Placebos, placebo effect, and the response to the healing situation: the evolution of a concept

Epilepsia, 42(12):1614–1625, 2001
Blackwell Science, Inc.
International League Against Epilepsy
Placebos, Placebo Effect, and the Response to the Healing *Yiannis G. Papakostas and †Michael D. Daras *Department of Psychiatry, Athens University Medical School, Athens, Greece; and the †Neurological Institute, Columbia-Presbyterian Medical Center, New York, New York, U.S.A. Summary: In spite of its impressive progress, medicine has
been established as the gold standard for the assessment of any been strongly criticized for relying on its modern biomedical new treatment. Recently a new trend regarding placebos seems tradition to the neglect of the psychosocial aspects of health.
to have emerged. The placebo and other nonspecific effects This neglect may account for patients’ dissatisfaction and even- elicited by the “healing situation” have been independently tual use of alternative health approaches. The concept of pla- subjected to scientific study. Progress in this area may promote cebo has sustained dramatic “protean” metamorphoses through useful clinical applications, enabling physicians to broaden the ages. For centuries, placebos have been regarded as pow- their perspectives on the healing process. We present the his- erful deceptive therapies. From the middle of the twentieth torical changes of the concept of placebo and the ethical issues century, however, conventional medicine has used placebos as raised by their use. Key Words: Placebo—Nocebo—
methodologic tools to distinguish the specific from the nonspe- Randomized controlled trials—Equivalence trials—Ethics— cific ingredients in treatments. In modern medical research, the double-blind, placebo-controlled, randomized clinical trial has Socrates: “good doctors . . . (it) is not possible for them to response to the healing situation, as well as to discuss attempt to treat the head by itself, apart from the certain ethical issues raised from their use.
body . . . or the body without the soul . . . . and the treat-ment of the soul by means of certain charms, and thesecharms are words of the right sort . . . ” Plato, Charmides (1) STAGE I: PLACEBOS
The history of placebo follows the same pattern as other Basic concepts
phenomena in science recognized as “artifacts” (2,3). For In a group of untreated patients with depression, it is centuries, healers, including physicians, have used inac- expected that a certain percentage will improve after a tive placebos or “dummy pills” worldwide, although the few weeks. With the administration of an inactive pill, official medical community ignored them (4–6). Next, the percentage may be higher, and even higher if an the nonspecific placebo effect was studied, isolated, and antidepressant is used. This example illustrates at least eliminated in the randomized controlled trials as a nui- three hypotheses that may explain patients’ improvement sance factor, in contrast to the “specific” pharmacologic after a certain therapy (12–16). First, a whole spectrum effect (7,8). Finally, the placebo effect or “the response of “nonspecific” factors such as the natural course of the to the healing situation” has recently become a subject of disease, the regression toward the mean, and other overt study in its own right, along with the exploration of its or covert influences may contribute to this healing through the time course of an untreated illness (17). Sec- The purpose of this article is to present the historical ond, in the case of the active drug, the patients improved transition from placebos, to the placebo effect, and to the because of the “specific” or “characteristic” effects of thetreatment that can be scientifically studied, isolated, and predicted through specifically designed studies, the ran- Address corresponding and reprint requests to Dr. M. D. Daras domized controlled trials (RCTs) (7,8). Finally, in the at Columbia-Presbyterian Medical Center, The Neurological Institute, case of the “inert” therapy, it is obvious that the inactive 710 West 168th Street, New York, NY 10032, U.S.A. E-mail:[email protected] pill, because it lacked a specific pharmacologic effect, acted therapeutically through its “symbolic power” and includes objective end points (such as blood pressure, its impact on the patient’s imagination, beliefs, expecta- EEG, serum hormone concentration), whereas the im- tions, and emotions (7,12). Patients were under the im- provement can be long lasting (10,30,31). Double-blind pression that they were taking an active or specific drug RCTs have shown that placebos can have healing effects with known efficacy for their condition. This therapy in diverse conditions as angina pectoris (32), epilepsy with the inert sugar pill lacking specific activity repre- sents a typical example of what is called placebo The notion, however, of a powerful placebo has not therapy, and the “dummy” pill is called a placebo (from been unchallenged. By reanalyzing the classic Beecher’s the Latin: “I shall please”) (4). Besides the administra- review (25), Kienle and Kiene (35) pointed to Ն19 pos- tion of a drug, this nonspecific “therapy” can take the sible reasons for the changes in the placebo-treated form of a physical (i.e., manipulation), psychological groups. They argued that the significance of the placebo (i.e., conversation), or any other form of therapy.
effect has been grossly overestimated and that the re- The elicited and measurable response after the admin- ported placebo effect may not be a distinctive one but istration of a placebo, the ingestion of the inert pill, is may be part of a spectrum of care and treatment (35).
called a placebo response. It may not necessarily be elic- They further consider these reports as misleading, per- ited by the symbolic characteristics of the inert therapy petuating widespread misconceptions about the magni- but may constitute a “time effect,” such as spontaneous tude of the placebo effect. Indeed, the placebo response improvement (18). Placebo effect is the nonspecific psy- rates quoted from clinical trials may indiscriminately in- chophysiologic effect produced by placebos through the clude both the spontaneous and natural recovery from symbolic significance of the administered therapy (5).
disease and the “true” placebo effect (15,17,35). Review The “true” placebo effect may be estimated as the dif- of clinical trials with patients randomly assigned to either ference between the magnitude of the placebo response a placebo or no treatment showed small differences in (“perceived placebo effect”) minus the magnitude of the clinical outcome, leading to the conclusion that there is “time effect” on untreated patients (“other nonspecific little evidence to support the assertion that placebos have effect”) (17). In practice, however, distinguishing the powerful clinical effects (36). Therefore, far from being “true” placebo effect from the “perceived placebo effect” resolved, the question of the existence and magnitude of is not easy (13,19,20). Actually, because these terms are a true placebo effect is valid and should be addressed by hard to define precisely (13,21), it was suggested that Nocebo (from the Latin: I shall harm) and nocebo The ethics of giving placebos
effects are terms less familiar to physicians, used in cases There are at least two distinct uses for placebos: in when patients attach negative meanings and emotions to research as tools for reducing bias in clinical trials and a treatment perceived as harmful. In contrast, the defini- in clinical practice for therapeutic purposes. In both tion of a treatment as “placebo” depends on the degree of cases, major ethical issues may arise whenever two key patients’ positive (beneficent) meanings and emotions to- considerations or requirements, beneficence and au- ward it (22). The most common adverse effects of systemic tonomy, are violated. (37). Beneficence refers to the phy- placebos are headache and drowsiness (23). Placebos used sician’s fundamental duty to respond according to the in healthy volunteers usually cause negative effects because best interests of an individual patient by providing the healthy persons anticipate no medical benefit (24).
best available treatment and optimal care to this patient In dealing with placebos, at least two important ques- (37,38). This old principle dates back to the Hippocratic tions arise: first, does a clinically significant “true” placebo code urging physicians to provide benefits to the patient effect exist, and second, because their use, either in clinical and balancing them against risks: “As to diseases, make practice or in research, almost invariably implies some a habit of two things, to help, or at least to do not harm” form of deception, what are the ethical issues involved? (␣␴␬␧ı´␯ ␲␧␳ı´ ␶␣ ␯␱␴␩´␮␣␶␣ ␦␷´␱, ␻␸␧␭␧´␧␫␯ ␩´ ␮␩ ␤␭a´␲␶␧␫␯) (39). Conflicts may arise whenever the phy- Does a clinically significant “true” placebo
sician has the additional role of a researcher whose pri- effect exist?
