Preventing coercion in human trials
Participants join human trials for a wide range of reasons: some are motivated by personal or family health experiences, others by financial compensation, and many by a combination of both. This creates a genuine ethical challenge. How do you make participation accessible and attractive without crossing into territory that compromises voluntary consent?
Understanding the difference between coercion and undue influence
These two concepts are often conflated, but they describe genuinely different problems, and the confusion isn't limited to people outside the field. A national survey of IRB members and human subjects protection professionals found expansive and often inconsistent views about the two, with many respondents treating coercion and undue influence as points on the same continuum. The researchers argued this is a category error: coercion is specifically a threat of harm, while undue influence is a cognitive distortion that skews how someone weighs risk and benefit. Because a payment offer is an offer rather than a threat, the paper's authors argue payment itself is never technically coercive, even though it can still amount to undue influence if it's large enough to override someone's honest risk assessment. That distinction matters in practice, not just semantically.
Coercion involves threats or pressure that eliminate real choice. Examples include:
- A clinician implying that refusing to participate could affect the quality of care received
- An employer or benefits provider making participation a condition of something else
Undue influence is more subtle. It happens when incentives or persuasion are so powerful that they override a person's ability to weigh up risks objectively. Examples include:
- Payments large enough to make economically disadvantaged participants overlook risks they would otherwise decline
- Language that implies participation guarantees better outcomes or access to treatments not otherwise available
Both undermine the foundation of ethical research: the idea that people choose to participate freely and with genuine understanding. But they call for different fixes. Coercion is addressed by removing threats and power imbalances from the situation entirely. Undue influence is addressed by calibrating incentives so they inform a decision rather than overwhelm it.
| Coercion | Undue influence | |
|---|---|---|
| Mechanism | A threat of harm | A cognitive distortion in weighing risk and benefit |
| Can payment cause it? | No, an offer isn't a threat | Yes, if large enough relative to the participant's circumstances |
| Fix | Remove the threat or power imbalance | Calibrate the incentive to time and inconvenience, not motivation |
Practical strategies for ethical conduct
Strengthen the consent process. Plain language, interactive Q&A, and adequate time to decide are the basics. In longer studies, revisiting consent periodically matters too. People's situations and understanding change over time.
Get compensation right. Payment should reflect time, inconvenience, and out-of-pocket costs. It should not be structured as a reward for completion alone, and it should be communicated transparently so participants understand it before they decide to enrol.
Make withdrawal genuinely consequence-free. This must be stated clearly and repeatedly. If participants believe withdrawing will affect their medical care, access to services, or relationship with the research team, the consent process has failed.
Train the people who interact with participants. Ethics training should not be a box ticked at onboarding. Regular refreshers, and genuine attention to how participant conversations actually go, help teams recognise and avoid patterns that drift toward pressure.
Take power dynamics seriously. In studies involving patient populations, employees, or community members with existing relationships to the research team, extra care is needed. Independent advocates, ethics committee oversight, and third-party support for participants provide some protection.
Use technology to support transparency. Digital consent platforms with video explanations, readable summaries, and anonymous feedback options give participants more control over how they absorb information and raise concerns without feeling observed.
Protect vulnerable populations specifically. Children, people with cognitive impairments, and others who may have limited capacity to assess risks independently need additional safeguards: guardian oversight, tailored communication, and more conservative incentive structures.
Build accountability into the process. Regular ethical audits, participant feedback mechanisms, and published outcomes all contribute to a culture where ethical conduct is expected rather than assumed. The goal is not just to avoid obvious violations but to maintain an environment where participants genuinely feel that their welfare comes first, and where the team itself has a clear, shared understanding of exactly which risk it's actually protecting against at each stage of the study.