Answer by Tirumalai Kamala:
Placebos are used in both clinical trials and general practice. Arguably the most prescribed drug in human history (1), the history of medicine is in large part the history of the placebo (2). What’s the ethical dilemma of placebos in clinical trials? To understand this we need to examine the history of informed consent, and understand what’s a placebo, where it stands in current biomedical research and how its use creates an ethical dilemma.
The problem starts with the definition of placebo. There isn’t one. Rather there are many.
The word placebo derives from the Latin word, ‘placere’ meaning ‘I will please’ or ‘I will do good’, a word opposite to nocebo derived from the Latin word ‘nocere’ meaning ‘I will harm’. One definition of placebo is ‘An inert substance usually prepared to look as similar to the active product investigated in a study as possible‘ and the placebo effect is ‘any (usually beneficial) changes that occur within a group ‘treated’ with placebo‘ (3). I wrote one definition implying there are many and therein lies one of several ethical issues about placebos. A recent review (4) helpfully collated different definitions of placebo used by many world-renowned biomedical research organizations and health policy-making bodies.
Ironically, even different institutes within the same organization, the US-based NIH, have multiple definitions. Why are multiple definitions a problem? Unlike our past or rather because of it, in biomedical research today, safety of patients is sacrosanct. The more ambiguous the definition of a foundational term like placebo, the greater the possibility of misunderstanding and miscommunication. If the experts themselves don’t agree on a common definition of placebo, an intervention that requires the approval of clinical trial participants, what chances then that the latter, likely not experts, fully understand what their consent means? Such a pitfall could further undermine public trust in doctors and clinical trials (5).
The clinical trial as we now know it is of recent vintage. It starts with the UK’s Medical Research Council 1947 study of streptomycin treatment for tuberculosis (6). Since 1947, clinical trial methodology has improved vastly to minimize bias in design, management and interpretation. Now health-care providers need to provide clinical trial participants information necessary for their autonomous decision-making, i.e. informed consent forms are now mandatory. In this mix, we have placebos.
By enabling a causal association between treatment and outcome, placebos have been instrumental in making clinical research more rigorous and scientifically sound, with a caveat of course. After all, the ‘placebo effect’ tells us in no uncertain terms that we don’t fully understand how our body, brain and mind work together. The unpredictable ‘placebo effect’ also negates economic arguments that are typically made in favor of placebo-controlled trials, namely, that it requires fewer participants and is faster compared to trials that compare two or more treatments (e.g., active controlled or superiority or non-inferiority trials) (7).
However, why do clinical trials require informed consent (approval by participants) in the first place?
Brief history of informed consent in clinical trials
German Nazi doctor experiments on concentration camp inmates and the Tuskegee Syphilis Project are representative examples of egregious medical practice abuse. Human autonomy, the overarching victim, thus became the central tenet in modern medical ethics through a convoluted and not simplistically linear process. While the established a set of research ethics for human experimentation in 1947, isn’t it ironic that the was conducted continuously from 1932 until 1972? In this study, African-Americans diagnosed with syphilis were deliberately left untreated in order to follow natural disease progression. This example of egregious post-WWII violation of human rights proved that the Nuremberg Code alone was insufficient to prevent wholesale abuse in human experimentation. While public revelation of the Tuskegee project in 1972 drove the creation of new standards in research ethics requiring the treatment of research participants as autonomous agents, it was only in 1997 that US president Bill Clinton apologized to the Tuskegee study participants and their families. However, neither of these two egregious examples of abuse alone suffice to understand how we arrive at today’s ‘informed consent’. For this we need to delve into the law and lawsuits.
The first piece of ‘Informed consent’. When and how the phrase came into existence.
The US case Salgo v. Leland Stanford, Jr. University Board of Trustees coined the phrase, ‘informed consent’. The plaintiff, Mr Salgo underwent a now-defunct treatment ‘involving puncturing the aorta through the back in order to inject a radio-opaque dye, and was left with permanent paralysis of the legs. According to the direction given to the jury: “The physician has . . . discretion [to withhold alarming information from the patient] consistent, of course, with the full disclosure of facts necessary to an informed consent” ( 317 P.2d 170 (Cal. Ct. App.) at 181)’(8). In this jury instruction, there is an obvious contradiction of terms between ‘discretion to withhold alarming information‘ and ‘full disclosure of facts necessary to an informed consent‘. Nevertheless, this and subsequent cases codified that disclosure needed to conform to ‘professional practice standard’, namely, what a reasonable health-care practitioner would do under similar circumstances (8).
The second piece of ‘Informed consent’: The ‘reasonable person standard’.
In the US case, Canterbury v. Spence ( 464 F.2d 772 (D.C. Cir.)), the patient fell out of his hospital bed after undergoing a laminectomy (surgical removal of a vertebral bone called the lamina) and suffered major paralysis (8). The patient had not been warned about ‘the possibility of this rare outcome‘. This case deemed the professional practice standard inadequate in that it failed to respect the patient’s self-determination, and it gave way to the patient- centered ‘reasonable person standard‘, i.e. what any reasonable patient would consider necessary and sufficient to know, rather than what professionals might consider necessary to disclose (9). According to the philosopher and his colleague Helga Kuhse, ‘this single move served to overcome three main weaknesses of the professional practice standard: first, that agreed professional standards of disclosure were typically set too low to satisfy patient demand for information; second, that there were no agreed standards for new procedures; and, third, that patients were put at a significant disadvantage in having to rely upon expert witnesses (usually other health-care practitioners) in disputes about standards of care’ (8).
