The Promise on the Form: How the Crimes of Nuremberg Produced Informed Consent – and Why the Signature Still Means Less Than We Think

The Hook: The Rules Were Already There

On February 28, 1931, the German Reich Minister of the Interior issued a circular with the cumbersome title Guidelines for Novel Therapy and for the Conduct of Scientific Experiments on Human Beings. Measured against anything else that existed in the world at the time, it was a remarkably modern document. It distinguished therapeutic treatment from pure experiment. It required the consent of the research subject — consent given in "unambiguous" form after the person had been instructed "in an appropriate manner." It prohibited experiments on the dying. It demanded particular restraint with children and adolescents. And it expressly forbade exploiting a person's social hardship to induce participation.

These guidelines were never formally repealed. They remained in force, at least on paper, throughout the period in which human beings were killed in low-pressure chambers at Dachau to study pilot survival at high altitude; in which women at Ravensbrück had wounds deliberately inflicted and infected with wood splinters and glass to test sulfonamides; in which people at Buchenwald were infected with typhus to evaluate vaccines.

That is the most uncomfortable finding in all of research ethics, and it stands at the beginning of this article because it frames everything that follows: a body of rules on paper that nobody enforces protects nobody. The history of informed consent is therefore not the history of ever better wording, but the history of a laborious transition from good intentions to institutional mechanisms: from a guideline to a code, from a code to a declaration, from a declaration to a statute, from a statute to a committee that is allowed to say no before the study begins.

And even at the end of that road, in 2026, stands a form that most people do not understand, that they confuse with treatment, and that they sign anyway. This article tells both halves of the story: the triumph and the unsolved problem.

Part 1: Before Nuremberg – Why Germany Had the Rules and They Did No Good

The Neisser Case and the Prussian Directive of 1900

The prehistory begins with a scandal. Albert Neisser, discoverer of the gonococcus and one of the best-known dermatologists of his day, had tried in Breslau to develop a syphilis vaccine. To that end he injected cell-free serum from syphilis patients into other patients, among them prostitutes and children, without asking them. Several fell ill. Neisser concluded that they would have been infected through their occupation anyway; the public drew a different conclusion.

The consequence, in 1900, was an instruction from the Prussian Minister for Religious, Educational and Medical Affairs to all clinic directors: interventions for purposes other than diagnosis, therapy, or immunization are prohibited if the person concerned is a minor or not fully competent, if she has not been "properly instructed," and if she has not expressly agreed. Authorization also had to be granted personally by the clinic director and documented.

That is, in substance, informed consent. It was formulated in 1900, forty-seven years before Nuremberg.

Why That Was Not Enough

The Prussian directive and the Reich Circular of 1931 shared the same weakness: they were administrative rules without sanction, without an independent reviewing body, and without any role for an outsider. Oversight rested with the clinic director — that is, with precisely the person whose scientific interest was supposed to be scrutinized. The medical ethicist Hans-Martin Sass analyzed the Reich Circular at length in the Journal of Medicine and Philosophy in 1983 and classified it as a remarkably progressive but ineffective document.

The lesson is as sober as it is general: an ethics that makes the actor the judge of his own case works exactly as long as it is not needed. The moment interests diverge, it is worthless. Anyone who designs compliance processes today, organizes code reviews, or builds four-eyes principles into an approval chain knows the same logic: self-control is not control.

Part 2: Nuremberg 1947 – The Code Nobody Applied to Themselves

The Trial

On December 9, 1946, proceedings opened before an American military tribunal in Nuremberg in the case of United States of America v. Karl Brandt et al., the so-called Doctors' Trial. Twenty-three defendants, mostly physicians, stood accused, among them Hitler's escort physician Karl Brandt. The verdict came on August 19, 1947: seven death sentences, several life sentences, seven acquittals.

The defense had an argument that gave the judges pause. It claimed that the camp experiments differed in no fundamental way from what researchers were doing everywhere in the world: American prison inmates too had participated in malaria studies; elsewhere too, research was conducted on the defenseless. There simply was, the defense argued, no recognized standard against which the defendants could be measured.

