Broad Consent and Its Limits

Part of Consent and Vulnerable Populations: Data, Samples and Secondary Use.

About 6 minutes. Free, in your language, no account needed.

Why this matters

Broad consent asks a person to agree, now, that their samples or data may be used for future research that cannot yet be described specifically.

It is genuinely useful. Biobanks and cohorts would be impossible without it, and much valuable research depends on material collected before the question was known.

It is also the consent mechanism that most strains the concept, because a person is agreeing to something nobody can fully describe, which sits uncomfortably with informed consent's central requirement.

What makes broad consent defensible

Not the breadth. The governance around it.

A described scope. Not "any future research," which is not a description. A stated field, disease area, or type of research, so the person knows roughly what they are agreeing to.

An oversight body that reviews each use. The participant is not approving each study; a committee is, against the scope the participant agreed to. That committee is doing the work the participant cannot, and its independence and diligence are what make the arrangement honest.

Stated limits. What will not be done. Commonly: no recontact without further consent, no transfer outside defined boundaries, no use for specified purposes the participant may object to.

A real withdrawal route, known to participants and usable, with an honest account of what withdrawal can and cannot achieve.

Transparency about what has actually been done. Participants able to find out what their material has been used for.

Broad consent without these is not broad consent. It is an open-ended permission with nobody exercising judgment on the participant's behalf.

The uses people actually object to

Research suggests most people are willing for their samples to be used broadly, and that their willingness is not unconditional. The objections cluster:

Commercial use, particularly where a company profits and the participant does not. This is the most consistently raised.

Research they consider objectionable, which varies by person and may include particular topics, and this is why a described scope matters more than the breadth.

Transfer to other countries, particularly where the participant has reason to distrust the destination.

Identification, including re-identification from supposedly anonymous data, which is increasingly feasible.

Uses affecting their group, including research producing findings about a population that the population considers stigmatizing. This one cannot be addressed by individual consent at all and is why community engagement matters for research on identifiable populations.

A consent process that is silent on these is silent on exactly what participants care about.

Genetic data and the group problem

Genetic data is about a person and also, necessarily, about their relatives, who never consented.

A finding about an individual can carry implications for siblings, children and parents. A finding about a population can carry implications for everyone in it.

Individual consent cannot resolve this, because the affected parties are not present and cannot be. The mechanisms that partly address it are community engagement for population-level research, governance restricting what may be studied and published, and honesty with participants that their data carries information about people who did not agree.

The withdrawal that cannot fully work

Participants are usually told they may withdraw.

For samples and data, withdrawal is partial. A sample already consumed cannot be returned. Data already included in a completed and published analysis cannot be removed. Data already shared with another researcher may not be retrievable.

What can happen: no further use, destruction of remaining material, and removal from future datasets.

Telling participants they may withdraw, without explaining these limits, promises something that cannot be delivered. The honest version is more complicated and it is the only one that is true.

In practice

A biobank's consent describes a scope, names the oversight committee and its composition including lay members, states that commercial use is permitted and that the participant may decline that specifically, states that material will not leave the country without further approval, and explains withdrawal accurately including what cannot be undone.

It also commits to an annual public summary of what the material has been used for.

Roughly 12 percent of participants decline commercial use while consenting to everything else. Under a bundled consent, all of them would have been recorded as agreeing.

That 12 percent is the argument for the separate option. It is not a marginal group; it is one participant in eight whose actual preference the simpler form would have misrepresented.

What this looks like when it goes wrong

A cohort study collects samples with a consent stating they may be used for future research into health and disease.

Fifteen years later the samples support a study of genetic associations with a behavioral trait, and the findings attract public attention as suggesting something about a group to which many participants belong.

The research was approved and the consent language covers it on any ordinary reading.

Participants are distressed. They agreed to research into health and disease, and understood that to mean research that would help treat illness. A study associating their group with a behavioral trait is not what they had in mind, and the harm falls on people who share the characteristic and never participated at all.

The consent was valid and the scope description was too broad to inform anyone, and no governance step asked whether this use was within what participants would have understood themselves to be agreeing to.

Key takeaways