Part 1 — The Why

A History of Research Ethics

The rules in this course can feel like a wall of requirements: review boards, consent forms, conflict declarations, registration, safety reports. Learned as a list, they are impossible to hold in your head and easy to treat as paperwork.

They are easier to understand — and much harder to forget — once you know that each one is the corrected version of a specific disaster. None of them was written in advance. Every one was written afterwards, by people looking at harm that had already happened and deciding it must not happen again.

The Problem With Good Intentions

The history of research ethics is not mainly a story about evil scientists. It is mostly a story about ordinary ones — sometimes distinguished, often well-intentioned — who believed that the importance of their goal justified their methods.

That is what makes it instructive rather than merely disturbing. The people in these cases were not cartoon villains you can be sure you are nothing like. They were researchers who convinced themselves that the knowledge they were chasing was worth the cost to the person in front of them. The frameworks exist because good intentions, on their own, have repeatedly failed to protect anyone.

This module walks the chain of failures in order. Each one exposed a gap the previous safeguards had left open, and each gap produced the next rule. By the end, the modern requirements should read less like a checklist and more like a set of scars.

Learning Objectives

After this module, you can:

Trace how each documented failure produced a specific safeguard, rather than treating the rules as arbitrary
Explain what the Nuremberg Code and the Declaration of Helsinki each established, and why
Identify the three failures the Tuskegee study exposed and the Belmont principles written in response
Recognise that consent and ethics review alone did not prevent later failures — and name the safeguards Vioxx and Gelsinger drove
Distinguish the ethical "why" from the operational "how," and locate where Swiss law sits — all in force at once

Nuremberg: Consent Becomes Non-Negotiable

At the end of the Second World War, the scale of what had been done in the name of medical research became public. In German concentration camps, physicians had conducted experiments on prisoners without their consent: immersing them in ice water to study hypothermia, depressurising them to study high-altitude physiology, deliberately infecting wounds, and testing surgical procedures — many of which killed or permanently maimed the people subjected to them.

In 1946–1947, twenty-three German physicians and administrators were tried before a United States military tribunal at Nuremberg. Sixteen were convicted; seven were sentenced to death.

Case study: the Nuremberg Doctors' Trial

The defence at Nuremberg was not that the experiments had not happened. It was that they were not categorically different from research being done elsewhere. The defendants argued that experimentation on prisoners without consent was an accepted practice in several countries at the time, and that the potential military and medical value of the results justified the methods.

The tribunal rejected this argument completely. In doing so it established a principle that sits underneath everything in this course: that an ethical obligation to a research participant does not dissolve because the practice is common, because the science is valuable, or because the participant is someone society has decided to disregard. The standard is not what other researchers are doing. The standard is the rights of the person being experimented on.

As part of its judgment, the tribunal set out ten points for permissible human experimentation. These became known as the Nuremberg Code (1947) — the first time the ethics of research were stated as binding principles in international law rather than left to professional custom.

The Nuremberg Code was aimed squarely at the most extreme abuses, and on consent it was decisive. But it was written by a war-crimes tribunal, not a system for running everyday medical research. It said nothing about who should review a study before it starts, nothing about conflicts of interest (a researcher's stake in a particular result), and nothing about what happens when the people being studied are not prisoners but ordinary patients who trust their doctors. Those gaps would be filled — but, each time, only after they had been paid for.

Helsinki: The Profession Writes Its Own Rules

The Nuremberg Code came from outside medicine, imposed by judges. The medical profession's own answer came in 1964, when the World Medical Association adopted the Declaration of Helsinki — a statement of ethical principles for medical research, written by physicians for physicians.

Helsinki accepted Nuremberg's foundation of voluntary consent and then went further, addressing the everyday machinery of research that a war-crimes judgment had never needed to touch. It has been revised repeatedly as new problems emerged; the current version was adopted in 2024, and it remains the primary international ethical reference for medical research involving people.

The most important advances Helsinki introduced over the Nuremberg Code were these:

Independent ethics-committee review before a study begins. A research protocol (the study's written plan — design, procedures, and justification) must be submitted to an independent committee for consideration and approval before the research starts, and no later change to the protocol may be made without that committee's approval. Declaration of Helsinki
A continuous risk–benefit duty. Research is only permissible if the importance of the objective outweighs the risks and burdens, and those must be monitored throughout. When the risks are found to outweigh the benefits, the researcher must assess whether to continue, modify, or stop immediately. Declaration of Helsinki
Scientific soundness as an ethical requirement. A study must have a rigorous design based on a thorough knowledge of the literature and adequate prior laboratory and, where appropriate, animal work. Bad science is not ethically neutral — it burdens participants for nothing. Declaration of Helsinki
Specific protection for people in situations of vulnerability, who may be at greater risk of being wronged and need considered safeguards. Declaration of Helsinki

It also set limits Nuremberg never reached: on when a placebo may be used instead of proven treatment, on the obligation to arrange continued access to a beneficial intervention after a trial ends, and on registering and honestly publishing research. Those last two — registration and publication — look like housekeeping until you see what happens without them, which is the subject of the Vioxx case later in this module.