So the judges formulated that standard themselves, before delivering the verdict. Ten points, later called the Nuremberg Code. The first reads with a clarity that still resonates: The voluntary consent of the human subject is absolutely essential.

The Ten Points in Brief

Note the distribution: points 1 and 9 concern autonomy; points 2 through 8 and 10 concern risk, qualification, and scientific quality. The Code is therefore not a pure autonomy document. It says: consent is necessary, but not sufficient. A poorly designed experiment remains impermissible even if everyone agreed. This dual structure will stay with us all the way to the Belmont Report.

The Failure

Internationally the Code was a watershed. The Geneva Conventions of 1949 and the International Covenant on Civil and Political Rights of 1966 incorporated the prohibition of medical experimentation without consent; the Code is today widely regarded as part of customary international law.

In practice, however, little happened at first. The reason is psychologically almost trivial and yet decisive: physicians and researchers in the United States, in Britain, in France read the Code as a document about war criminals. A set of rules laid down by a court to convict murderers is, on that reading, addressed to murderers, not to decent scientists. George Annas and Michael Grodin described this mechanism in the New England Journal of Medicine in 1997: the Code was admired and ignored.

This is a pattern familiar well beyond medicine. After every major data protection scandal, after every spectacular supply-chain attack, the first reaction of many organizations is: it's different here. Distancing oneself from the extreme case is the most reliable method of learning nothing from the extreme case.

Part 3: Helsinki 1964 – Physicians Write Their Own Rules

Because the Code came from lawyers, the World Medical Association wrote its own version, by physicians for physicians. The Declaration of Helsinki was adopted in June 1964 by the 18th General Assembly in Helsinki. It has since been revised ten times: Tokyo 1975, Venice 1983, Hong Kong 1989, Somerset West 1996, Edinburgh 2000, Washington 2002, Tokyo 2004, Seoul 2008, Fortaleza 2013, and Helsinki again in October 2024.

Two innovations of the 1975 version still shape practice today. First, the requirement that every research project be described in a written protocol and submitted to an independent committee for review. That is the birth of the ethics committee as an institutional body. Second, the requirement that journals should not publish work that contravenes the principles of the Declaration. With that, ethics acquired for the first time an enforcement lever that actually reaches researchers: publication.

The revision of 2000 (Edinburgh) triggered the fiercest controversy to date, because it restricted placebo-controlled trials wherever a proven effective treatment exists. The dispute flared over HIV trials in Africa in which control groups received placebo while effective therapies had long been standard in the United States. The underlying question is: is the comparator the local standard of care or the best available anywhere in the world? The Declaration chose the global best and thereby collided head-on with the practice of regulatory research.

The 2024 Revision

On October 21, 2024, the General Assembly in Helsinki adopted a version that marks several shifts going well beyond cosmetics:

• Language: subjects become participants throughout. This is not word-policing but the codification of a change in role that has been emerging for thirty years.
• Vulnerability (paragraphs 19–20): The Declaration moves away from fixed lists of "vulnerable groups." Vulnerability is understood situationally, and at the same time it is stressed that excluding such groups from research can perpetuate or exacerbate their health disparities. Protection no longer automatically means exclusion.
• Ethics committees (paragraph 23): They must be independent, able to resist pressure from sponsors and institutions, and include at least one member of the general public.
• Post-trial provisions (paragraph 34): Sponsors and researchers must make arrangements so that participants who still need an intervention identified as beneficial can obtain it after the trial ends.
• Environment (paragraph 11): Research must be conducted so as to avoid or minimize harm to the environment. This is the first ecological obligation in a core document of research ethics anywhere.

Part 4: Beecher 1966 – The Scandal Inside the House

The real rupture came not from Europe but from Boston. Henry K. Beecher, anesthesiologist at Massachusetts General Hospital and one of the most respected clinicians in the United States, published a paper on June 16, 1966, in the New England Journal of Medicine (volume 274, pages 1354–1360) under the plain title Ethics and Clinical Research.