With the adoption of a ‘reasonable person standard’, patients have greater decision-making control about their own health care. In theory, anyway. What does this mean in practical terms though? Health-care providers have to disclose to the patient the four elements necessary for informed patient consent: the nature (therapeutic/not), risks, alternatives and benefits of the procedure and/or treatment. Disclose the nature of the treatment: Is it therapeutic or not? This is how we arrive at the current ethical dilemma of the placebo. As with many ethical dilemmas that ensue from the tussle between individual rights and collective good, the participants of a placebo-controlled trial bear the risks. Let’s examine who benefits.
Who benefits from a placebo-controlled trial, public health (the participants, i.e. patients, nation(s)) or the sponsor? Some case studies highlight a) inevitable conflicts of interest and b) developed vs developing world disconnects in access to consistent standard of care.
Neither individuals nor public health, only the sponsor. In Bolivia, Discovery labs was testing a new surfactant, Surfaxin, for respiratory distress syndrome (RDS). However, its intent was not to develop the drug to benefit premature Bolivian infants with RDS but to market it in high-income countries (10). Violation of ethics? No doubt.
Individuals pay but public health benefits. In the Romanian government-funded, Bucharest Early Intervention Project (11), abandoned children were randomly assigned to either foster or institutional care even though US childcare experts agreed foster care was better. The result helped Romanian policy-makers establish a foster care program in Bucharest, and an ethicist concluded the study had sufficient ‘social value‘ to justify its design (12). Millum and Grady state ‘Social value is a fundamental, but under-analyzed, concept in research ethics‘ (7). I bet the children left to institutional care would vociferously beg to differ. Another fundamental informed consent issue that this case highlights? Presumably, usually parents or caregivers, i.e. individuals, proffer informed consent on behalf of minors. In the case of these and other orphans, presumably the state, manifestly not an individual, consents? Is individual and state informed consent the same thing? The study (11) itself is mum on this issue. Yet another ethical conundrum.
Double standards in medical research.The US 076 regimen gave Zidovudine (AZT) intravenously prenatally, during delivery, and postpartum (13). Though it reduced perinatal (vertical) HIV transmission by approximately 2/3rds and had become the standard of care in developed countries, the needed infrastructure, comprehensive prenatal care, and drug cost couldn’t be met in developing countries, which had the majority of perinatal HIV infections. Hence controversial placebo-controlled trials in Sub-Saharan Africa gave a single nevirapine dose to pregnant women during labor and their infants within 72 hours of birth (14). What’s egregious about that? As Millum et al write, ‘it was known at the time that single-dose nevirapine would not be as effective as more comprehensive and much more expensive treatment regimens that also targeted transmission during pregnancy‘ (15). Like it or not, when it comes to public health, we live in a world of haves and have-nots.
According to latest guidelines, when is a placebo currently acceptable?
A brief history of the modern clinical trials and the codified set of rules for their conduct, namely, the World Medical Association’s 1964 (DoH) and its subsequent revisions (7 as of 2013) reveal plenty of controversy such as the 5th revision in 2000 which was approved without consensus from national medical associations in the aftermath of the controversial sub-Saharan nevirapine studies on vertical HIV transmission.
The 7th version of the DoH, adopted in October 2013 at the 64th WMA General Assembly in Fortaleza, Brazil, says, ‘Medical research involving human subjects may only be conducted if the importance of the objective outweighs the risks and burdens to the research subjects. (Paragraph16)‘ (15). It also asserts that ‘placebos, no intervention or any intervention less effective than the best proven one may be used only when the patients who receive them will not be subject to additional risks of serious or irreversible harm as a result of not receiving the best proven intervention‘ (16). ‘Less effective than the best proven‘, a phrase that explicitly codified double standards in medical research. In response, the Latin American and Caribbean Medical Confederations did not approve the wording of placebo use in the DoH 2013 because ‘the poor and vulnerable populations, discriminated by their lack of resources, cannot be subjected to biomedical research that have levels of safety less than those applied to more developed societies’ (17, 18).
Ambiguity about the ‘risk of serious or irreversible harm‘ is another big problem with the latest DoH stance on placebo. Is it risk of serious or irreversible harm to not treat a cut or skin biopsy or to not treat an HIV positive pregnant woman? In the arena of global politics, ‘risk of serious or irreversible harm‘ becomes a vast, seemingly insurmountable chasm between the optimal and the dubious in wealthy and poor countries, respectively. Hellman et al also point out that the 2013 DoH excluded the division between therapeutic and non-therapeutic studies, further increasing participant vulnerability (18).