The rule

Declaration of Helsinki (2024), Article 7: the purpose of research "can never take precedence over the rights and interests of individual research participants." Declaration of Helsinki

Article 9: the responsibility for protecting participants "must always rest with physicians or other researchers and never with the research participants, even though they have given consent." Declaration of Helsinki

Article 9 is the one to hold on to. It means consent does not transfer the duty of protection onto the participant. When someone signs a consent form, the obligation to look after them does not move to them — it stays with you. This applies whatever your role: a study nurse or coordinator who notices something wrong carries a duty to act on it, even though ultimate responsibility for the trial sits with the investigator.

Helsinki gave medicine a coherent ethical framework: review before you start, weigh risk against benefit continuously, protect the vulnerable, keep the participant's interests above the study's. On paper, by 1964, the protections looked broadly modern. The next failure showed how little "on paper" is worth.

Tuskegee and Belmont: When the Rules Already Existed

The hardest case in this history is not one where the rules had not yet been written. It is one where they had — and a public health service ignored them for decades.

Case study: the Tuskegee Syphilis Study (1932–1972)

In 1932 the United States Public Health Service began a study in Macon County, Alabama, of the natural course of untreated syphilis. It enrolled around 600 Black men — roughly 400 who had syphilis and about 200 without it as a comparison group.

The men were never told they had syphilis. They were told they were being treated for "bad blood," a vague local term, and were given placebos, ineffective tonics, and diagnostic procedures described to them as treatment. The actual purpose was to observe the disease as it progressed, untreated, to death and autopsy.

The study was designed to withhold treatment — but in its early years there was no reliably effective one. That changed in the 1940s. By 1947 penicillin was the established, curative standard of care for syphilis. The study did not stop. The men were not offered penicillin; investigators actively worked to keep them from receiving it elsewhere. The study continued for another twenty-five years.

It ended in 1972 only because a Public Health Service employee, troubled by what he had seen, gave the story to the press. By then many of the men had died of syphilis, and the disease had been passed to wives and children. The study ran for forty years, through the writing of the Nuremberg Code and the Declaration of Helsinki, entirely within a government health agency.

Tuskegee produced three distinct failures, and it is worth separating them, because each maps onto a protection in force today:

Deception in consent. The men were lied to about their diagnosis and about what was being done to them. There was no informed consent in any meaningful sense — the opposite of Nuremberg's first principle.
Withholding available treatment to preserve the data. Once penicillin existed, every additional day of observation was bought with a participant's health. The study's scientific goal was placed directly above the welfare of the people in it.
The systematic targeting of a vulnerable population. The study was built on the assumption that poor Black men in the rural South could be enrolled, deceived, and denied care without consequence — that their interests counted for less.

The public reckoning, when it came, was severe. The United States passed the National Research Act (1974), which created the system of institutional review boards, and commissioned a report on the ethical principles that should govern all research with human subjects.

The rule

That report — the Belmont Report (1979) — distilled the lesson of Tuskegee into three principles that remain the ethical backbone of human-subjects research. Belmont Report

Respect for persons: individuals are treated as autonomous agents and given the information and freedom to decide for themselves; those with diminished autonomy are entitled to extra protection. (Tuskegee violated this through deception.)
Beneficence: maximise possible benefit and minimise harm. (Tuskegee violated this by withholding a known cure.)
Justice: the benefits and burdens of research are distributed fairly, not loaded onto whoever is easiest to exploit. (Tuskegee violated this by targeting a population it expected to face no consequences for harming.)

These principles were not produced in a seminar. Each one is the negative image of something that was actually done to real people. You will meet them again, expanded into the four-principle framework that underpins Swiss research law, in Module 6.

The lasting lesson of Tuskegee is not that the rules were missing. The Declaration of Helsinki had existed since 1964, and the men were still being denied penicillin in 1972. A framework that no one is accountable for enforcing protects no one. That is precisely why the modern system does not stop at stating principles — it attaches them to named people, independent review, and inspection. Good principles without enforcement are exactly what Tuskegee had.

The Belmont Report named three principles, each the negative image of one of Tuskegee's failures. Which pairing matches each failure to the principle it most directly offends?