Beecher did something nobody had done before: he collected published studies, appearing in reputable journals, from respected American institutions, and used them to show that people there had been exposed to substantial risk without disclosure and without consent. He began with seventeen examples, reached fifty without difficulty, and printed twenty-two for reasons of space. In only two of the original fifty papers was consent mentioned at all.

The examples are still hard to read sixty years later. Example 1: soldiers with streptococcal infections were denied penicillin in order to study the natural course of the illness and the risk of rheumatic fever; several developed exactly that complication. Example 16: in an institution for children with intellectual disabilities (Willowbrook State School on Staten Island), children were given hepatitis virus to study the course of infection and later immunoglobulin prophylaxis; for a period, admission to the institution was in practice tied to participation. Example 17: twenty-two elderly, chronically ill patients at the Jewish Chronic Disease Hospital in Brooklyn were injected with live cancer cells to study immune response; they were told it was a skin test, and the word cancer was deliberately avoided.

Beecher's central thesis, however, is more nuanced than its reception often suggests. He considered fully informed consent practically unattainable in many situations — "consent in any fully informed sense may not be obtainable" — and therefore relied primarily on the integrity of the investigator: the "intelligent, informed, conscientious, compassionate, responsible" researcher was, in his view, a more dependable safeguard than the form.

I am of the opinion that this was Beecher's greatest error, and at the same time the reason his paper accomplished so much. For his own twenty-two examples came without exception from intelligent, informed, respected researchers. He had thereby, without intending it, supplied the empirical proof that personal integrity fails as a sole protective mechanism. The institution that emerged was not the better researcher but the external reviewer.

The effect, in any case, was immediate: as early as 1966 the US Public Health Service required research projects to be reviewed by a committee of the investigator's own institution before they began. That is the origin of the Institutional Review Board (IRB).

Part 5: Tuskegee – Forty Years, Six Hundred Men, No Consent

What Happened

In 1932 the US Public Health Service began an observational study of the natural course of untreated syphilis in Macon County, Alabama. It enrolled 600 African American men, mostly poor sharecroppers: 399 with syphilis and 201 without, as a control group. They were told they were being treated for "bad blood," a common catch-all term at the time for a range of complaints. They received free meals, medical examinations, and burial assistance.

There is no evidence that informed consent was ever obtained. The men did not know they were part of a study, did not know they had syphilis, and did not know that the examinations offered to them served knowledge rather than their cure. The painful lumbar puncture was announced to them in a letter as "special free treatment."

From the mid-1940s onward, penicillin was the safe, curative standard therapy for syphilis. It was not offered to the participants. More than that: the study leadership took steps to prevent the men from being treated elsewhere.

How It Ended

Peter Buxtun, an employee of the Public Health Service, had raised objections internally since 1966 and had twice been rebuffed by review panels that recommended continuation. In 1972 he went to the press. On July 25, 1972, Jean Heller's report appeared on the Associated Press wire. The public outcry ended the study within months; an ad hoc advisory panel recommended termination that same year.

The consequences: in 1973 a class action that ended in a settlement of ten million dollars plus lifetime medical care for survivors and family members. In 1974 the National Research Act, which gave IRBs a statutory basis and established the national commission that five years later produced the Belmont Report. In 1997 the formal apology by President Bill Clinton at the White House.

Why This Case Lands Differently Than Nuremberg

Tuskegee could not be dismissed as the crime of a foreign regime. It was a study run by the country's own public health service, published in American journals over forty years, reviewed internally several times and each time judged worth continuing. That destroyed the narrative of the decent researcher for good.

And the damage was not merely historical. Tuskegee remains to this day a measurable factor in the mistrust of African American patients toward medical research, with concrete consequences for recruitment into clinical trials. Ethical violations leave a mortgage that outlives generations. Anyone building a system in which trust is a precondition of function should factor that in.

Part 6: Belmont 1979 – Three Principles That Created Order

The National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, established by the National Research Act, met in February 1976 for an intensive retreat at the Belmont Conference Center in Maryland. The result, published in the Federal Register on April 18, 1979, is officially titled Ethical Principles and Guidelines for the Protection of Human Subjects of Research and is known as the Belmont Report.

