Part 1 — The Why

The Importance of Research and Research Ethics

Medicine spent most of its history confident and wrong. The treatments that failed were rarely the work of bad doctors — they were the work of good ones who trusted experience and tradition without ever putting them to a fair test.

Clinical research exists to solve that problem. But when you enrol people in a trial — asking them to accept experimental risks in order to generate knowledge for people they will never meet — you take on obligations that are not optional, not transferable, and not discharged by good intentions alone.

Learning Objectives

After this module, you can:

Explain why untested clinical intuition has repeatedly caused harm — and what replaced it
Spot the tension between scientific progress and participant protection when it surfaces in a real decision
Apply the three conditions that make research on humans justifiable, and spot when one has collapsed
Locate where ultimate responsibility for a trial sits, and why it cannot be delegated away
Identify what GCP is, which version is in force, and where its requirements come from

Why Medicine Needs Systematic Testing

Bloodletting was practiced for two thousand years and probably killed more patients than it helped. Radical mastectomy — removing not just the breast but the chest wall muscles and axillary lymph nodes — was the standard surgical treatment for breast cancer for most of the twentieth century. It caused significant disfigurement and long-term disability. When a randomised trial was eventually done, it showed the procedure offered no survival benefit over far less invasive surgery.

In each case, the harm was not the product of malice. It was the product of treating clinical impression as evidence — and clinical tradition as proof.

Clinicians saw patients who recovered and concluded the treatment worked. They could not easily see the patients who would have recovered regardless, or the ones who did not survive to return. This is not a failure of intelligence. It is how human cognition works when data is not systematically collected and compared. It is precisely why systematic testing exists.

What this looks like in practice

You're working on a new trial of an anti-inflammatory compound. The preclinical data look solid; the protocol is well-designed. Three months after enrollment opens, the sponsor's safety team sends an updated Investigator's Brochure (the sponsor's safety dossier on the compound) noting a new signal from a companion study: elevated liver enzymes in a subset of participants at higher doses.

Your enrolled participants are on a lower dose. You have not seen anything at your site. Your first instinct might be: we are fine, let's continue.

That instinct is the same one that kept radical mastectomy in practice for decades. The question is not whether you have personally observed harm — it is whether the evidence base that originally justified enrollment is still intact, and a new safety signal in a related population changes that calculus.

So you act on it — now, not at the next scheduled visit. You make sure the new risk information is documented and the signal reaches the sponsor, that the question of changing dose or monitoring is raised, and that whether those already enrolled need re-consenting gets checked. How much of that is yours to decide and how much to escalate depends on your role; the trigger for action does not — it is the change in the evidence base, not a harm event at your site.

The Tension at the Heart of Clinical Research

Clinical research contains a structural conflict that cannot be designed away. The goal of research is to generate knowledge that benefits future patients. The obligation of the researcher is to protect the participant in front of them, now. These two imperatives are not always aligned.

This is not a design flaw. It is an honest description of what research is. You are asking people to accept uncertainty — sometimes real risk — for the potential benefit of people they will never meet. That is a significant thing to ask of another human being.

The rules in this course do not eliminate this tension. They define the conditions under which it is acceptable. They mark the point at which the scales tip — where the risk to this participant outweighs the potential benefit to future ones — and specify what you are required to do when you reach it. Your job is not to resolve the tension. Your job is to navigate it honestly, with your obligations to the participant as the non-negotiable baseline.

The rule

Declaration of Helsinki, Article 7: The primary purpose of medical research involving human participants is to generate knowledge to advance individual and public health. These purposes can never take precedence over the rights and interests of individual research participants.

Declaration of Helsinki, Article 9: It is the duty of physicians involved in medical research to protect the life, health, dignity, integrity, autonomy, privacy, and confidentiality of personal information of research participants. The responsibility for the protection of research participants must always rest with physicians or other qualified individuals — never with the research participants, even if they have given consent. Declaration of Helsinki

Article 9 is worth reading carefully. The obligation to protect participants does not transfer to the participant when they sign a consent form. It stays with you.

What this looks like in practice

It is Friday at 5pm. A study coordinator calls: a participant enrolled in your Phase II oncology trial (early efficacy-and-safety testing) has just been told by their treating oncologist that a new standard-of-care treatment became available — one that did not exist when they consented. The participant wants to withdraw from the trial immediately and start the new treatment.

You are mid-enrollment. Losing this participant will affect data completeness. The sponsor's project manager has already sent a message asking whether this will affect the timeline.

The rules are unambiguous. The right to withdraw at any time, without penalty or disadvantage, is absolute — guaranteed in Swiss law, and covered in depth in Module 7. HRA The participant's welfare — including access to newly available standard treatment — is your primary obligation. It is not a factor to be weighed against data completeness or sponsor timelines.

Notice if you feel even a moment's pull toward discouraging the withdrawal. That pull is the structural tension making itself felt. The rules were not written for the easy cases. They were written for exactly this one.