Thalidomide: Testing Before Market

The failures so far were about how participants are treated. The next one was about something different: whether a drug has been adequately tested before it is given to anyone at all. It reshaped not research ethics but the entire system of drug regulation that sits around clinical trials.

Case study: thalidomide (1957–1961)

Thalidomide was marketed from 1957, first in West Germany and then across much of Europe and beyond, as a sedative that was also effective against the nausea of early pregnancy. It was promoted as remarkably safe — supposedly impossible to overdose on — and was widely taken by pregnant women.

Its effects on the developing fetus had never been adequately tested. Taken during a specific window of early pregnancy, thalidomide interfered with the development of limbs and organs. An estimated 10,000 children were born with severe malformations — most characteristically, severely shortened or absent limbs — and many more pregnancies ended in miscarriage and stillbirth before the cause was identified in 1961.

It could have been far worse. In the United States, a medical reviewer at the regulator refused to approve thalidomide, unsatisfied with the safety data and concerned about reports of nerve damage in adults. Her refusal held the drug off the U.S. market and spared a generation of American children — and showed exactly what an independent gatekeeper, empowered to demand evidence, is for.

Thalidomide exposed a gap that consent rules and ethics committees did not address: the inadequacy of pre-market testing. The drug had been approved on limited animal and human data, with no systematic testing in the population most affected — pregnant women — and no requirement to prove safety in a relevant model before exposing the public on a massive scale.

The regulatory response was structural and permanent. Countries rewrote their drug laws to require proof of both safety and efficacy, supported by adequate pre-clinical data, before a medicine could be sold. This is the principle the Declaration of Helsinki now states as an ethical requirement in its own right: that research rest on a thorough knowledge of the literature and adequate prior laboratory and animal work. Declaration of Helsinki The drive to harmonise these national rules across countries eventually produced the international Good Clinical Practice standard — ICH-GCP — whose structure and content are the subject of Module 8.

Vioxx: When the Data Itself Is the Problem

By the end of the twentieth century the framework looked complete. Consent was required, ethics committees reviewed protocols, drugs were tested for safety and efficacy before approval. The next failure happened anyway — and it happened not because a rule was missing but because the evidence a rule depends on had been distorted.

Case study: Vioxx (1999–2004)

Vioxx (rofecoxib) was a painkiller approved in 1999 and marketed by Merck for arthritis and acute pain. It belonged to a class designed to relieve pain with less stomach irritation than older anti-inflammatories, and it became one of the most widely prescribed drugs in the world, taken by millions of people.

In 2000 a large trial called VIGOR, designed to study Vioxx's gastrointestinal safety against an older drug (naproxen), found that participants on Vioxx had substantially more heart attacks. Rather than reading this as a signal that Vioxx might raise cardiovascular risk, the result was publicly explained the other way around — as evidence that naproxen was unusually protective of the heart. When the VIGOR results were published, the reported count of heart attacks did not include several cardiovascular events that were known before submission; the journal that published it later issued a formal "expression of concern" about the omission.

The drug stayed on the market for four more years. It was finally withdrawn in 2004, after a separate trial (run to test whether Vioxx could prevent colon polyps) showed a clear increase in heart attacks and strokes among long-term users — the same signal VIGOR had produced, now impossible to explain away. Merck withdrew Vioxx worldwide.

An FDA safety reviewer subsequently estimated that Vioxx may have contributed to tens of thousands of excess heart attacks and strokes in the United States alone — an estimate, not a precise count, but one that conveys the scale. The litigation that followed ran for years.

The failure in Vioxx was not a lack of consent and not an absent ethics committee. It was the integrity of the evidence. A safety signal was reinterpreted to protect a product, an unfavourable result was presented incompletely, and the surveillance that should have caught the problem in marketed use was too slow. Patients and prescribers cannot weigh a risk they have not been honestly shown.

Vioxx is the reason several modern safeguards exist in the form they do: mandatory trial registration before enrolment, the duty to publish unfavourable results, and active monitoring of a drug's safety after it reaches the market. The day-to-day disciplines that protect evidence at your level — accurate, complete, unalterable records — are the subject of Module 12, and how safety signals must be reported during a trial is Module 9. The principle underneath all of them is what Vioxx violated: the data must tell the truth, including the parts no one wants to hear.

Gelsinger: Conflicts of Interest and Honest Reporting

The last case is the most modern, and it brings together two threads the others raised but never centred: what happens when the people running a study have a financial stake in its success, and what happens when known risks are not honestly disclosed.