Its achievement is reduction: instead of a list of commandments, it formulates three principles and assigns each a concrete application.

Three things about this architecture are remarkable.

First, the separation of practice and research. The report opens by drawing the boundary between practice (intervention for the benefit of an individual patient, with a reasonable expectation of success) and research (testing a hypothesis to produce generalizable knowledge, usually under a protocol). Wherever even one element of research is present, the review obligation applies. This boundary matters so much because it runs exactly where the roles of physician and investigator diverge: the physician owes this patient the best available option, the investigator owes the public a clean result. From the physician's point of view randomization is an imposition; from the investigator's it is a duty.

Second, the decomposition of consent into three components: information, comprehension, voluntariness. This is the report's sharpest analytical contribution. A form that contains everything but is not understood satisfies only a third of the requirement. Consent that is understood but given under pressure, likewise.

Third, the elevation of justice. Tuskegee was no accident of selection. Poor, Black, uneducated farm laborers were chosen because they were available and defenseless. Prison inmates, institutionalized children, soldiers, patients in public clinics: the history of research violations is with great reliability a history of research on the powerless. The Belmont Report turns that into an independent criterion of review.

Principlism: Four Principles as a Working Language

That same year, 1979, saw the publication of Principles of Biomedical Ethics by Tom Beauchamp — himself involved in drafting the Belmont Report — and James Childress. The book adds a fourth principle to Belmont's three and arranges them differently:

1. Respect for autonomy
2. Non-maleficence
3. Beneficence
4. Justice

The decisive move is that none of these principles is absolute. They bind prima facie: each holds unless it collides with another, and in case of conflict the balance must be struck and the balancing justified. Principlism is therefore not a theory that says what is right, but a shared language in which utilitarians, Kantians, and virtue ethicists can argue about the same case without first settling their metaethics. That explains both its sweeping practical success in ethics committees worldwide and the most common criticism of it: if you never fix the weighting, you can justify almost any outcome by appeal to principles.

Part 7: The Sobering Record – Why the Signature Means Little

So far the story reads as progress: from the ineffective guideline through the code and the declaration to the statute, the committee, the catalogue of principles. Now comes the part rarely heard in commemorative lectures.

Problem 1: Nobody Can Read the Forms

Ethics committees and institutions have long recommended writing consent forms at an eighth-grade reading level. An analysis of 5,239 consent documents from studies registered on ClinicalTrials.gov reached the following result in 2024: the mean reading level was grade 10.99 (standard deviation 2.45), and 91 percent (4,768 of 5,239) of the documents were above eighth grade (Zai, Faro & Allison, Journal of Clinical and Translational Science 8:e125, 2024).

This is not a marginal problem. It means that the Belmont condition of "comprehension" is systematically missed for the overwhelming majority of clinical trials — measurably, documented, and known for decades. The reason is structural: the forms are not written for participants but for the legal department. Every lawsuit, every audit finding, every regulatory inquiry adds a paragraph. Nobody ever removes one.

The US regulatory answer to this problem is found in the revised Common Rule that took effect in 2018 (published January 19, 2017; general compliance date January 21, 2019): consent must now begin with a key information section, a concentrated summary of the points a reasonable person needs first in order to decide whether to read any further at all. That is a remarkably honest regulatory capitulation: it concedes that the rest of the document will not be read anyway.

Problem 2: The Therapeutic Misconception

Even those who read the form in full and understand it often fall prey to a misunderstanding that better wording can hardly fix. In the early 1980s Paul Appelbaum, Loren Roth, and Charles Lidz described the therapeutic misconception: participants' assumption that decisions within the study are made individually for their personal benefit, as they would be in clinical care.

This is not a gap in factual knowledge but a false framing of the situation. People are perfectly aware that randomization occurs and nonetheless assume someone will see to it that they get the right arm. They know what a placebo is and nonetheless believe their physician would not let them end up in the placebo group if it harmed them. The misconception arises because the same person in the same white coat, in the same hospital, using the same language, performs two incompatible roles. Disclosure fights against the entire context, and often loses.

