US Pharm. 2011(36)12:Epub.  

Physicians and medical ethicists throughout history have recognized that a positive emotional state regarding a treatment is associated with a better response to that treatment. This phenomenon, known as the placebo effect, shaped early physician-patient relationships. Early treatises on medical ethics advised physicians about the dangers of providing patients with negative information. Over the last century, the rise of the informed consent doctrine has increased the power that patients have over their own medical treatment. Today, the informed consent doctrine mandates the disclosure of the nature, risks, and alternatives of a medical treatment. The placebo effect suggests that the disclosure of these facts affects the efficacy of the treatment itself. As new research confirms the existence of the placebo effect, and even explains the pathways by which the placebo effect works, medical professionals should be aware that placebos are a significant therapeutic tool. However, health practitioners must also strike a delicate balance to make sure they do not cross the legal boundaries of placebo use.

Placebo Versus Placebo Effect

A placebo is a particular treatment prescribed by a physician that has no known pharmacologic effect for its prescribed purpose. The term placebo is of Latin origin and literally translates as “I shall please.”1 There are two varieties of placebo. A “pure” placebo is an inert substance without known pharmacologic effect. Examples of pure placebos are sugar pills and saline injections. “Impure” placebos are prescribed medications that have a known pharmacologic effect on some illnesses but not on others. Examples of impure placebos include antibiotics and vitamins. The placebo effect, on the other hand, refers to the patient’s response to a treatment that is attributable to some reason other than the treatment’s pharmacologic effect. To illustrate this point, a drug that has a pharmacologic effect for its prescribed purpose is not a placebo. However, such a drug can have a placebo effect in addition to its pharmacologic effect. Procedures such as sham surgery, acupuncture, hypnosis, and other interventions that make patients feel better are generally believed to work through the placebo effect as well.

Early Acceptance of the Placebo Effect

The placebo effect has been observed and used in treatment since antiquity. Ancient physicians recognized the importance of a patient’s confidence in the physician and his treatment. The Roman physician Galen stated: “He cures most successfully in whom the people have the most confidence.”2 Nearly 500 years earlier, the Greek physician Hippocrates, who is best known for his Hippocratic Oath, warned that “many patients…have taken a turn for the worse…by the declaration…of what is present.”3 In the 19th century, French physician Armand Trousseau recognized that patients were more optimistic about new drugs than about older drugs. He advised his fellow physicians to “treat as many patients as possible with the new drugs while they still have the power to heal.”4

In 1803, Thomas Percival, a respected staff physician at the Manchester Infirmary in England, published the first modern code of medical ethics. Percival recognized the power of the placebo effect, and in his code he declared that the physician should provide patients with hope and comfort and “counteract the depressing influence of maladies.” Critics who read Percival’s work urged him to limit the obligation to convey hope only “as far as truth and sincerity will admit,” but Percival believed that if a physician was motivated by a desire to do good rather than to deceive, a falsehood was not a lie but was praiseworthy.5 The American Medical Association’s (AMA) first Code of Ethics in 1847 was heavily influenced by Percival’s Medical Ethics. In fact, the code stated: “The life of a sick person can be shortened not only by the acts, but also by the words or the manner of a physician.  It is therefore, a sacred duty to guard himself carefully in this respect, and to avoid all things which have a tendency to discourage the patient and depress his spirits.”6

By 1957, the AMA had amended the code three times, each time condensing and removing controversial provisions regarding the physician-patient relationship. By 1980, the code had been condensed to 250 words. Principle II of the code directed physicians to “deal honestly with patients.”7 The AMA’s shift from an appreciation of the placebo effect to a focus on patient autonomy reflected the establishment of the informed consent doctrine. The legal doctrine of informed consent is based on the idea that every person has a right to determine what is done to his own body. The precise rules governing how much information must be given vary by the laws of each state, but there is general agreement that before providing medical treatment, a physician must explain to the patient the nature and purpose of a proposed treatment, the probable risks and benefits, the probability of success, the availability of reasonable alternative courses of action, and, in many cases, the risk of refusing treatment. Despite an appreciation for the placebo effect dating back hundreds of years, the informed consent doctrine has questioned the legality of placebo use.