mary concern is to produce valid, generalized knowledge It is generally accepted that placebos are associated that may benefit future patients (40–42). Hence, research with powerful nonspecific effects (3,25–29). According differs from optimized individual clinical treatment as it to Oh (27), placebo is “the most effective medication typically contains an element of increased risk. Admin- known to science, subjected to more clinical trials than istration of placebo instead of standard treatment, as it any other medication, yet generally always does better usually occurs in research, may be considered unethical than anticipated. The range of susceptible conditions ap- because it places some people at risk for the good of pears to be limitless.” The response is not limited to others, future patients, and society (43). According to the subjective (such as pain or anxiety) end points but also Hippocratic code and the Declaration of Helsinki, the physician’s role as a healer must take precedence over the brain associated with the perception of a stimulus as his role as a researcher. The advancement of scientific “meaningful” in a healing-related sense, thus producing knowledge, therefore, must always be secondary to the end-organ changes (57). Various “mind–body” healing concern for the individual, whose well-being is the phy- pathways of the brain such as the endogenous opioid sician’s highest priority. Nevertheless, physicians also (58), the neuroendocrine (59), and the psychoneuroim- have an ethical obligation to find better treatments for mune pathways (60) have been implicated. Mental and diseases (44); thus one duty may be at times in discord emotional stimuli related to the placebo phenomena can reach the immune system through the autonomic nervous Autonomy refers to the patient’s right to be informed system and the neuroendocrine outflow via the pituitary, about treatment options, the pros and cons of each, and to linking the brain to the immune system (60,61).
decide about health matters without coercion (37,38).
Although nothing is expected to be absorbed, distributed, For centuries, beneficence, in the form of benevolent and metabolized by the ingestion of an “inert” substance, paternalism, not truthfulness, was the modus operandi of placebos nevertheless display certain “pharmacologic,” physicians (7,45). During the second half of the twenti- behavioral, or physiologic characteristics noted in active eth century, however, autonomy became a major issue, drugs (62). “Peak effects,” “cumulative effects,” “ carry- and the new medical ethics replaced “beneficence” with over effects,” “tolerance,” and “dependence” are some of “informed consent” (7). Violation of autonomy (decep- the subjective or behavioral characteristics of placebo tion) may occur whenever placebos are used without administration (63,64). Similarly, placebo-generated per- adequate patient information and consent. Because some ceptions can result in the production or release of active form of deception is believed to be necessary for place- materials, or may influence the enterohepatic circulation bos to exert their effect, the patient’s autonomy is threat- or the local blood flow and the rate of the agonist turn- ened by a (benevolent) paternalistic attitude that may over, thus leading to physiologic alterations (62).
influence physicians to administer placebos against thepatient’s knowledge and consent (45). Deception can be STAGE TWO: PLACEBO EFFECT
detected by evaluating the patient’s understanding and For centuries, placebos were the most physicians voluntary participation, the two main components of the could to offer their patients and through them maintain modern informed-consent process (46). The issue of au- their reputation (5,26). In that regard, the history of tonomy becomes even more pressing in individuals with medicine was the history of placebos (5). After World questionable capacity to consent (47).
War II, as medicine adopted biologic principles, the “art”of healing was replaced by the “science” of therapeutics Mechanism of the placebo action
(65,66), and the traditional beneficence paternalism As an example of the mind–body relations and com- yielded to informed consent (67). As a result, the use of patible with a the holistic view (27), the placebo effect placebos for therapeutic purposes started to decline.
offers unique opportunities to study the interplay of en- Treatments with “specific” action replaced “nonspecific vironmental, mental, and brain processes (48). Because therapies,” and the principle of autonomy overrode the the placebo effect seems to exist across all medical dis- principle of benevolence. Therapeutic or even diagnostic ciplines, a single universal theory that can explain all use of placebo in noninformed patients, with rare excep- placebo effects is unlikely (4). Hence, several psycho- tions in well-defined cases, is considered ethically unac- logical and biologic explanations have been proposed ceptable and may have detrimental effects on the (2,50). Currently two psychological theories, the classic patient–therapist relationship (20,41,68–70). Nonethe- conditioning theory and the expectation theory, have less, even today there are instances in which clinicians been used to account for the placebo effect.
prescribe placebos, intentionally or unintentionally, in a According to the classic conditioning theory, a variety covert (deceptive) form and frequently in the least- of inert substances, procedures, persons, or places, if appropriate cases (71). This practice illustrates ignorance associated with effective treatments (e.g., penicillin, in- and prejudice (41) and can be considered nothing more sulin, morphine) come to function as conditioning than “a face-saving disguise for medical impotence” stimuli eliciting similar responses to the unconditioning (72). For example, controlled studies have shown that stimuli (51–53). Conversely, the expectation theory pro- vitamin C for cold prevention is ineffective, as are small poses that a special type of cognition called expectations doses of antidepressants, tricyclics in particular, for de- or expectancies triggers the placebo response (54,55).
pression; yet both are common clinical practices (73,74), It has been suggested that specific expectations, which occasionally leading to serious side effects (75). It should reflect knowledge about the therapeutic agent, the circum- be emphasized, however, that even if placebos were stances under which it is administered, and the condition to given in a direct, nondeceptive way (i.e., patients were be treated, are directly linked to the placebo effect (56).
informed about the inertness of the given substance), a The biologic approach aims at identifying pathways in placebo effect might still be obtained (76,77). This raises the possibility that in appropriate clinical circumstances, (93). The first outcome suggests a specific pharmaco- a placebo could be among the therapeutic choices if of- logic effect exhibited by the new drug, whereas the op- fered to patients in a nondeceptive way (3,28). The clini- posite suggests lack of such an effect and/or even cal applications, however, of the nondeceptive use of pla- adverse effects. A demonstration of a difference can cebo, although theoretically intriguing, is open to question.
serve as compelling evidence, defined as evidence that Before World War II, a medical therapy was legiti- “on its own, without the need for any supplementary mate if it had a positive outcome. But because both inert sanguine assumption based on assertions about condi- and active therapies were observed to produce results tions external to an experiment, leads to one and only beyond their predicted physiologic properties, the need logical conclusion about the meaning of the evidence to study and control them was formally acknowledged adduced in the RCT. . . ” (93). The absence of difference and pursued. Subsequently a new trend developed in is not so informative, because the new drug may lack medicine. The long-standing tradition of clinical use of efficacy, but not necessarily so, for various reasons (93).
placebos to promote the healing process was replaced by The main disadvantage of an inert placebo use in the investigators’ attempt to disentangle, isolate, and re- RCTs is the risk and consequences of a nonspecific treat- move the placebo and other nonspecific effects from the ment in the placebo-arm patients. Apparently, “the ethi- “specific” action (10,78). Beginning with the streptomy- cality of a placebo arm in a trial is inversely related to the cin trial (79), a new standard was established, and a availability of an excellent standard treatment, the seri- medical intervention was scientifically acceptable only ousness of the condition under study, and the probability if it successfully passed the powerful screening proce- of irreversible harm resulting from the denial of thera- dure of RCT (7,8). In spite of their limitations (80–82), peutic benefit from conventional therapy” (94). Two ad- RCTs are widely acknowledged as the most secure ditional issues not unique to placebo-controlled trials source for valid interpretation of the treatment effect apply to any RCTs not involving placebos. First, the (83,84) and the first step in improving the efficacy of blind RCTs include an element of deception, as patients medical care (85). The random assignment of individuals do not know the type of treatment they receive. This, involved in RCTs, the sole defining characteristic of however, may be compensated for by the fact that opti- these trials (86), is the best way to achieve a balance mally consenting patients are considered “partners” in between groups for known and unknown factors that the deception (95). The rationale of the placebo-arm use influence outcome, so that any difference can be ex- in double-blind RCTs is to maintain blinding and mini- plained only by the treatment (87). It permits a compari- mize bias (96). Similarly, concerns about whether the son of the effects of two or more treatments “ceteris consent to participate was obtained under the optimal paribus” (all other facts being equal) (88).