Today, health-care providers need to tell placebo-controlled trial participants they could receive either a treatment or a placebo. Do participants balk? Certainly but it seems to vary by disease type and severity. While a study suggests 70% of cancer patients would likely decline to participate in a placebo-controlled trial (19), 24% of hypertension patients are likely to do so (20).
Final word: Placebos remain a persistent ethical dilemma. We just added another wrinkle with bureaucracy.
Our history teaches us that when we codify processes and develop contingencies to accommodate ethical codes and legal requirements, we tend to become complacent and end up going through the motions without fully comprehending the poignant ethical and moral compulsions that necessitated them. In other words, a definition of bureaucracy. We see a similar effect at play with our current use of placebos in clinical trials. The meager research currently available on how clinical trial participants are informed about placebos and their effects suggests that placebos and their risks are poorly explained (4, 21, 22). This, even without getting into vast global cultural differences. For example, imagine informed consent in cultures that believe evil spirits or witchcraft cause diseases or where the concept of lock and key privacy (…of informed consent forms) is alien or where native healers are accorded respect on par with modern medical practitioners or where illiteracy is prevalent. Given we do live in a world of such enormous cultural, resource and opportunity differences, from benign neglect to active deception, the entire ethical gamut with respect to placebos is still very much at play.
- Shapiro, Arthur K., and Elaine Shapiro. “The powerful placebo.” From ancient priest to modern (1997).
- Požgain, Ivan, Zrinka Požgain, and Dunja Degmečić. “Placebo and nocebo effect: a mini-review.” Psychiatria Danubina 26.2 (2014): 100-107.
- Day, Simon. Dictionary for clinical trials. John Wiley & Sons, 2007.
- Hernández, Astrid, et al. “The Definition of Placebo in the Informed Consent Forms of Clinical Trials.” PloS one 9.11 (2014): e113654.
- Louhiala, Pekka, Harri Hemilä, and Raimo Puustinen. “Clinical use of placebo treatments may undermine the trust of patients: a response to Gold and Lichtenberg.” Journal of medical ethics (2014): medethics-2014.
- Marshall, Geoffrey, et al. “Streptomycin treatment of pulmonary tuberculosis: a Medical Research Council investigation.” BMJ 2.4582 (1948): 769-782.
- Millum, Joseph, and Christine Grady. “The ethics of placebo-controlled trials: methodological justifications.” Contemporary clinical trials 36.2 (2013): 510-514.
- A Companion to Bioethics, Second Edition. Helga Kuhse, Peter Singer, editors.
- Faden, Ruth R., Tom L. Beauchamp, and Nancy M. King. “A history and theory of informed consent.” (1986).
- Lurie P, Wolfe SM. Commentary 9.2. The Developing World as the “Answer” to the Dreams of Pharmaceutical Companies: The Surfaxin Story. In: Lavery JV, Grady C, Wahl ER, Emanuel EJ, editors. Ethical Issues in International Biomedical Research: A Casebook. Oxford University Press; 2007. pp. 159–170.
- Nelson, Charles A., et al. “Cognitive recovery in socially deprived young children: The Bucharest Early Intervention Project.” Science 318.5858 (2007): 1937-1940.
- Rid, Annette. “When is research socially valuable? Lessons from the Bucharest Early Intervention Project: commentary on a case study in the ethics of mental health research.” The Journal of nervous and mental disease 200.3 (2012): 248-249.
- Connor, Edward M., et al. “Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment.” New England Journal of Medicine 331.18 (1994): 1173-1180.
- Marseille, Elliot, et al. “Cost effectiveness of single-dose nevirapine regimen for mothers and babies to decrease vertical HIV-1 transmission in sub-Saharan Africa.” The Lancet 354.9181 (1999): 803-809.
- Millum, Joseph, David Wendler, and Ezekiel J. Emanuel. “The 50th anniversary of the Declaration of Helsinki: progress but many remaining challenges.” Jama 310.20 (2013): 2143-2144.
- World Medical Association (WMA). Declaration of Helsinki. Amended by the 64th WMA General Assembly, Fortaleza, Brazil, October 2013. WMA Archives, Ferney-Voltaire, France.
- Confederacion Medica Latinoamericana y el Caribe (CONFEMEL). Declaracion de Pachuca Sobre la Revision de Helsinki. 22 and 23 November 2013. (Accessed May 5, 2014.
- Hellmann, Fernando, et al. “50th Anniversary of the Declaration of Helsinki: The Double Standard Was Introduced.” Archives of medical research 45.7 (2014): 600-601.
- Jefford, Michael, and Rosemary Moore. “Improvement of informed consent and the quality of consent documents.” The lancet oncology 9.5 (2008): 485-493.
- Halpern, Scott D., et al. “Hypertensive patients’ willingness to participate in placebo-controlled trials: implications for recruitment efficiency.” American heart journal 146.6 (2003): 985-992.
- Bishop, Felicity L., et al. “Informed consent and placebo effects: a content analysis of information leaflets to identify what clinical trial participants are told about placebos.” PloS one 7.6 (2012): e39661.
- Keränen, Tapani, et al. “Placebo-controlled clinical trials: how trial documents justify the use of randomisation and placebo.” BMC medical ethics 16.1 (2015): 2.