The Three Conditions That Justify Research on Humans

Why is it ethically acceptable to expose research participants to the uncertainties and risks of an experimental intervention?

Because the alternative — treating patients with interventions that have not been properly tested — carries its own risks, as the history of medicine makes clear. Knowledge generated from well-designed research can prevent harm at a scale that justifies the burden placed on present participants.

But this justification only holds when three conditions are satisfied simultaneously. None of them is optional. If any one collapses, the ethical foundation of the study collapses with it.

Here is the complete set, before going deeper on each:

The research question has genuine scientific merit
The risk-benefit balance is favourable
Participation is truly voluntary

These are not independent virtues that can compensate for each other. A scientifically sound study with coerced participants is not ethically justifiable. A study where participants consented freely but whose design cannot generate useful knowledge burdens those participants for nothing. A study with both merit and voluntary participation, but an unfavourable risk-benefit balance, cannot ethically be conducted. All three must hold, at the same time, throughout the trial.

Condition 1: Genuine Scientific Merit

A study designed to confirm a predetermined answer is not research — it is marketing with a protocol. A study whose results will only be published if they are favourable cannot justify the burdens it places on participants. A study that replicates what is already well-established, without adding meaningful new knowledge, wastes participants' risk and time.

Scientific merit is assessed by the ethics committee before a trial begins. But it is not simply an administrative gate. It is the first ethical question you should ask yourself before agreeing to run a study: if it cannot realistically generate knowledge that will be used — because the question is wrong, the design is flawed, the sample is too small, or the sponsor has no intention of publishing unfavourable results — there is no ethical justification for what it asks of participants.

In Switzerland, the ethics committee is required to assess scientific validity as part of its review. A protocol that cannot answer its own question will not receive approval. KlinV

Condition 2: Favourable Risk-Benefit Balance

Favourable does not mean zero risk. Almost every clinical trial involves some risk; the question is whether the potential benefits — direct benefits to participants and potential benefits to future patients — are proportionate to the risks and burdens imposed.

Risk-benefit is not a one-time calculation completed at protocol approval. It is a continuous obligation throughout the life of the trial. If new safety data emerges mid-trial — a signal from a companion study, a finding published in the literature, an unexpected pattern in your own safety data — the balance may shift. If the standard of care changes and the control arm is no longer clinically defensible, you may be obligated to amend or stop.

The moment the risk-benefit balance tips against continued participation, you have an obligation to act. Not to note it and continue. Not to wait for the next scheduled monitoring visit. Swiss law also sorts trials into risk categories that scale the regulatory requirements to the risk involved — covered in depth in Module 6.

Condition 3: Truly Voluntary Participation

Voluntary means more than "no one directly threatened the participant." It means:

Free from undue inducement — payment or benefits so substantial that they compromise the person's ability to weigh the risks freely
Free from therapeutic misconception — the mistaken belief that research participation is equivalent to receiving individualised medical treatment, rather than an intervention whose primary purpose is to generate knowledge
Free from institutional coercion — the particular risk when participants are the investigator's own patients, students, or employees, and may feel implicit pressure to participate because of the relationship

If participation is not genuinely free, the entire ethical justification for the research dissolves, regardless of how sound the science is.

A trial is scientifically rigorous, has ethics-committee approval, and poses no more than minimal risk. To recruit fast, the investigator enrols her own clinic patients, telling each that she 'would be very grateful for their help' and that taking part 'can only be good for their care.' Enrolment fills quickly. Which of the three conditions has collapsed?

What this looks like in practice

A new sub-investigator joins your site. During onboarding, she asks: "If I follow the protocol exactly and act in good faith, am I protected if something goes wrong?"

The answer is: following the protocol is necessary but not sufficient. The protocol reflects the risk-benefit calculation made when it was designed and approved. Your obligation is to the participant in front of you — which means that if you observe something the protocol did not anticipate, or if the evidence base underpinning the protocol changes, you have an obligation to act on it.

The protocol is not a ceiling on your ethical obligations. It is a floor.

This applies to consent as well. A participant may have signed a valid consent form at enrollment. If the study has materially changed since then — new safety information, a protocol amendment, a changed risk profile — the consent process may need to be repeated. The form does not do your ethical work for you.

Common mistakes

Treating ethics committee approval as permanent clearance. Approval is based on evidence and design at the time of submission. New data changes the picture. The obligation to reassess does not expire when the trial opens.
Equating a lower risk category with fewer ethical requirements. Risk category governs which specific procedures apply — it does not waive any of the three conditions. A low-risk study still requires genuine scientific merit, a favourable risk-benefit balance, and truly voluntary participation.
Letting the consent form do the ethical work. A signed form documents that a consent process occurred. It does not establish that participation was genuinely voluntary, that the participant fully understood the risks, or that the three conditions remain satisfied as the trial progresses.