Modern Therapeutic Placebo Use

Trousseau’s observation that new drugs are more effective than older drugs is based on the fact that although the pharmacologic effect of a drug remains constant, the placebo effect wanes over time. In fact, it has been argued that all OTC drugs act mainly through the placebo effect, despite the fact that they contain ingredients that have known pharmacologic effects.8 Many believe that some of the most widely used drugs in modern medicine are completely inert, and estimates suggest that 30% to 45% of all prescription drugs are nothing more than placebos.9

The actual amount of therapeutic placebo use in the United States is somewhat unknown, but may be more common than generally believed. A 1979 survey examined the use of pure placebos among 60 physicians and 39 nurses during their postgraduate training. The surveyors found that 78% of the physicians and 82% of the nurses had ordered or administered at least one placebo medication for pain relief.10 In several cases, the placebos were used to test whether patients were faking their symptoms. In a 1999 survey of 74 interns at a New York community hospital, 64% of the interns were aware of pure placebo use in clinical practice, and 16% had ordered placebo administration during their internships.11 In roughly half of the situations in which interns had ordered placebos, attending physicians were aware of it and none had objected. In 2007, 466 physicians from internal medicine departments of three Chicago medical schools were surveyed. The results revealed that 231 of the physicians responded, and 45% of these physicians reported prescribing or administering a placebo. The survey also found that 96% of the respondents believed placebos could have therapeutic effects, and only 12% believed that placebo use should be categorically prohibited in medical care.12

Placebo Efficacy

Placebos have proven effective in many areas of medicine, but have been most noticeable in the treatment of pain. A study of placebo administration at a Canadian teaching hospital found that 89% of pure placebos administered were given to treat pain.13 In 1985, Frederick Evans reviewed six double-blind studies comparing morphine to a placebo and concluded that placebos are 56% as effective as morphine in reducing pain.14 Around the same time, researchers at the University of California, San Francisco, attempted to determine an appropriate control group in studies of placebo-induced analgesia. Among other things, they concluded that a placebo given for pain may be as effective as 8 mg of morphine.15

Whereas most studies on placebo effectiveness against pain look at self-reports of patients to measure pain, a probative 2001 study also looked at patient behavior. Known by economists as revealed preference, this theory considers the actions of subjects rather than only their reports. The study enrolled 38 patients recovering from thoracic surgery for lung cancer in the surgery ward of an Italian hospital. Patients were given an unknown solution that was actually saline via intravenous drip. They were permitted to request buprenorphine, an analgesic. Patients were divided into three groups: the first group was told nothing about the unknown solution, the second group was told that the saline was either a powerful painkiller or a placebo, and the third group was told that the saline was a powerful painkiller.  The investigators measured both the amount of buprenorphine requested and the self-reports of pain intensity. The third group requested the least buprenorphine, followed by the second group and the first group, respectively. However, all three groups reported roughly the same level of pain. The patients’ request for pain relief was therefore based less on their own determination of pain and more on their faith in the painkiller they believed they were receiving. The researchers concluded that the placebo effect of the suggestion given to the third group was the equivalent of four additional doses of buprenorphine over 70 hours.16

Mechanisms of the Placebo Effect

The connection between the mind and body was well known at the times of Hippocrates and Galen.  Modern medicine, however, treats the mind and body as separate. In an editorial in the Journal of the American Medical Association, professor and psychiatrist David Spiegel wrote: “We have been closet Cartesians in modern medicine, treating the mind as though it were reactive to but otherwise disconnected from disease in the body.”17 According to Walter Brown, professor of psychiatry at Brown University, simply expecting an improved outcome can lead to an improved outcome. “There is certainly data that suggest that just being in the healing situation accomplishes something. Depressed patients who are merely put on a waiting list for treatment do not do as well as those given placebos.”18

One explanation for the placebo effect is that placebos produce endorphins. Endorphins, discovered in the 1970s, are broadly defined as any of a group of endogenous peptides found especially in the brain that bind chiefly to opiate receptors and produce some pharmacologic effects, such as pain relief, like those of opiates. Endorphins are chemically similar to opium-delivered narcotics and attach themselves to the same receptor sites in the brain as morphine. Research has demonstrated that naloxone, a drug that blocks opioid analgesia, also blocks placebo analgesia. In a 1978 study, all patients were given a placebo following a dental procedure. For the second treatment, patients received either another placebo or naloxone. Of the patients who received the naloxone, only those who responded to the first placebo experienced increased pain.19 The research demonstrates that placebos produce actual relief from pain in the form of endorphins. This is consistent with Professor Brown’s observations: “When placebos are given for pain management, the course of pain relief follows what you would get with an active drug. The peak relief comes about an hour after it’s administered, as it does with the real drug, and so on. If placebo analgesia was the equivalent of giving nothing, you’d expect a more random pattern.”20

Another possible explanation for the placebo effect is conditioned response. In a 1990 study, pain was artificially generated by administering local electric shock to each subject’s wrist. The researchers then reduced the stimulus, applied a sham analgesic cream, and told the subjects it was a mild analgesic. In a subsequent trial, the researchers applied the cream but returned the electrical stimulus to its original level. The subjects reported that the pain in the subsequent trial was less than the pain in the original stimulus.21 The application of the placebo cream with which the subjects had already had a positive experience led the subjects to an expectation of decreased pain.