Problem 3: Consent to What? The Havasupai Case

In 1989 the Havasupai Tribe, who live at the bottom of the Grand Canyon, asked Arizona State University for help: diabetes was rising dramatically in their community of roughly 650 people. Anthropologist John Martin and geneticist Therese Markow started a study; about a hundred tribal members gave blood samples and signed a consent form that spoke of studying "the causes of behavioral/medical disorders." Many of the signatories had not completed high school; for many, English was a second language.

The samples were subsequently used, without further inquiry, for research on schizophrenia, on inbreeding within the community, and on the tribe's origin and migration history. The last point was particularly wounding to the Havasupai: the genetic findings on migration across the Bering Strait contradicted their own origin account, according to which the tribe came into being in the canyon.

Havasupai Tribe v. Arizona Board of Regents ended in April 2010 with a settlement: $700,000 for 41 tribal members, return of the blood samples, scholarships, and support for building a health clinic. The decisive legal and ethical lesson is not that anyone lied, since the wording of the form arguably covered the secondary uses. The lesson is: a consent broad enough to permit everything informs about nothing. That is precisely the structural weakness of broadly framed consent, whether in genetics or in a privacy policy.

Part 8: The Present – Consent in the Age of Datasets

The original use case for informed consent was an intervention on a body at a point in time. The normal case today is a dataset that outlasts decades, that can be linked with other datasets, and whose future uses are simply unknown at the moment of collection. With that, the model's founding assumption collapses: you cannot disclose something nobody knows yet.

Three answers compete.

The 2018 Common Rule formally introduced broad consent (45 CFR 46.116(d)) and with it two new exemption categories for secondary research. What is striking is how little it is used in practice. The main reason is a design detail with large effects: if a person declines broad consent, the ethics committee may not subsequently grant that person a waiver of consent. Institutions would therefore have to track every refusal permanently and exclude it from all future analyses. Faced with that burden, many institutions have foregone broad consent altogether.

One empirical finding argues for dynamic consent and is rarely cited, and it convinced me. Pacyna and colleagues resurveyed participants in a DNA biobank and compared their answers with those given at enrollment: 1,164 usable questionnaires, a 72 percent response rate. Forty percent gave a preference about the availability of their samples at the second time point that was inconsistent with the one expressed at enrollment; in 94 percent of those cases the change was toward greater openness (European Journal of Human Genetics 28:1168–1177, 2020). Consent, then, is not a fact you establish once but a state that drifts. A rapid review by Lay, Gasparini, Siero, and Hughes (Research Ethics, 2024) screened 1,611 papers, included twelve relevant studies, and judged the obstacles to dynamic consent surmountable, provided digital channels do not remain the only ones.

The Arc Back to One's Own Field

The parallels to information technology are too obvious to pass over. Consent under the GDPR (Art. 4(11), Art. 7) demands the same three components as the Belmont Report — informed, intelligible, freely given — plus the right to withdraw at any time. And it fails at the same three points: nobody reads the notice, nobody understands the purpose description, and consent is rarely "free" when the alternative is going without the service.

The purpose limitation principle is the data protection version of the Havasupai lesson; the right of withdrawal is dynamic consent in statutory form; and the AI Act's demand for transparency and human oversight of high-risk systems is functionally what the Declaration of Helsinki introduced in 1975 with the independent committee: a body allowed to say no before the start.

The Central Takeaway

If you take one lesson from this story, take this one: consent is not a document but a state — and institutions protect more reliably than intentions. The Nuremberg Code failed on the assumption that rules suffice. Beecher erred on the assumption that character suffices. What actually worked was structural: a committee that reviews before the start, an outsider on that committee, a publication ban for violations, a statute with teeth.

Three transferable practices follow.

First: separate the roles you are currently occupying. The sharpest insight of the Belmont Report is the boundary between treatment and research, because duties change at that line. The same separation applies wherever someone simultaneously produces and reviews: anyone who approves his own code, assesses his own risk, or measures the effectiveness of his own initiative is in the position of the 1931 clinic director. Name the role, and bring in someone who has nothing to lose.