Case study: Jesse Gelsinger (1999)

Jesse Gelsinger was an eighteen-year-old with a genetic liver disorder, OTC deficiency, which left his body unable to fully clear ammonia. His was a comparatively mild form, controlled with a strict diet and medication. In September 1999 he volunteered for an early-phase gene-therapy trial at a university institute, testing the safety of delivering a corrected gene using a modified virus. Because his condition was stable, he stood to gain nothing medically himself; he took part to help develop a treatment for severely affected newborns.

Four days after the modified virus was infused, Jesse died of a massive immune reaction that led to multiple organ failure. He was the first person widely recognised to have died as a direct result of gene therapy.

The investigation that followed found a series of failures that had nothing to do with the inherent risk of the science:

Known risks had not been disclosed. Earlier participants had suffered serious reactions, and animals had died at higher doses in prior studies. This information had not been adequately reported to regulators or reflected in what volunteers were told before consenting.
Eligibility rules appear to have been broken. Jesse's blood-ammonia level at the time of dosing may have exceeded the limit the protocol set for entry — meaning he arguably should not have been treated that day.
There was a financial conflict of interest. The lead researcher and the university held equity stakes in a company positioned to profit if the therapy succeeded. That interest was not adequately managed or made clear to participants.

The regulator shut down the institute's human trials. The case became the reference point for reforming how financial conflicts of interest are disclosed and managed in human research, and how honestly risk must be reported.

Gelsinger is not a story about a reckless field. Early-phase trials of genuinely novel therapies are necessary and are run safely all the time. It is a story about what a financial stake in the outcome can do to judgement — about the pressure to keep a promising programme moving, to read eligibility generously, and to not dwell on the bad news. The harm did not come from the experiment being attempted. It came from the people running it having reasons, some of them financial, to see it succeed.

How the Layers Fit Together

Read separately, these cases are six disasters. Read in order, they are a single argument: each failure exposed a gap the last set of rules had left open, and each gap was closed by a new layer of protection. Nuremberg secured consent. Helsinki added independent review and a continuous duty of care. Tuskegee showed that principles without enforcement are worthless, and Belmont named the principles. Thalidomide forced real pre-market testing. Vioxx demanded honest data and registration. Gelsinger demanded that conflicts of interest be disclosed and risks reported truthfully.

Those layers did not replace one another. They stack, and for a clinical trial in Switzerland today they apply at the same time, each doing a different job:

The Declaration of Helsinki is the ethical why — the principles, and the reasons behind them.
ICH-GCP (E6(R3)) is the operational how — the international standard that turns those principles into auditable, day-to-day practice. Its structure is the subject of Module 8.
Swiss law is the legal must — binding obligation enforced by the authorities. The Human Research Act (HRA) places the interests of the individual above those of science and society (Art. 4), the same principle Helsinki states in Article 7; and the Clinical Trials Ordinance (KlinV) requires that trials be conducted in accordance with Good Clinical Practice (Art. 5). HRA KlinV The full Swiss legal framework is the subject of Module 6.

1947
Nuremberg Code
Voluntary consent as an absolute precondition
1962 onward — drug-law reform
National drug-law reform (after thalidomide, 1957–1961)
Proof of safety and efficacy required before a drug may be marketed
1964 → 2024
Declaration of Helsinki
Independent review before a study begins; a continuous risk–benefit duty; honest publication
1974–1979
National Research Act; Belmont Report
Review boards; the principles of respect, beneficence, and justice
2004 onward
Post-Vioxx transparency reforms
Trial registration; publication of negative results; post-market surveillance
2025
ICH E6(R3)
The current operational GCP standard

None of this is bureaucracy for its own sake. Every entry in that timeline is the place where someone was harmed badly enough that the system changed to stop it happening again. When a rule in this course seems excessive, the useful question is not "is this necessary?" but "which failure is this the answer to?" — because there almost always is one.

Module Summary

Every rule in clinical research is the corrected version of something that went wrong. Learned as history, the requirements stop being an arbitrary list and become a connected argument — which is also the easiest way to remember them, and to apply them when a situation arises that no checklist anticipated.

You can now:

Trace how each documented failure produced a specific safeguard, rather than treating the rules as arbitrary
Explain what the Nuremberg Code and the Declaration of Helsinki each established, and why
Identify the three failures the Tuskegee study exposed and the Belmont principles written in response
Recognise that consent and ethics review alone did not prevent Vioxx or Gelsinger — and name the safeguards they drove
Distinguish the ethical "why" from the operational "how," and locate where Swiss law sits — all in force at once

If there is one thing to carry forward: the protections are not there to slow you down. They are there because, every time they were absent, someone paid for it — and the whole point of knowing the history is to recognise the next gap before it costs anyone again.

Last reviewed 2026-04 against DECLARATION_OF_HELSINKI · ICH_E6_R3