Additional evidence that the placebo effect is explained by patient expectations is shown by the fact that the placebo effect varies with the degree of medical intervention. “The bigger and more dramatic the patient perceives the intervention to be, the bigger the placebo effect. Big pills have more than small pills, injections have more than pills and surgery has the most of all.”22 This is because patients believe that drugs are more effective at higher doses, and that major and risky medical interventions are not performed unless necessary.

Negative Effects of Placebos

Placebos may directly or indirectly harm patients. Surveys have noted adverse side effects from placebos, such as nausea, dermatitis, and headaches.23,24 An indirect form of harm is the danger that physicians will prescribe placebos without completing a thorough medical and psychological evaluation. The availability of a placebo may lead to diminished diagnostic vigilance by the physician and may result in a treatable medical disorder being overlooked. Another form of indirect harm is a sense of dependency with medication. Simmons has remarked that modern patients “have come to expect, some even to demand, a medication for every symptom.”25 If physicians come to rely on placebos when no known treatment exists, they are only encouraging the idea that drugs can cure most illnesses. Particularly for those patients who are seeking medication, unnecessary medication could lead to a dependence on or addiction to drugs. Finally, the use of impure placebos such as antibiotics could cause significant downstream harm to society through the creation of antibiotic resistant bacteria.

Legality of Placebo Use

Although the AMA’s Code of Medical Ethics does not itself carry the force of law, courts generally give it weight in resolving issues in the area of informed consent. In Culbertson v. Mernitz, the Indiana Supreme Court stated: “The 1992 Code of Medical Ethics, as prepared by the Council on Ethical and Judicial Affairs of the American Medical Association, sets forth the medical profession’s standard on informed consent.”26 In November 2006, the AMA adopted its current stance on placebo use after a proposal by the Council on Ethical and Judicial Affairs to prohibit placebo deception. The report, issued in June 2007, finds the use of deceptive placebos to be ethically improper. The AMA advocates that revealed placebos can still be effective and allows placebo use “for diagnosis or treatment only if the patient is informed of and agrees to its use.”27 However, it has been largely accepted that deceptive placebos produce greater benefits. In fact, revealed placebos may lead to side effects, skepticism about the medical system, and loss of trust in physicians, possibly rendering the placebos more harmful than no treatment at all.

Considering the significant use of placebos in the U.S. and the AMA’s stance on placebos, it is surprising to find an almost complete lack of case law on the issue. There are several theories for why this is so. First, litigation in the medical context arises from a harmful or poor result. When placebos are successful or have no effect, no litigation is expected. When placebos do cause direct harm, not including violations of patient autonomy, the harm is usually minor side effects not worthy of a law suit. Secondly, with regard to impure placebos, patients may have difficulty demonstrating that the physician actually prescribed a placebo. Since impure placebos are prescribed medications that have a known pharmacologic effect on some illnesses, physicians can often present some plausible theory as to why the medication could have a positive effect.  These cases are most simply litigated as malpractice cases where the physician fell below the standard of care and failed to adequately treat a condition.28 If the physician actually committed malpractice, determining placebo treatment becomes unnecessary.

A final reason is the difficulty of proving that the physician caused the patient’s harm by failing to disclose the use of a placebo. In a traditional informed consent case, a lawyer can show that a physician negligently failed to disclose important information to the patient. Cases involving placebos are difficult for lawyers because the entire basis for the placebo effect is a lack of disclosure to the patient.

Duty to Disclose?

Two schools of thought have arisen regarding the scope of a physician’s duty to disclose. The majority of jurisdictions employ the “reasonable physician” standard. Under this standard, a physician is required to disclose the amount and type of information that a reasonable physician in the same or similar circumstances would have disclosed. Given the little knowledge medical experts have on the issue, lawyers tend to avoid these arguments in the context of placebos. A minority of jurisdictions accept the “reasonable patient” standard. This standard requires the physician to disclose the material information that a reasonable patient would need to make a decision. Courts differ on whether “reasonable patient” refers objectively to a reasonable patient or subjectively to the patient at issue. In either case, the causation analysis is poorly suited to the placebo context because it requires a determination that a patient would have consented to a therapy that requires nondisclosure. As one analyst put it: “[D]oing so is logically impossible. One simply cannot disclose to a patient a treatment that by definition requires non-disclosure.”29 Thus, the informed consent doctrine and the placebo effect inherently conflict.