Who Is Responsible

Every clinical trial site has one person who is ultimately responsible for the conduct of the trial at that site: the Principal Investigator. This is not an administrative designation. It is a legal and ethical role with specific obligations that cannot be transferred, proportionally shared, or satisfied by delegation alone. The PI may delegate trial-related activities, but retains ultimate responsibility and must maintain appropriate oversight of them; the mechanics of delegation and the delegation log are covered in Module 3. ICH E6(R3)

Where you stand in relation to that responsibility depends on your role. If you are the PI, it is yours, and it cannot be delegated away. If you are a sub-investigator, study coordinator, or study nurse, it is precisely not yours — and your duty is to understand where that line falls, to work within the scope delegated to you, and to flag uncertainty rather than assume.

What this looks like in practice

A PI delegates eligibility screening and consent discussions to an experienced study coordinator and a sub-investigator — reasonable, and exactly what delegation is for. Six weeks in, a monitoring visit finds that two enrolled participants did not meet a key inclusion criterion: the sub-investigator had misread it, and no one had checked those files against the protocol.

The finding is recorded against the PI — not the coordinator, not the sub-investigator who misread it. Delegating the task moved the work; it did not move the accountability. What the PI cannot hand off is the oversight: the spot-checks, the clear delegation log, the culture where staff flag uncertainty instead of guessing. Build that system and it holds; skip it and the gap is the PI's — no matter whose hands the task passed through.

When things go wrong in clinical trials — when participants are harmed, when data are falsified, when consent is bypassed — an investigator is almost always implicated. Not always through deliberate misconduct. More often through inadequate training, under-resourced teams, unclear delegation, or a gradual normalisation of shortcuts over time.

This course exists to prevent that kind of harm — not by giving you a compliance certificate, but by giving you the framework, the vocabulary, and the reasoning to ask the right questions before things go wrong.

Every safeguard in this course is a responsibility attached to a specific person. Part of your job — whatever your role — is knowing exactly which of those responsibilities are yours, and where the line runs to the person who carries ultimate responsibility for the site.

What Is GCP?

Good Clinical Practice (GCP) is an international quality standard for designing, conducting, recording, and reporting clinical trials. It exists to ensure two things simultaneously: that the data generated are credible enough to inform regulatory decisions, and that the people who contributed to generating them were protected throughout.

GCP is not a single law. It is a set of principles agreed internationally and then incorporated into national regulation. The international standard is ICH E6, published by the International Council for Harmonisation (ICH) — a body that brings together regulators and industry representatives from the US, Europe, and Japan to align standards across jurisdictions. ICH guidelines are adopted globally and form the basis of national GCP requirements in Switzerland, the EU, the US, and most other major markets.

The current version is ICH E6(R3), which came into force in August 2025 and superseded E6(R2). ICH E6(R3) The revision introduced a more explicit risk-proportionate and quality-by-design approach — more emphasis on critical thinking and less on one-size-fits-all procedures. This course is built to E6(R3). Training records citing only E6(R2) are no longer adequate.

GCP covers the full lifecycle of a clinical trial. It defines responsibilities for the sponsor who designs and funds the trial, the investigator who conducts it, and the ethics committee that reviews it. It sets out requirements for the protocol, informed consent, data integrity, adverse event reporting, monitoring, and essential documents. The structure of ICH E6(R3) — its sections, its principles, and how they translate into day-to-day practice — is the subject of a dedicated module later in the course. What matters here is this: when this course says something is "required," that requirement comes from GCP, from Swiss law, or from both.

What inspectors look for

Can the PI articulate why the trial they are running is scientifically necessary — not just that it was approved, but what question it answers and why that question is worth the burden placed on participants?
Is there documented evidence that the risk-benefit assessment was conducted before enrollment opened, and that it has been formally revisited when new safety information became available?
Is GCP training current and explicitly scoped to ICH E6(R3)? Training records referencing only E6(R2) are no longer adequate.
Does the PI understand the distinction between ethics committee approval — which covers the study design at the time of submission — and their continuous personal obligation to participants throughout the trial?
Does the site's delegation log accurately reflect who is responsible for which tasks, and does the PI understand that delegation does not transfer their ultimate accountability for site conduct?

Module Summary

Clinical research is necessary. It is ethically justifiable under specific conditions. And those conditions must be actively maintained throughout the life of the study — not established at approval and then assumed to hold.

You can now:

Explain why untested clinical intuition has repeatedly caused harm — and what replaced it
Spot the tension between scientific progress and participant protection when it surfaces in a real decision
Apply the three conditions that make research on humans justifiable, and spot when one has collapsed
Locate where ultimate responsibility for a trial sits, and why it cannot be delegated away
Identify what GCP is, which version is in force, and where its requirements come from

If there is one thing to carry into the rest of this course: the participant in front of you is not a means to an end. The knowledge generated by research serves future patients. But your obligation to protect the person enrolled in your trial is unconditional — not conditional on the sponsor's timeline, not conditional on what their withdrawal would cost the dataset, not conditional on anything.

Last reviewed 2026-06 against ICH_E6_R3 · DECLARATION_OF_HELSINKI · HRA · KLINV