggest that §5 shows that placebo treatments used in this limited way in clinical practice are in the collective good. Having satisfied the conditions of being initially plausible and being in the collective good, placebo treatments in clinical practice are permissible. It is impossible not to have placebo treatments as they are the inseparable context of any treatment. Yet, meeting my criteria for permissibility gives doctors scope to exaggerate or overstate their treatment recommendations in contexts that coincide with the rule of thumb plausibility. And, moreover, to orchestrate the entire atmosphere of a treatment to induce placebo effects. Clinical research will help pinpoint how to construct the best, most appropriate placebo treatments for the illnesses at hand.
Placebos are inert substances such as sugar pills. Placebo treatments are the positive context of medicine. They might or might not involve placebos. Placebo treatments try to induce a relaxed state of optimism in the patient, which can elicit placebo effects. Placebo effects are meaning responses to the patient’s experience of treatment; they are neurobiological changes that result in clinical improvement. Placebo effects are not pharmacologic effects.
Classic RCT incorporates conditions such as double blinding that serve to make them poor settings to investigate placebo treatments and placebo effects. Classic RCT seek to eradicate treatment bias, but when investigating placebo effects it is the treatment bias that is of interest. The simplest way to test placebo effects is with a three-arm RCT that administers one group a conventional treatment, another a placebo treatment and the third no treatment.
Clinical equipoise is not a justifiable position when considering clinical research into placebo treatments and placebo effects. This research is not aimed at showing placebo treatments to be superior but to better understand them. Patients participating in clinical research must give their informed consent therefore placebo research is not unethical and permissible.
I suggest that if the efficacy of placebo treatments can be shown to be initially plausible and if placebo treatments in clinical practice can be shown to be in the collective good, then they are permissible. If placebo treatments are permissible then clinical research into placebo effects is called for to improve the efficacy of those treatments.
Treatments that seem to be placebo treatments, Homeopathy and Psychotherapy, have been shown to elicit healing effects in clinical practice. The initial plausibility of placebo effects means placebo effect research is permissible. The fact that this initial plausibility only extends to subjective conditions and that the patient has to expect clinical improvement, leads us to a rule of thumb plausibility. The rule of thumb plausibility for the efficacy of placebo treatments makes the idea of placebo treatments in clinical practice defensible.
Adopting the utilitarian view gives us license to think that being in the collective good is a justification of using placebo treatments in clinical practice. The rule of thumb plausibility outlined in §4 suggests that, in the particular conditions of subjective illness and positive context of treatment, doctors would expect patients to experience clinical improvement. So, when these conditions are present the doctor is not in equipoise and would not consider there to be a human cost to administering placebo treatments. The collective good of placebo treatments in clinical practice would be limited in the sense that we can only consider placebo treatments to be efficacious for subjective illness (so not illness such as Malaria or cancer) and because placebo treatments have to be administered in conjunction with orthodox treatments in order to retain the public trust in the institution of medicine necessary to induce some placebo effects. But despite these limitations, there are clinical benefits to placebo treatments in clinical practice. Placebo treatments can restore the autonomy loss a patient suffers as a result of their illness. Combining placebo treatments with orthodox treatments enables medicine to fulfil its duel role as science and succour. Moreover, placebo treatments can be given suitable objectivity for clinical practice. There are reasons why not permitting placebo treatments may be bad for the collective. Placebo treatments are already used routinely in clinical practice and for this use to continue unchecked perhaps risks patient safety. Similarly, much of the power of Alternative Medicine could stem from its use of placebo treatments. Alternative Medicine endangers patients by engaging with pseudoscience. If placebo treatments in clinical practice is in the collective good then clinical research into placebo effects will be in the collective good because it would be in the collective good to know and understand as much as possible in order to fashion placebo treatments that maximise clinical improvement. Research into placebo effects will help orthodox medicine to understand how better to account for their impact on results. If placebo effects exacerbate or obscure treatment effects by potential orthodox treatments then patient safety could be risked in clinical practice because they are not actually receiving the best most appropriate treatment. I suggest that these benefits, even together with the limitations, point to placebo treatments in clinical practice being in the collective good.