conditions of understanding and voluntary participation Randomized controlled trials
in a placebo-arm RCT are legitimate, but also apply to The RCT technology offers two distinct ways to show the effectiveness of a new treatment: comparing it witheither a placebo or an active therapy. Each approach has Active control “equivalence” trials These trials are used to show that the new therapy is equivalent to or not worse than a known effective treat- These trials aim to show that the experimental treat- ment, either as an alternative or as having advantages in ment is superior to the “nonspecific” elements solidified safety, convenience, or cost (97,98). Because “equiva- and assessed through the use of a placebo arm. They lence” or “noninferiority” is the most probable expecta- intend to prove if the experimental drug has any specific tion (superiority is seldom asked), these trials have been or pharmacologic effect (89). The nonspecific elements called “active controlled equivalence studies” (99).
in RCTs often include many influences, such as support, Unfortunately, a study that successfully shows attention, education about the treatment or the illness, “equivalence” is uninformative, as it does not by itself and so on (49,90). Hence, in placebo-controlled RCTs, demonstrate that both treatments were effective. It also the experimental drug is compared with the “nonspecific can mean that both were ineffective in the study treatment” and not with “nothing” (16,42,91). It is pos- (93,97,100–103). The conclusion that the new treatment sible that the placebo effect elicited under these circum- is effective on the basis of its similarity to the active stances may be even more powerful than that of control requires the critical assumption that the active treatment had an effect in that particular study. Only in Assuming that a placebo-arm study was designed and trials in which “sensitivity” is proven (that is the standard performed in a scientifically reasonable manner, a statis- treatment is superior to placebo) can correct conclusions tically significant difference favoring either the experi- of equivalence be made. But such an assumption cannot mental drug over placebo or vice versa is informative be maintained, however, as it is known that one third to one half of RCTs are negative, even when the standard make sacrifices for the good of others (violation of the therapy is known to be clinically effective (104). Clearly, beneficence principle). These volunteer patients are ex- a third group taking placebo is needed. Conversely, even posed to the risks of harm from lack of treatment during if the experimental therapy is less effective than the their participation in the clinical trial. To support their active one, the possibility exists that it may still have arguments, these critics usually cite the following sen- clinical use for a subpopulation or because of a better tence from the Declaration of Helsinki in 1997: “In any side-effect profile. These advantages cannot be known medical study, every patient including those of a control group should by assured of the best proven diagnostic Active control equivalence trials can be informative in and therapeutic method” (117). Furthermore, the recent many areas such as treatment of cancer, infectious, and fifth revision of the Declaration of Helsinki in 2000 cardiovascular diseases, provided that assay sensitivity is (118), that medical researchers and ethicists have agreed not in doubt (97). Temple and Ellenberg (97) claimed on, emphasizes in much clearer terms than ever before that “the larger the effect size, the less study-to-study the already restricted conditions allowing the use of pla- variability in outcomes, and the fewer the instances of cebo in RCT. Researchers, drug companies, and journal unexplained failure of the control agent to show superi- editors may be profoundly affected, as the new principles ority to placebo in well-controlled studies, the more per- are expected to tighten the rules for clinical research and suasive is the case for using this design.” The active use of placebos, putting new limitations on the risks to control “equivalence” studies differ from “superiority” which patients may be exposed (119–122). Ethicists also studies (i.e., studies designed to detect differences such raise objections to the use of placebo-controlled RCTs on as the placebo RCTs), in several important ways. The methodologic grounds. Several researchers (89,90,93,94, design and analysis of the former, as compared with the 97,100–102,123,124), however, have firmly disputed latter, are more complex and less precise scientifically their assertion that placebo-controlled studies are not sci- (98,105). Among their disadvantages, the approval of drugs that are ineffective or have serious adverse effectsremains the main concern (75,93,106). In addition, more The future of placebos in research
patients are exposed to a possibly ineffective and dan- The World Medical Association (WMA) decision is at gerous drug as compared with placebo-controlled RCTs.
odds, and clashes with the opinion of those who consider They also may incur significant cost increase and delay these recommendations impractical, interconflicted, and in the process of drug development (44). These draw- unattainable (90,94,97,124–127). A literal interpretation backs raise ethical concerns, as it is obviously unethical of the Declaration means that if an effective treatment to engage human subjects in research lacking scientific exists for any condition, however benign it may be, the validity and unable to suggest interpretable conclusions.
use of placebo control is unethical. Several problems The most obvious advantage of the active control RCTs arise with such a strict interpretation: first, the use of is that they pose no ethical dilemmas, if there is a “genu- placebo control will be permissible in only a few cases ine uncertainty” about which of the two, or more, treat- (116,128). The main bulk of research will be instead ments tested is preferable [a condition known as conducted through active control equivalence trials with all the aforementioned problems, including the approvalof ineffective therapies. As Khan and Brown (90) stated, Placebo-controlled trials disputed
“The elimination of placebo controls in favor of non- Researchers and U.S. Food and Drug Administration inferiority trials is not only bad science, it is bad for the regulators (106) prefer and/or require placebo-controlled public’s health.” Second, a strict interpretation of the RCTs at least during the initial phase of establishing the declaration would exclude not only the use of placebo, efficacy and safety of a new drug and its superiority but also active-control trials, because patients taking the against placebo. The use of randomized placebo-controlled, investigational drug instead of the established therapy double-blind clinical trial is considered the gold standard would not receive the “best” proven, current treatment in research designs, as it is believed to provide the most (94,97). Any nontherapeutic or nonmedically indicated definite test of efficacy (93,98). This doctrine, however, investigational procedure, such as magnetic resonance has recently been disputed (108,109), and the RCT trials imaging (MRI) or positron emission tomography (PET), using placebo arms have come under increasing attack might also be restricted in clinical research if the same on ethical and, perhaps less convincingly, methodologic standard of individualized beneficence applied as in grounds (110–116). Critics of the placebo-controlled clinical medicine (42). Finally, a strict interpretation RCTs consider it unethical to give patients placebos seems to preclude the free will of patients who, for their when an approved therapy exists, even if the required own reasons, choose to participate, such as self-interest, informed consent has been obtained, because patients are altruism, or the desire to “be treated as special patients” asked to participate in a medical trial with the intent to (129–131). This may appear as another form of revived beneficent “paternalism” conflicting with and overpow- the RCT, when randomization is inappropriate, impos- ering the “autonomy” principle (100).
sible, or unacceptable. Thus in spite of their limitations, The confusion over the use of placebo forced the well-designed observational studies can provide valuable WMA recently to establish a task force to investigate the information, because their results may not overestimate issue and refine changes to the Declaration of Helsinki the magnitude of the therapeutic effect, as previously (132). Meanwhile, the International Conference for Har- monization (ICH) released a statement that considers It seems but erroneous that the use of placebos in ethical the use of placebo-arm RCT, even if an effective RCT, the principal research tool or the gold standard, treatment is available for the condition under study, pro- through which medicine in the twentieth century had vided that no serious harm is expected from withholding made all its great therapeutic discoveries, is seriously available treatment, and patients are fully informed about challenged (7). Some expect that the use of placebos in existing therapies and consequences of delayed treatment research will progressively change or decrease as medi- (133). This report is in agreement with the appraisal of cal knowledge accumulates (105,114), whereas others Temple and Ellenberg (97) from the FDA on the ethical believed that the continued use of a placebo control and scientific issues of placebo-controlled trials.