Second: measure comprehension, not the signature. The figure 10.99 against a target of 8 is a metric for a failed procedure that has been running in formally correct fashion for decades. Wherever you obtain consent — for a study, for data processing, for a permission in your software — do not ask whether people agreed but whether they understood. A key-information section at the very top, in short sentences, with the three points a reasonable person wants to know first, is cheaper than any page you append at the end.

Third: write purposes narrow enough to exclude something. A purpose description that covers every conceivable future use is the "behavioral/medical disorders" formula of the Havasupai form. It is legally convenient and informationally worthless. The test is simple: name a use that your purpose description would not cover. If you cannot think of one, your purpose informs about nothing.

And plan for change. Forty percent of biobank participants thought differently later than they did at the start. A withdrawal path as easy as the consent path is not a nicety; it is the only way to keep a consent true over time.

A Question to Reflect On

The history of informed consent has an uncomfortable through-line: every major advance followed an exposed scandal, and every scandal was caused by people who considered themselves decent and were seen that way by their peers. Beecher found his twenty-two examples not in obscure corners but in the leading journals of his day.

Suppose that in twenty years someone writes our era's Beecher paper: a collection of twenty-two entirely normal, published, approved practices in handling other people's data, at which readers will then shake their heads in disbelief. Which three practices from your own working life — in collecting, storing, reusing, or training models on other people's data — are probably on that list? And what exactly is stopping you today from ending one of them?

Cross-References in the Vault

• The Gentle Push: Why a Pre-Ticked Box Achieves More Than Any Argument — and Whether It Is Permissible to Nudge People Toward Their Own Good – the publicity principle and the question of when influence becomes manipulation; the same line that runs through the voluntariness of consent.
• Behind the Veil of Ignorance: How Rawls Reinvented Justice – and Why His Thought Experiment Still Echoes in the Ethics of Artificial Intelligence – the theoretical foundation of Belmont's justice principle and of the question who bears the burdens of research.
• Meant or Merely Foreseen: The Doctrine of Double Effect and the Power of Intention – why the actor's intention counts morally, and why, as Beecher showed, it does not suffice as the sole safeguard.
• The Child in the Road: Moral Luck and the Limits of Responsibility – the role of chance in moral judgment, which also decides which study ends as a scandal and which goes unnoticed.
• The Child in the Pond: How a Three-Page Essay by Peter Singer Launched a Global Movement — and Why the Question of How Far Our Duties Reach Echoes All the Way into the Boardrooms of Tech Billionaires – the counterposition: if research saves lives, how far does the duty to enable it reach?
• The Algorithm as Judge: Article 22 GDPR, the SCHUFA Ruling, and the Right to a Human Decision – the European continuation of the same idea: a procedural right against decisions about oneself.
• The Noise That Protects Privacy: Differential Privacy and the Art of Revealing Nothing About the Individual – the technical route to permitting secondary use without needing consent for every single case.
• The Way Out of the Cloud: Why the Data from Your Machines Belongs to You, What a Cloud Migration May Cost From January — and Why That Legal Question Has Very Technical Consequences – access rights instead of ownership rights in data, the second major answer to the same problem.
• The Programmable Scissors: CRISPR and the Rewriting of Life – the technology where the limits of consent show most sharply: who consents on behalf of future generations?
• The Pyramid of Risk: How the EU AI Act Tames Artificial Intelligence – and Why It Concerns the Whole World – prior review of high-risk systems as the functional counterpart of the ethics committee.