The rule of thumb plausibility for the efficacy of placebo treatments in clinical practice and that such use is in the collective good makes the use of placebo treatments in clinical practice permissible.
•Allan, L. and Siegel, S. ‘A Signal Detection Theory Analysis of The Placebo Effect’, Evaluation & The Health Professions, Vol. 25, No. 4, December 2002, p. 410-420
•Benedetti et al. ‘Neurobiological Mechanisms of the Placebo Effect’, The Journal of Neuroscience, November 9, 2005, 25(45): 10390-10402
•Benedetti, F (2009), Placebo Effects: understanding the mechanisms in health and disease, Oxford University Press.
•Brody, H (2002), ‘The Placebo Effect: Implications for the Study and Practice of Complementary and Alternative Medicine’, The Role of Complementary & Alternative Medicine: accommodating pluralism, ed. Callahan, D., Georgetown University Press.
•Brooks, M (2010), Thirteen Things That Don’t Make Sense, Profile Books Ltd.
•Collins, H and Pinch, T (2005), Dr Golem: How To Think About Medicine, University of Chicago Press.
•Fässler et al. ‘Placebo Interventions in Practice: a questionnaire survey on the attitudes of patients and physicians’, British Journal of General Practice, February 2011; 61: 101-107.
•Goldacre, B (2009), Bad Science, HarperCollinsPublishers.
•Golomb et al. ‘What’s in Placebos: Who Knows? Analysis of Randomised Controlled Trials, Annals of Internal Medicine (2010), Vol. 153, No. 8; 153:532-535.
•Honderich (2005), The Oxford Companion to Philosophy, Oxford University Press.
•Howick, J (2008), Philosophical Issues in Evidence-Based Medicine: Evaluating the Epistemological Role of Double Blinding and Placebo Controls, PhD Thesis submitted to LSE.
•Howick, J. ‘Escaping the Placebo Prison’, British Medical Journal, (2009) Vol. 338, Issue 7704.
•Hróbjartsson, A and Brorson, S (2002), ‘Interpreting Results from Randomised Clinical Trials of Complementary/Alternative Interventions: The Role of Trial Quality and Pre-Trial Beliefs’, The Role of Complementary & Alternative Medicine: accommodating pluralism, ed. Callahan, D., Georgetown University Press.
•Kaptchuk et al. ‘Components of placebo effect: randomised controlled trial in patients with irritable bowel syndrome’, British Medical Journal, (2008); 336:999.
•Kaptchuk, T. ‘The Placebo Effect in Alternative Medicine: Can the Performance of a Healing Ritual Have Clinical Significance?’ Annals of Internal Medicine, June 4 2002, Vol. 136, No. 11 817-825.
•Nitzan, U and Lichtenberg, P ‘Questionnaire Survey on the Use of Placebo’, British Medical Journal, (2004); 329:944-6
•Owens et al. ‘The irritable bowel syndrome: long-term prognosis and the physician-patient interaction’, Annals of Internal Medicine, January 15 1995, Vol. 122, No. 2 107-112.
•Penston, J (2003), Fiction and Fantasy In Medical Research: The Large-Scale Randomised Trial, The London Press Ltd.
•Private Eye article, ‘Eye Told You So’, Medicine Balls, 7.03.2003, Private Eye.
•Spiegel, D. and Harrington, A. ‘What is the Placebo Worth? The Doctor-Patient Relationship is a Crucial Part of Its Value’, British Medical Journal, Vol. 336, Issue 7651
•Spiers, A. ‘Placebos are Valuable’, British Medical Journal (2009),http://www.bmj.com/content/338/bmj.b1568.full/reply (09.03.2011).
•Sullivan, M. ‘Placebo Controls and Epistemic Control In Orthodox Medicine’, The Journal of Medicine and Philosophy (1993) 18: 213-231.
•Truog, R. ‘Informed Consent and Research Design in Critical Care Medicine’, Critical Care 1999, Vol. 3.
•Whitmarsh, T (2002), ‘The Nature of Evidence in Complementary and Alternative Medicine: Ideas from Trials of Homeopathy in Chronic Headache’, The Role of Complementary & Alternative Medicine: accommodating pluralism, ed. Callahan, D., Georgetown University Press.