The Gatekeeper Problem

Part of Consent and Vulnerable Populations: Cluster Trials and Simplified Consent.

About 5 minutes. Free, in your language, no account needed.

Why this matters

Every cluster trial has someone who says yes on behalf of the cluster, and that person's position is genuinely difficult.

They are asked to agree to something affecting people who have not been asked. They frequently have authority over those people in other respects, which is why they were approached. And the trial team needs their agreement to proceed, which puts them under a pressure nobody names.

Handling that relationship well is a large part of whether a cluster trial is ethically sound in practice rather than only on paper.

What a gatekeeper legitimately decides

Whether the cluster participates. Whether this clinic, school or village is a site.

Whether the intervention may be delivered there, where it is delivered to the setting.

Whether the organization's own data may be used, where the data belongs to the organization rather than to individuals.

Practical arrangements: access, scheduling, and how the research fits with the setting's work.

These are real decisions and they require someone with authority. A researcher cannot deliver an intervention in a clinic without the clinic agreeing.

What a gatekeeper cannot decide

Whether an individual participates in data collection about them.

Whether an individual receives an individually delivered intervention, where individual refusal is possible.

Whether an individual's personal data may be used, which is theirs.

Anything on behalf of a person who could be asked.

The line is between the setting and the person. A gatekeeper's authority is over the setting, and it does not reach the people in it as individuals.

The pressures on a gatekeeper

Worth naming, because they affect the quality of the agreement.

Benefit to the institution. Research brings funding, equipment, training and status. A clinic director agreeing may be agreeing partly for reasons unconnected to the research question.

Relationship with the researchers, who may be colleagues, funders or people the gatekeeper wants to keep on good terms with.

Asymmetry of information. The research team understands the protocol; the gatekeeper usually does not, and may not know what to ask.

No advocate for the people affected. The gatekeeper is being asked to protect the interests of people who are not in the room, without necessarily having the standing or the information to do it.

What good practice looks like

Be explicit about what is being asked. State plainly that the gatekeeper is agreeing to the cluster's participation and is not consenting for individuals, and that individual consent or objection arrangements apply separately.

Give them time and the ability to consult. A gatekeeper who can discuss the proposal with colleagues or with a representative body is in a better position than one asked to decide alone.

Make the individual-level arrangements visible to them. How people will be informed, how they may object, and what happens if many do.

Do not use the gatekeeper to recruit. A clinic director instructing staff to enroll patients converts their institutional authority into pressure on individuals, which is precisely the harm the gatekeeper distinction exists to prevent.

Accept refusal cleanly. A gatekeeper who declines has exercised their role properly, and a team that returns with a better offer has told them their decision was a negotiation.

In practice

A cluster trial approaches schools. The research team writes a gatekeeper information document distinct from the parent and child materials.

It states what the school is being asked to agree to, that the school is not consenting on behalf of any child, that parental consent and child assent will be sought separately, what the school's involvement requires in time and space, and what happens if few families consent.

It also states, explicitly, that the school will not be asked to encourage participation and that no information about which families declined will be shared with the school.

One head teacher raises something the team had not considered: children whose families decline will be visibly not participating during the sessions, which is a form of exposure. The design changes to run sessions in a way that does not separate participating and non-participating children.

That improvement came from treating the gatekeeper as a party with a perspective rather than as an access route.

What this looks like when it goes wrong

A cluster trial in workplaces obtains agreement from employers and delivers a health intervention at work.

Data is collected through questionnaires distributed by line managers, who are asked to ensure a good response rate.

Employees complete them. The response rate is excellent.

Employees are answering personal health questions, distributed by the person who manages them, in a setting where their employer has agreed to the research and their manager is measured on the response rate.

No employee was compelled. Every one of them was in a position where declining was visible to their manager and where the manager had an interest in their answer.

The employer's agreement was proper for what it covered. Using the management line to distribute and chase the questionnaires converted institutional agreement into individual pressure, which is exactly what the gatekeeper's limited authority is supposed to prevent.

Key takeaways