It seems that a universal agreement exists that place- bos can certainly be used, when no effective treatment STAGE III: RESPONSE TO THE
exists, and should not to be used in life-threatening situ- HEALING SITUATION
ations, when standard treatment exists (37,97,116,128).
The task, therefore, of the research community is to In spite of its impressive progress, medicine has been evaluate specific cases that fall in between, balancing strongly criticized that by relying on its modern biomedi- ethical and scientific considerations. Examples of the cal tradition has neglected the psychosocial aspects of types of data that the research community may need, health (142). The price is significant, leading among oth- come from the work of Al-Khatib et al. (134) and Khan ers to patients’ dissatisfaction and use of alternative et al. (135). They evaluated through meta-analysis the health approaches associated with allegedly powerful risks of using placebo controls as opposed to the active nonspecific effects (67,143,144). The last approaches are drug condition in short–term treatment of mild to mod- often described with such desirable characteristics as erate hypertension and acute treatment of depression, Hippocratic, integrative, preventive, and holistic, whereas correspondingly. They both found no additional risks conventional medicine suffers by being considered ga- from the use of the placebo as compared with those of the lenic, analytic, curative, and specific (145).
The accumulated knowledge on placebo and the other Furthermore, researchers should consider reducing the nonspecific effects of medical care may offer medicine a number of patients exposed to placebos. For example, unique opportunity scientifically to explore and expand because measurement errors are compensated by increas- the therapeutic processes within a broader biopsychoso- ing the number of studied subjects (136), improvement cial framework and, therefore, a more integrative, yet of raters’ reliability through systematic training or use of scientific, approach. Physicians may learn how to use multiple raters will reduce placebo requirements. Next, these effects in a nondeceptive and ethically acceptable adaptive allocation designs aiming at reducing the num- manner. This can be possible because the administration ber of patients receiving either ineffective and/or poten- of a placebo is neither a necessary nor a sufficient con- tially dangerous treatments may be considered (137). In dition for the initiation of a placebo effect (3,146). Not the “play the winner” design, the probability of ending only placebos but also any feature of the nonspecific up in one arm of the trial or the other depends on how elements (concomitants) of the physician–patient en- well previous subjects responded to the treatment (138).
counter may recruit the healing response. Hence, it is In such a paradigm, the better the experimental treatment, probably more appropriate to speak of the “response to the greater the odds that the subsequent patient will be the healing situation” than of the “placebo response” (3).
assigned to this study arm, and, therefore, the smaller the Instead of, therefore, being conceptualized as the product number of patients taking placebos. These types of adaptive of placebo administration (5), the placebo effect can be allocation designs, however, are at their very beginning, are redefined as any effect attributable to the symbolic im- logistically complex, and lack supportive evidence (137).
portance of a treatment, treatment setting, or treatment Other strategies, common in cancer, heart failure, and process (11). This reconceptualization bypasses the con- epilepsy trials that seem to pose no ethical dilemmas, are troversial issue of the existence or the magnitude of the the “add-on” designs in which the patient’s baseline placebo effect as it encompasses all the nonspecific in- medication is maintained and then either a placebo or the fluences. In addition, because the elicited effect is not new investigational drug is added (97). Finally, clinically solely dependent on placebo administration, it has the significant evidence may come from sources other than potential of being ethically immune.
It seems that a new era in the life of an artifact has priate emotional involvement appears to be a basic, come of age, and the growing debate on ethics (87,108) nonspecific, healing-promoting component, yet neglected as well as the changing face of medical curricula (147) during the modern era. Instead, “seeing” became the pri- may be instrumental in this transition. The phenomena of mary tool in knowing and understanding during the heal- placebo and the other nonspecific effects have become ing endeavors (151). It has been estimated that the mean objects of scientific scrutiny on their own right.
length of time that elapses before doctors interrupt thepatient’s first response is 18 s (152). As Sir Theodore Harnessing the placebo effect
Fox wrote, “Lack of time made us all bad doctors” (153).
By recognizing the importance of the response to the The recent rise of medical interest in the lost tradition of healing environment, the clinician should be interested in narrative may force physicians to listen more construc- developing strategies to optimize this effect in clinical tively to their patients (154–157). Both patients’ verbal practice (3,9–11,148). The new “healing art,” however, and nonverbal communication should be taken into ac- is in its infancy, lacking a conceptually elaborated and count, with encouragement to voice their agenda con- empirically tested body of knowledge. As opposed to a cerning covert fears and hopes (158).
placebo-elicited effect that has been systematically stud- Because doctors and patients talk to each other with ied, there are only few studies of the “response to the different “voices,” it is the doctors’ duty to smooth com- healing situation” (3). It is still unknown which of the munication by explaining in clear and comprehensible nonspecific factors and to what extent each of them con- language the diagnostic, prognostic, and therapeutic as- tributes to the healing effect. Future research should ad- pects of the illness. Additionally, physicians must be dress these questions in studies comparing two situations sensitive to cultural differences and understand how pa- differing in only one nonspecific factor (14,49). The tients’ models of health and disease affect their beliefs need for a multidisciplinary approach to formulate the and behaviors (159–161). Patients vary in their desire most important questions pertinent to the nonspecific for involvement in decision making, so the degree of factors in the healing process is mandatory (2,50,92, their participation cannot be predetermined, being in need of constant revision throughout the therapeutic en- At the moment, various attempts and suggestions are deavor and monitored according to individual needs based on physicians’ clinical experience and some prin- ciples drawn from psychological, conditioning, or expec-tation theories (3,9,11). It is possible that in the future, The patient and his or her illness
this healing potential may be deliberately mobilized onan evidence-based manner or invoked in other ways, such as biochemical. Psychosocially or biochemically At one time it was thought that only a specific type of enhanced, the response to the healing situation can be individual responded to placebo, a consistent placebo incorporated into standard pharmacologic and surgical reactor (165,166), but such an individual has not been treatments to act synergistically with them, thus pro- identified, and no type of personality is prone to a con- viding the medical healing art with a strong integrative sistent placebo-elicited effect (167). Hence, physicians and holistic trend. At least theoretically, the placebo ef- must try to elicit the “placebogenic” and/or “nocebo- fect depends on the patient, the dispensing physician, genic” aspects of each individual patient with a particular and prominent between them, the quality of their rela- disease, positive and negative nonspecific effects in the patient’s previous beliefs and experiences with treat-ments (background placebo variables), as well as atti- Physician–patient relationship
tudes toward the present illness, and faith in or distrust of The value of a sound patient–physician relationship a particular physician or form of treatment (situational has been historically recognized and is considered the placebo variables) (9,168). This type of history will pro- cornerstone for achieving, maintaining, and improving vide the physician with valuable information to deter- health (149). We do not know the key elements of such mine the conditions favoring a placebo effect.
an ideal relationship and whether they influence the pla-cebo effect.