Medical professionals using therapeutic placebos should be aware of the AMA’s stance that deceptive placebos are ethically improper. Moreover, the informed consent doctrine mandates the disclosure of the nature, risks, and alternatives of a medical treatment. Nonetheless, deceptive placebos are commonly used in the medical arena, and research continues to confirm the existence of the placebo effect and explain the pathways by which it works. Since the informed consent doctrine is based on negligence principles, medical professionals should not prescribe placebos when a known treatment exists. Failure to prescribe the known treatment could result in the physician falling below the standard of care. Instead, therapeutic placebos are best limited to circumstances where the expected benefits from a placebo outweigh the expected benefits of an established treatment, or where no known treatment exists at all.

Professor Brown proposes a model wherein physicians tell the patient that the pills do not contain any drug, but provide hope by stating: “We don’t know exactly how they work; they may trigger or stimulate the body’s own healing processes. We do know that your chances of improving with this treatment are quite good.”20 Such an approach would take advantage of the placebo effect while disclosing the fact that the pills do not contain any actual drug. Finally, physicians should be aware of the placebo effect in the course of all of their treatments. The optimism and confidence physicians portray toward a treatment can affect the patient’s outcome, as recognized by the ancient physicians Hippocrates and Galen.


1979;91:106-110. 1. Knapp M. Placebo therapy and the law: prescribe with care. Am J L and Med. 1982,8:371-373.
2. Harington A. The Placebo Effect: An Interdisciplinary Exploration. 1st ed. 1997.
3. Hippocrates. Decorum. Jones W, trans. Cambridge, MA: Harvard University Press; 1967:297.
4. Laprata J. Healing: The Coming Revolution in Holistic Medicine. New York, NY: McGraw Hill; 1978:89.
5. Percival T. Medical Ethics: A Code of Institutes and Precepts Adapted to the Professional Conduct of Physicians and Surgeons. Special ed, Pellegrino E, ed. 1985,1803:1.
6. Am. Med. Assn, Code of Ethics ch. 1847. I, art. I, para. 4.
7. Katz J. The Silent World of Doctor and Patient. New York: NY: The Free Press; 1984:23.
8. Jones J. Do over-the-counter drugs act mainly as placebos? Yes. In: Lasagna L. Controversies in Therapeutics. 1980, 26.
9. Simmons B. Problems in deceptive medical procedures: an ethical and legal analysis of the administration of placebos. J Med Ethics, 1978,4:172.
10. Goodwin JS, et al. Knowledge and use of placebos by house officers and nurses. Ann Int. Med.
11. Berger JT. Placebo medication use in patient care: a survey of medical interns. West J Med. 1999;170:93-94.
12. Sherman R, Hicker J. Academic physicians use placebos in clinical practice and believe in mind-body connection. J Gen Int Med. 2008;23:7.
13. Gray G, Flynn P. A survey of placebo use in a general hospital. Gen Hosp Psych, 1981;3:199-201.
14. Evans FJ. Expectancy, Therapeutic instructions, and the placebo response. In: White L, et al. Placebo: Theory, Research and Mechanisms. 1985:215-219.
15. Levine JD, Newton CG. Influence of the method of drug administration on analgesic response. Nature. 1984;312:755.
16. Pollo A, et al. Response expectancies in placebo analgesia and their clinical relevance. Pain. 2001;93:77-80.
17. Speigel D. Healing words: emotional expression and disease outcome. JAMA. 1999;281:1328-1329.
18. Talbot M. The placebo prescription. New York Times. January 9, 2000:36.
19. Levin JD, et al. The mechanism of placebo analgesia. Lancet. 1978;23:654.
20. Talbot M. The placebo prescription. New York Times. Jan. 9, 2000:38.
21. Voudories NJ, et al. The role of conditioning and verbal expectancy in the placebo response. Pain. 1990;43:126-127.
22. Talbot M, The placebo prescription. NY Times. Jan. 9, 2000:35.
23. Adler HM, Hammett VO. The doctor-patient relationship revisited: an analysis of the placebo effect. Ann Int Med. 1973;78:595-598.
24. Bourne H. The placebo—a poorly understood and neglected therapeutic agent. Rational Drug Therapy. 1971;5:1-6.
25. Simmons B. Problems in deceptive medical procedures: an ethical and legal analysis of the administration of placebos. J Med Ethics. 1978,4:172.
26. Culbertson v. Mernitz , 602 N.E.2d 98, 103 (Ind. 1992).
27. Council on Ethical & Judicial Affairs, Rep. 2-A-06, Opinion 8.083. Am. Med. Assn: Withholding Information from Patients (Therapeutic Privilege), 2006.
28. Kolber AJ. A Limited Defense of Clinical Placebo Deception. 26 Yale L. & Pol’y Rev. 2007;75:79.
29. Kolber AJ. A Limited Defense of Clinical Placebo Deception. 26 Yale L. & Pol’y Rev. 2007;75:109.

To  comment on this article, contact