Sources

1. United States Holocaust Memorial Museum: The Nuremberg Code (trial USA v. Karl Brandt et al., opened December 9, 1946, verdict August 19, 1947, 23 defendants; the ten points verbatim). encyclopedia.ushmm.org; assessment: Annas, G. J. & Grodin, M. A. (1997). Fifty Years Later: The Significance of the Nuremberg Code. New England Journal of Medicine 337(20), 1436–1440. NEJM; on its reception: Beyond Nazi War Crimes Experiments: The Voluntary Consent Requirement of the Nuremberg Code at 70, PMC
2. Sass, H.-M. (1983). Reichsrundschreiben 1931: Pre-Nuremberg German Regulations Concerning New Therapy and Human Experimentation. The Journal of Medicine and Philosophy 8(2), 99–111. Oxford Academic; PubMed 6350522. Source text of the guidelines of February 28, 1931: Deutsche Digitale Bibliothek
3. Beecher, H. K. (1966). Ethics and Clinical Research. New England Journal of Medicine 274(24), 1354–1360 (22 printed examples from an original 50; consent mentioned in only 2 of the 50). NEJM; full-text PDF: Observatorio de Bioética
4. Centers for Disease Control and Prevention: About the Untreated Syphilis Study at Tuskegee (1932–1972, 600 participants: 399 with and 201 without syphilis; no documented informed consent; terminated after press coverage in 1972; Clinton's apology 1997). cdc.gov; whistleblower Peter Buxtun: PBS NewsHour
5. The National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research (1979). The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Research. Federal Register, April 18, 1979 (retreat at the Belmont Conference Center, February 1976). HHS Office for Human Research Protections
6. World Medical Association: WMA Declaration of Helsinki – Ethical Principles for Medical Research Involving Human Participants (adopted June 1964 in Helsinki; latest revision adopted October 21, 2024, in Helsinki; para. 11 environment, paras. 19–20 vulnerability, para. 23 ethics committees, paras. 25–32 consent, para. 34 post-trial provisions). wma.net – full text; press release on the 2024 revision: wma.net
7. Beauchamp, T. L. & Childress, J. F. (1979; currently 8th edition, 2019). Principles of Biomedical Ethics. Oxford University Press (four principles: autonomy, non-maleficence, beneficence, justice, each binding prima facie).
8. On the Havasupai case: Genetic Research among the Havasupai: A Cautionary Tale. AMA Journal of Ethics 13(2), 2011, 113–117 (study begun 1989, ~100 consents for "behavioral/medical disorders", secondary use for schizophrenia, inbreeding, and migration history, April 2010 settlement of $700,000). AMA Journal of Ethics; Garrison, N. A. et al. (2017). Genomic Justice for Native Americans. PubMed 28216801
9. U.S. Department of Health and Human Services: Companion Q&As about the Revised Common Rule (published January 19, 2017; general compliance date January 21, 2019; key information section, broad consent, new exemptions, single-IRB requirement). HHS OHRP (PDF); on the limited uptake of broad consent and the refusal-tracking problem: Understanding Broad Consent. Ochsner Journal 20(1), 2020, 81–86. Ochsner Journal; regulatory text: 45 CFR § 46.116
10. Zai, A. H., Faro, J. M. & Allison, J. (2024). Unveiling readability challenges: An extensive analysis of consent document accessibility in clinical trials. Journal of Clinical and Translational Science 8, e125 (5,239 documents, mean reading level grade 10.99 ± 2.45; 91% above grade 8). Cambridge Core (PDF)
11. Pacyna, J. E. et al. (2020). Assessing the stability of biobank donor preferences regarding sample use: evidence supporting the value of dynamic consent. European Journal of Human Genetics 28, 1168–1177 (1,164 questionnaires, 72% response rate, 40% inconsistent preferences, 94% of those toward greater openness). Nature; PubMed 32327712
12. Lay, W., Gasparini, L., Siero, W. & Hughes, E. K. (2024). A rapid review of the benefits and challenges of dynamic consent. Research Ethics 21(1) (1,611 papers screened, 12 studies included). SAGE Journals
13. Appelbaum, P. S., Roth, L. H. & Lidz, C. W. on the therapeutic misconception; measurement instrument: Appelbaum, P. S., Anatchkova, M., Albert, K., Dunn, L. B. & Lidz, C. W. (2012). Therapeutic misconception in research subjects: Development and validation of a measure. Clinical Trials 9(6), 748–761. SAGE Journals; overview: Clinical Trials and Medical Care: Defining the Therapeutic Misconception. PLOS Medicine 4(11), e324, 2007. PLOS Medicine

← All articles