During the past three decades, the patient–physician Although the establishment of predictable placebo re- relationship has changed more than during the preceding sponse patterns to particular disorders has been proven 25 centuries. It cannot any longer be seen as a kind of difficult, the physician should be aware of the limits and parent–child relationship, but rather as one between two potentials of placebo use in each illness. The most thor- morally autonomous adults (150). And although the de- oughly studied placebo-responsive disorders are chronic sirable characteristics change over time, the respect for pain and psychiatric illnesses (23). Generally, the pres- the patient’s autonomy is the core of such an ideal rela- ence of anxiety and pain, the involvement of the auto- tionship (149). Empathic “listening” that requires appro- nomic nervous system, and the immunobiochemical processes are believed to respond favorably to placebo, (172), or cholesterol for ischemic heart disease (173), whereas hyperacute illnesses (i.e., heart attack), chronic seems to be relevant to the issue of referral.
degenerative diseases, or hereditary diseases are ex-pected to resist (27). Conversely, different diseases invoke different meanings for each individual. An en- Practitioners tend to prescribe more medications than counter with a physician is most likely to produce a patients expect, possibly because they fail to consider the placebo effect when it changes the meaning of the illness patients’ reason for coming (174). From a placebo per- experience for that individual in a positive direction. This spective, a prescription, if needed, should preferably is most likely to occur when the patient is listened to and combine the specific ingredients with nonspecific, receives an explanation for the illness that makes sense; placebo-promoting concomitants revealed during the the patient feels care and concern by the physician and placebo history-taking procedure (9). The prescription others in the environment, and experiences an enhanced should be offered with realistic optimism because the sense of mastery or control over the illness or its symp- placebo effect depends on the confidence of the therapist toms (11). The physician, however, must be cautious (175). Based on the placebo-promoting effect of predic- not to raise false hopes that may produce further disap- tive accuracy, the physician should inform the patient about the time course of the effect and possible adverseeffects of the treatment (9). A patient who responded The physician and the treatment setting
well to a particular treatment in the past should be con- The presence of a physician is the most common fac- sidered to receive the same in case of a recurrence, es- tor able to induce a placebo effect, the most frequently pecially if the disorder is relatively minor, definite used “drug” in general practice (57). Furthermore, a treatment does not exist, the prior treatment was not dan- well-informed physician, without misconceptions, gerous, or the patient is convinced that the treatment (31,169) may intentionally use placebo knowledge to optimize the placebo effect in various ways (3,9–11,51).
Perceptual characteristics of the treatment Predictive accuracy as a placebo-elicited tool The potentially important effect of these characteris- According to learning theory, there are at least two tics is recognized in RCTs, in which various measures reasons for the physician’s strong “placebogenic power” are taken to isolate them. From the plethora of these (51). First, the physician possesses treatments with “spe- factors and guided by information obtained during the cific” effects tested by RCTs therapeutic ingredients. The placebo history, physicians may choose treatments more effective a treatment, the bigger the placebo effect whose perceptive characteristics are possibly associated and the placebogenic power of the physician. Second, the with strong placebo effects. For example, a new treat- physician has exceptional diagnostic and prognostic ment, irrespective of the specific effect, appears more abilities that represent another source of placebogenesis effective than an old one (5), or capsules more potent for the patient, who may confuse them with the physi- than tablets in anxiety states (176); generally, the more cian’s healing potential. Accurate diagnosis and progno- invasive a treatment or the more actively it involves the sis are, therefore, important tools at the physician’s patient, the larger the placebo effect (177). The color of disposition, because, if properly handled, they bear sig- the pill seems to influence its effectiveness (178), but nificant possibilities for eliciting a placebo effect (3,9).
consistent trends are not universally apparent (148).
The name of the drug is generally considered to pos- Appropriate referrals for testing–reassurance sess tremendous symbolic and practical significance.
While certain referrals and diagnostic tests are neces- Creating a generic name is a scientific task, but finding a sary and self-reassuring, oftentimes ordering laboratory brand name seems to be more an art (179). The FDA tests, primarily for reassurance purposes, may not be the forbids use of words that make promises and raise ex- best strategy and definitely is not without problems pectations and hopes about cure or safety, but drugs with (170). The diagnostic tests may leave a patient more allusive names implying actions far beyond the known anxious than before unless the true fears are addressed pharmacologic actions of a given product are not uncom- (171). Thus a false-positive test result will not allow reassurance; an inconclusive result may leave doubt andanxiety, and reassurance may not be entirely successful Conceptual characteristics of the treatment despite a normal test result (170). Obtaining a test should Besides its perceptual aspects, a treatment also has be avoided if an expert clinical opinion suffices because particular meanings for each patient. Physicians, how- of the risk of residual anxiety and the epidemiologic ever, who focus their efforts primarily on properties of hazard of a false-positive or inconclusive test result drugs and their effects often ignore the psychological and (170). The current skepticism about the benefit/risk ratio behavioral aspects of taking a medication (181). Thus on of procedures like mammography for breast cancer a personal level, the patient might hold a beneficent or harmful view on the treatment offered. On an interper- 2. Harrington A. The placebo effect: an interdisciplinary explora- sonal level, the physician may be perceived by the pa- tion. Cambridge, Massachusetts: Harvard University Press, 1997.
3. Brown WA. Harnessing the placebo effect. Hosp Pract 1998;33: tient as caring, and the treatment offered becomes a token of the physician’s care; the pill also may symbolize 4. Shapiro AK. A contribution to a history of the placebo effect.
the physician’s substitution of a personal relationship 5. Shapiro AK. Placebo effects in medicine, psychotherapy, and with the patient with something inanimate and inferior.
psychoanalysis. In: Bergin AE, Garfield SL, eds. Handbook of Finally, on a social level, the sick role assigned to a psychotherapy and behavior change: empirical analysis. New patient might be interpreted either in a constructive or 6. Frank JD. Persuasion and healing. Baltimore, MD: Johns Hop- nonconstructive (i.e., a stigma) way. The physician’s task is to explore these nonpharmacologic meanings, and 7. Kaptchuk TJ. Powerful placebo: the dark side of the randomised if negative (“nocebogenic”), to try to reconstruct them.
controlled trial. Lancet 1998;351:1722–5.
8. Kaptchuk TJ. Intentional ignorance: a history of blind assessment Such interventions may have a profound effect on the and placebo controls in medicine. Bull Hist Med 1998;72:389– outcome by influencing, for example, compliance.
9. Chaput de Saintonge DM, Herxheimer A. Harnessing placebo effects in health care. Lancet 1994;344:995–8.
Large-scale trials of drugs for heart disease have 10. Benson H, Friedman R. Harnessing the power of the placebo effect and renaming it “remembered wellness.” Annu Rev Med shown that patients who adhere to treatment, even when that treatment is a placebo, have better outcomes than do 11. Brody H. The placebo response. New York: Cliff Street Books, poorly complying patients (182). In view of the impor- 12. Brody H. Placebo effect: an examination of Grunbaum’s defini- tance of compliance, several attempts to increase it have tion. In: White L, Tursky B, Schwartz GE, eds. Placebo: theory, been undertaken with rather modest success (183). Modi- research, and mechanisms. New York: Guilford Press, 1985: fication of patient’s “nocebogenic” meanings regarding 13. Gotzsche PC. Is there logic in the placebo? Lancet 1994;344: the treatment might enhance compliance and the healing 14. Kleijnen J, de Craen AJM, van Everdingen J, et al. Placebo effect in double blind clinical trials: a review of interactions with medi- Conclusions
cations. Lancet 1994;344:1347–9.
Whereas the use of placebos as methodologic research 15. Turner JA, Deyo RA, Loeser JD, et al. The important of placebo effects in pain treatment and research. JAMA 1994;271:1609–14.
tools and the decline of their clinical use characterized 16. Preskorn SH. A dangerous idea. J Pract Psychiatry Behav Health the last half of the twentieth century, a recent trend has emerged. In the third millennium, placebo phenomena 17. Ernst E, Resch KL. Concept of true and perceived placebo ef- seem to enter their third phase, a phase of maturity in 18. Fisher S. Nonspecific factors as determinants of behavioral re- which the use of placebos for research has become more sponse to drugs. In: DiMascio A, Shader RI, eds. Clinical hand- parsimonious, whereas their clinical usefulness is in de- book of psychopharmacology. New York: Science House, 1970:17–39 cline. Instead, physicians have started to learn more 19. Gotzsche PC. Placebo effects: concept of placebo should be dis- about the therapeutic environment and its healing poten- tial, so in addition to their scientific approach, they will 20. Bailar JC. The powerful placebo and the wizard of Oz [Editorial]. N Engl J Med 2001;344:1630–2.
be able deliberately to use placebo-related knowledge 21. Grunbaum A. The placebo concept in medicine and psychiatry.
synergistically to be more effective and in an ethically acceptable way. It is the physician’s duty to respect the 22. Hahn RA. The nocebo phenomenon: concept, evidence, and im- plications for public health. Prev Med 1997;26:607–11.
benefit of “the response to the healing situation” and 23. Margo CE. The placebo effect. Surv Ophthalmol 1999;44:31–44.
bring its full advantage into everyday practice. Such a 24. Rosenzeweig P, Brohier S, Ziphel A. The placebo effect in task requires knowledge of the placebo phenomena. Cur- healthy volunteers: influence of experimental conditions on theadverse events profile during phase I studies. Clin Pharmacol rently, physicians receive little or no formal training in placebo knowledge and manipulation. This lack of edu- 25. Beecher HK. The powerful placebo. JAMA 1955;159:1602–6.
cation can be inferred from the many medical myths 26. Brody H. The lie that heals: the ethics of giving placebos. Ann surrounding the placebo phenomenon (31). A curriculum 27. Oh VMS. The placebo effect: can we use it better? BMJ 1994; for medical students and residents covering the concepts and clinical implications of placebo phenomena would 28. Brown WA. Placebo as a treatment for depression. Neuropsycho- pharmacology 1994;10:265–9.
fill in the gaps in conventional training and allow phy- 29. Enserink M. Can the placebo be the cure? Science 1999;284: sicians to broaden their perspective on the healing pro- 30. Gowdey CW. A guide to the pharmacology of placebos. Can Med 31. Wall PD. The placebo effect: an unpopular topic. Pain 1992;51: REFERENCES
32. Benson H, McCallie DP Jr. Angina pectoris and the placebo 1. Plato XII. (WR Lamb, transl.). Charmides, Alcibiades, Hippar- effect. N Engl J Med 1979;300:1424–9.
chus, the Lovers, Theages, Minos, Epinomis. Cambridge, MA: 33. UK Gabapentin Study Group. Gabapentin in partial epilepsy.
Harvard University Press, 1988:19–21.
34. Downer SM, Cody MM, McCluskey P, et al. Pursuit and practice 61. Kiecolt-Glaser JK, Glaser R. Psychoneuroimmunology: can psy- of complementary therapies by cancer patients receiving conven- chological interventions modulate immunity? J Consult Clin Psy- tional treatment. BMJ 1994;309:86–9.
35. Kienle GS, Kiene H. The powerful placebo effect: fact or fiction? 62. Weiner M, Weiner GJ. The kinetics and dynamics of responses to J Clin Epidemiol 1997;50:1311–8.
placebo. Clin Pharmacol Ther 1996;60:247–54 36. Hrobjartsson A, Gotzsche PC. Is the placebo powerless? An 63. Lasagna L, Laties VG, Dohan JL. Further studies on the “phar- analysis of clinical trials comparing placebo with no treatment.
macology” of placebo administration. J Clin Invest 1958;37: N Engl J Med 2001;344:1594–602.
37. Simon R. Are placebo-controlled clinical trials ethical or needed 64. Suchman AL, Ader R. Classic conditioning and placebo effects in when alternative treatment exists? Ann Intern Med 2000;133: crossover studies. Clin Pharmacol Ther 1992;52:372–7.
65. Collier J. Confusion over use of placebos in clinical trials. BMJ 38. Beauchamp TL, Childress JF. Principles of biomedical ethics. 4th ed. Oxford: Oxford University Press, 1994.
66. Spiegel H. Nocebo: the power of suggestibility. Prev Med 1997; 39. Hippocrates. Epidemics. Loeb Classical Library (WHS Jones, transl.). Cambridge, MA: Harvard University Press, 1984:164.
67. Kaptchuk TJ, Eisenberg DM. The persuasive appeal of alternative 40. Hellman S, Hellman DH. Of mice but not men: problems of the medicine. Ann Intern Med 1998;129:1061–5.
randomized clinical trial. N Engl J Med 1991;324:1585–9.
68. Bok S. The ethics of giving placebos. Sci Am 1974;231:17–23.
41. De Deyn PP, D’Hooge R. Placebos in clinical practice and re- 69. Price JR. The placebo effect: use outside trials may be unethical.
search. J Med Ethics 1996;22:140–6.
42. Miller FG. Placebo-controlled trials in psychiatric research: an 70. Devinksy O, Fisher R. Ethical use of placebos and provocative ethical perspective. Biol Psychiatry 2000;47:707–16.
testing in diagnosing nonepileptic seizures. Neurology 1996;47: 43. de Castro LD. Exploitation in the use of human subjects for medical experimentation. Bioethics 1995;9:259–68.
71. Dixon M. The placebo effect: wrong patients receive treatment.
44. Zipursky RB, Darby P. Placebo-controlled studies in schizophre- nia: ethical and scientific perspectives: an overview of conference 72. Pickering WG. Does medical treatment mean patient benefit? proceedings. Schizophr Res 1999;35:189–200.
45. Rawlinson MC. Truth telling and paternalism in the clinic: philo- 73. Truswell AS. Ascorbic acid and colds. N Engl J Med 1986;315: sophical reflections on the use of placebos in medical practice. In: White L, Tursky B, Schwartz GE, eds. Placebo: theory, research, 74. Donoghue JM. Sub-optimal use of tricyclic antidepressants in and mechanisms. New York: Guilford Press, 1985:403–18.
primary care. Acta Psychiatr Scand 1998;98:429–31.
46. Alderson P, Goodey C. Theories of consent. BMJ 1998;317: 75. Laporte J-R, Capella D. Useless drugs are not placebos: lessons from flunarizine and cinnarizine. Lancet 1986;ii:853–4.
47. Shore d, Hyman SE. An NIMH commentary on the NBAC report.
76. Park LC, Covi L. Nonblind placebo trial: an exploration of neu- Biol Psychiatry 1990;46:1013–6.
rotic patients’ response to placebo when its inert content is dis- 48. Kazdin AE. Pills or placebos? Science 1999;284:913.
closed. Am J Psychiatry 1963;12:336–45.
49. de Craen AJM, Kaptchuk TJ, Tijssen JGP, et al. Placebos and 77. Vogel AV, Goodwin JS, Goodwin JM. The therapeutics of pla- placebo effects in medicine: historical overview. J R Soc Med cebo. Am Fam Physician 1980;22:105–9.
78. Fish JM. Pills or placebos? Science 1999;284:913–5.
50. White L, Tursky B, Schwartz GE. Placebo: theory, research and 79. Medical Research Council. Streptomycin treatment of pulmonary mechanisms. New York: Guilford Press, 1985.
tuberculosis. BMJ 1948;ii:769–82.
51. Wickramasekera I. A conditioned response model of the placebo 80. Feinstein AR. An additional basic science for clinical medicine, effect: predictions from the model. In: White L, Tursky B, II: the limitations of randomized trials. Ann Intern Med 1983;99: Schwartz GE, eds. Placebo: theory, research, and mechanisms. New York: Guilford Press, 1985:255–87.
81. Black N. Why we need observational studies to evaluate effec- 52. Voudouris NJ, Peck CL, Coleman G. Conditioned placebo re- tiveness of health care. BMJ 1996;312:1215–8.
sponses. J Pers Soc Psychol 1985;48:47–53.
82. Cranberg L. Evaluating new treatments. BMJ 1998;317:1261.
53. Ader R. Conditioning effects in pharmacotherapy and the incom- 83. Chalmers I. Unbiased, relevant, and reliable assessment in health pleteness of the double blind, cross-over design. Integr Psychiatry care: important progress during the past century, but plenty of scope for doing better. BMJ 1998;317:1167–8.
54. Bootzin R. The role of expectancy in behavior change. In: White 84. Gluud C. Trials and errors in clinical research. Lancet 1999;354 L, Tursky B, Schwartz GE, eds. Placebo: theory, research, and mechanisms. New York: Guilford Press, 1985:196–210.
85. Andrews G. Randomised controlled trials in psychiatry: impor- 55. Swartzman LC, Burkel J. Expectations and the placebo effect in tant but poorly accepted. BMJ 1999;319:562–4.
clinical drug trials: why we should not turn a blind eye to un- 86. Keijnen J, Gotzsche P, Kunz RH, et al. So what’s so special about blinding, and other cautionary notes. Clin Pharmacol Ther 1998; randomisation? In: Maynard A, Chalmers I, eds. Non-random reflections on heath services research. London: BMJ Books, 56. Kirsch I. Specified nonspecifics: psychological mechanisms of placebo effects. In: Harrington A, ed. The placebo effect: an 87. Roberts C, Torgerson D. Randomisation methods in controlled interdisciplinary exploration. Cambridge, MA: Harvard Univer- 88. Lilienfeld AM. Ceteris paribus: the evolution of the clinical trails.
57. Brody H. The doctor as therapeutic agent: a placebo effect re- Bull Hist Med 1982;56:1–18.
search agenda. In: Harrington A, ed. The placebo effect: an in- 89. Leber P. The use of placebo control groups in the assessment of terdisciplinary exploration. Cambridge, MA: Harvard University psychiatric drugs: an historical context. Biol Psychiatry 2000;47: 58. Levine JD, Gordon NC, Fields HL. The mechanism of placebo 90. Khan A, Brown WA. The placebo enigma in antidepressant clini- analgesia. Lancet 1978;ii:654–7.
cal trials. J Clin Psychopharmacol 2001;21:123–5.
59. McEwen BS. Protective and damaging effects of stress mediators.
91. Weisman M. Pills or placebos? Science 1999;284:913–5.
N Engl J Med 1998;338:171–9.
92. Paterson C. Placebo effects: psychological, sociological and gen- 60. Ader R, Cohen N, Felten D. Psychoneuroimmunology: interac- eral practice research may elucidate effects. BMJ 1995;311:1640.
tions between the nervous system and the immune system. Lancet 93. Leber PD. Hazards of inference: the active control investigation.
Epilepsia 1989;30(suppl 1):S57–63.
94. Lasagna L. The Helsinki Declaration: timeless guide or irrelevant 127. Benatar SR, Singer PA. A new look at international research anachronism? J Clin Psychopharmacol 1995;15:96–8.
ethics. BMJ 2000;321:824–6.
95. Stanley B. An integration of ethical and clinical considerations in 128. Freedman B. Placebo-controlled trials and the logic of clinical the use of placebos. Psychopharmacol Bull 1988;24:18–20.
purpose. IRB: Rev Hum Subj 1990;12:1–6.
96. Chalmers TC. The clinical trial. Milbank Mem Fund Q Health Soc 129. Edwards SJL, Lieford RJ, Hewison J. The ethics of randomised controlled trials from the perspective of patients, the public, and 97. Temple R, Ellenberg SS. Placebo-controlled trials and active- the healthcare professionals. BMJ 1998;317:1209–12.
control trials in the evaluation of new treatments. Part 1: ethical 130. Appelbaum PS. Placebo controlled studies in schizophrenia: ethi- and scientific issues. Ann Intern Med 2000;133:455–63.
cal and scientific issues [Panel discussion]. Schizophr Res 1999; 98. Jones B, Jarvis P, Lewis JA, et al. Trials to assess equivalence: the importance of rigorous methods. BMJ 1996;313:36–9.
131. Madsen SM, Holm S, Davidsen B, et al. Ethical aspects of clinical 99. Makuch RW, Pledger G, Hall DB, et al. Active control equivalent trials: the attitude of participants in two non-cancer trials. J Intern studies. In: Peace K, ed. Statistical issues during research and development. New York: Marcel Dekker, 1990:225–62.
132. Ferriman A. WMA agrees to refine changes to Declaration of 100. Klein DF. Response to Rothman and Michels on placebo- controlled clinical trials. Psychiatry Ann 1995;25:401–3.
133. International Conference for Harmonization of Technical Re- 101. Tramer MR, Reynolds DJM, Moore RA, et al. When placebo quirements for the Registration of Pharmaceuticals for Human controlled trials are essential and equivalence trials are inad- Use (ICH), “Topic E 10: Choice of control group in clinical equate. BMJ 1998;317:875–80.
trials.” The European Agency for the Evaluation of Medical Prod- 102. Quitkin FM. Placebos, drug effects, and study design: a clini- ucts, London, July 2000. (www.ifpma.org/ich7.html) cian’s guide. Am J Psychiatry 1999;156:829–36.
134. Al-Khatib SM, Califf RM, Hasselblad V, et al. Medicine: 103. Lavori PW, Sugarman J. Placebo revised. Acta Psychiatr Scand placebo-controls in short-term clinical trials of hypertension. Sci- 104. Mattocks KM, Horwitz RI. Placebos, active controls, and the 135. Khan A, Warner HA, Brown WA. Symptom reduction and sui- unpredictability paradox. Biol Psychiatry 2000;47:693–8.
cide risk in patients treated with placebo in antidepressant clinical 105. Stein CM, Pincus T. Placebo-controlled studies in rheumatoid trials: an analysis of the Food and Drug Administration database.
arthritis: ethical issues. Lancet 1999;353:400–3.
Arch Gen Psychiatry 2000;57:311–7.
106. Temple R. Government viewpoint of clinical trials. Drug Info J 136. Perkins DO, Wyatt RJ, Bartko JJ. Penny-wise and pound-foolish: the impact of measurement error on sample size requirements in 107. Freedman B. Equipoise and the ethics of clinical research. N Engl clinical trials. Biol Psychiatry 2000;47:762–6.
137. Ranga K, Krishnan R. Efficient trial designs to reduce placebo requirements. Biol Psychiatry 2000;47:724–6.
108. Taubes G. Use of placebo controls in clinical trials disputed.
138. Yao Q, Wei LJ. Play the winner for phase II/III clinical trials. Stat Med 1996;15:24–13 (discussion 2455–8).
109. Reynolds T. The ethics of placebo-controlled trials. Ann Intern 139. Benson K, Hartz AJ. A comparison of observational studies and randomized, controlled trials. N Engl J Med 2000;342:1878–86.
110. Rothman KJ, Michels KB. The continuing unethical use of pla- 140. Concato J, Shah N, Horwitz RI. Randomized, controlled trials, cebo controls. N Engl J Med 1994;331:394–8.
observational studies, and the hierarchy of research designs.
111. Henry D, Hill S. Comparing treatments. BMJ 1995;310:1279.
N Engl J Med 2000;342:1887–92.
112. Freedman B, Weijer C, Glass KC. Placebo orthodoxy in clinical 141. Georgiou A. Placebo mania: high quality placebos should be research, I: empirical and methodological myths. J Law Med 142. Engel GL. The need for a new medical model: a challenge for 113. Freedman B, Weijer C, Glass KC. Placebo orthodoxy in clinical biomedicine. Science 1977;196:129–36.
research, II: ethical, legal, and regulatory myths. J Law Med 143. Ernst E. The role of complementary and alternative medicine.
114. Rothman K. Placebo mania. BMJ 1996;313:3–4.
144. Rees L, Weil A. Integrated medicine: imbues orthodox medicine 115. Macklin R. The ethical problems with sham surgery in clinical with the value of complementary medicine. BMJ 2001;322: research. N Engl J Med 1999;341:992–6.
116. Weijer C. Placebo-controlled trials in schizophrenia: are they 145. Fulder SJ, Munro RE. Complementary medicine in the United ethical? Are they necessary? Schizophr Res 1999;35:211–8.
Kingdom: patients, practitioners, and consultations. Lancet 1985; 117. World Medical Association. Declaration of Helsinki: recommen- dations guiding physicians in biomedical research involving hu- 146. Hrobjartsson A. The uncontrollable placebo effect. Eur J Clin man subjects. JAMA 1997;277:925–6.
118. World Medical Association. Declaration of Helsinki: ethical prin- 147. Jones R, Higgs R, de Aggelis C, Prideaux D. Changing face of ciples for medical research involving human subjects. JAMA medical curricula. Lancet 2001;357:699–703.
148. Ernst E, Herxheimer A. The power of placebo: let’s use it to help 119. Christie B. Doctors revise Declaration of Helsinki. BMJ 2000; as much as possible. BMJ 1996;313:1569–70.
149. Emanuel EJ, Dubler NN. Preserving the physician-patient rela- 120. Lancet. A fifth amendment for the Declaration of Helsinki [Edi- tionship in the era of managed care. JAMA 1995;273:323–9.
torial]. Lancet 2000;356:1123.
150. Lazaro J. Doctor’s status: changes in the past millennium. Lancet 121. Enserink M. Helsinki’s new clinical rules: fewer placebos, more disclosure. Science 2000;290:418–9.
151. Jackson SW. The listening healer in the history of psychological 122. Ready T. Placebo trials deemed unethical. Nat Med 2000;6:1198.
healing. Am J Psychiatry 1992;149:1623–32.
123. Clark PI, Leaverton PE. Scientific and ethical issues in the use of 152. Beckan HB, Frankel RM. The effect of physician behavior on the placebo controls in clinical trials. Annu Rev Health 1994;15: collection of data. Ann Intern Med 1984;101:692–6.
153. Dunning AJ. Status of the doctor: present and future. Lancet 124. Ellenberg SS, Temple R. Placebo-controlled trials and active- control trials in the evaluation of new treatments, Part 2: practical 154. Greenhalgh T, Hurwitz B. Why study narrative? BMJ 1999;318: issues and specific cases. Ann Intern Med 2000:133:464–70.
125. Levine RJ. The need to revise the Declaration of Helsinki. N Engl 155. Elwyn G, Gwyn R. Stories we hear and stories we tell: analysing talk in clinical practice. BMJ 1999;318:186–8.
126. Vastag B. Helsinki discord? A controversial Declaration. JAMA 156. Hurwitz B. Narrative and the practice of medicine. Lancet 2000; 157. Charon R. Narrative medicine: form, function, and ethics. Ann 171. Fitzpatrick R. Telling patients there is nothing wrong. BMJ 1996; 158. Barry C, Bradley CP, Britten N, et al. Patients’ unvoiced agendas 172. Gotzsche PC, Olsen O. Is screening for breast cancer with mam- in general practice consultations: qualitative study. BMJ 2000; mography justifiable? Lancet 2000;355:129–34.
173. Wald NJ, Law M, Watt H, et al. Apolipoproteins and ischaemic 159. Kleinman A, Eisenberg L, Good B. Culture, illness, and care: heart disease: implications for screening. Lancet 1994;343:75–9.
clinical lessons from anthropologic and cross-cultural research.
174. Bradley CP. Decision making and prescribing patterns: a litera- Ann Intern Med 1978;88:251–8.
ture review. Am Pract 1991;8:276–87.
160. Pachter LM. Culture and clinical care: folk illness beliefs and 175. Thomas KB. General practice consultations: is there any point in behaviors and their implications for health care delivery. JAMA being positive? BMJ 1987;294:1200–2.
176. Hussain MZ. Effect of shape of medication in treatment of anxi- 161. Carillo JE, Green AR, Betancourt JR. Cross-cultural primary ety states. Br J Psychiatry 1972;120:507–9.
care: a patient-based approach. Ann Intern Med 1999;130:829–34.
177. Kaptchuk TJ, Goldman P, Stone DA, et al. Do medical devices 162. Thompton H. Today’s patient: passive or involved? Lancet 2000; have enhanced placebo effects? J Clin Epidemiol 2000;53: 163. Britten N, Stevenson FA, Barry CA, et al. Misunderstanding in 178. de Craen AJM, Roos PJ, de Vries AL, et al. Effects of color of general practice prescribing decisions: qualitative study. BMJ drugs: systematic review of perceived effect of drugs and their effectiveness. BMJ 1996;313:1624–6.
164. McKinstry B. Do patients wish to be involved in decision making 179. Gunderson L. The complex process of naming drugs. Ann Intern in the consultation? A cross sectional survey with video vignettes.
180. Holm S, Evans M. Product names, proper claims? More ethical 165. Parkhouse J. Placebo reactor. Nature 1963;199:308.
issues in the marketing of drugs. BMJ 1996;313:1627–9.
166. Mortel CG, Taylor WF, Roth A. Who responds to sugar pills? Mayo Clin Proc 1976;51:96–100.
181. Barksy AJ. Nonpharmacologic aspects of medication. Arch Intern 167. Doongaji DR, Vahia VN, Bharucha MP. On placebos, placebo response and placebo responders. J Postgrad Med 1978;24: 182. Horwitz RI, Horwitz SM. Adherence to treatment and health outcomes. Arch Intern Med 1993;153:1863–8.
168. Andersen LO. Placebos in medicine. Lancet 1994;344:1642.
183. Haynes RB, McKibbon KA, Kanani R. Systematic review of 169. Goodwin JS, Goodwin JM, Vogel AV. Knowledge and use of randomized trials of interventions to assist patients to follow pre- placebos by house officers and nurses. Ann Intern Med 1979;91: scriptions for medications. Lancet 1996;348:383–6.
184. Steiner JF, Earnest MA. Lingua medica: the language of medi- 170. McDonald IG, Daly J, Jelinek VM, et al. Opening Pandora’s box: cation taking. Ann Intern Med 2000;139:926–30.
the unpredictability of reassurance by a normal test result. BMJ 185. Buckman R, Levith G. What does homeopathy do—and how?

Source: http://www.cognitivepsychotherapies.gr/files/papers/P_2001_01.pdf