How Much Recruitment Uncertainty Is Too Much Before a Study Begins?

You cannot eliminate recruitment uncertainty, so how much can you reasonably accept?

Before recruitment begins, you rarely know exactly how many people you will identify, reach, screen, find eligible, or enroll. Even well-planned studies encounter uncertainty because recruitment involves institutions, communication channels, eligibility criteria, human decisions, and time.

The presence of uncertainty is therefore not evidence that a study is infeasible.

The more important question is whether your study can tolerate being wrong about its recruitment assumptions. A project may be reasonably robust when several plausible recruitment outcomes still produce an adequate sample. Another may succeed only if nearly every assumption turns out favorably.

That difference matters. Recruitment uncertainty becomes concerning not simply when you lack perfect information, but when plausible departures from your assumptions could prevent the study from obtaining the evidence it needs.

Uncertainty becomes too much when the study depends on optimism

Recruitment risk depends on both uncertainty and its consequences

Some recruitment uncertainty is normal

Recruitment plans are forecasts. They depend on quantities that may not be known precisely before the study begins: how many people will encounter the invitation, how many will respond, how many will satisfy eligibility criteria, how many eligible individuals will consent, and how quickly enrollment will occur.

Even pilot and feasibility studies produce estimates with uncertainty. Preliminary recruitment performance should therefore inform planning without being mistaken for a guarantee of what will happen in the main study.

The objective is not to remove uncertainty. It is to determine whether the uncertainty threatens the study's ability to answer its research question.

Ask what happens if your assumption is wrong

Suppose two studies both expect 40% of eligible participants to enroll.

The first needs 100 participants and can realistically approach 1,000 eligible people. Even if enrollment falls to 20%, the available pool may remain sufficient.

The second also needs 100 participants but can approach only 260 eligible people. At the expected 40% enrollment fraction, approximately 104 would enroll. A relatively small shortfall would put the target out of reach.

The assumed rate is identical. The recruitment risk is not.

A highly uncertain assumption may be tolerable when the study has substantial margin. A moderately uncertain assumption may be dangerous when the study has almost none.

Look for stacked assumptions

Recruitment plans often appear plausible because assumptions are considered separately.

You may assume that a site will approve access, that 600 people can be reached, that half will qualify, that half of those will participate, and that recruitment can be completed within six months. None of those assumptions may appear extraordinary by itself.

But the final recruitment yield depends on all of them operating together.

When several uncertain assumptions all need favorable outcomes, the study can be much more fragile than any single estimate suggests.

Distinguish evidence from hope

Not every recruitment assumption deserves the same confidence.

An expected eligibility fraction derived from recent screening records at the actual study site is different from one borrowed from a study conducted in another country with different eligibility criteria. A recruitment rate observed using the same procedures is more informative than “the administrator thinks plenty of people will participate.”

When evaluating uncertainty, ask what supports each important assumption. Evidence may include site records, previous recruitment data, comparable studies, aggregate population information, or appropriately conducted feasibility work.

If eligibility itself is uncertain, it may be useful to estimate how many potential participants are likely to qualify. If the uncertainty concerns enrollment, examine whether eligible people are likely to agree to participate.

Use scenarios rather than one forecast

A single recruitment estimate can conceal how sensitive the study is to uncertainty. Scenario analysis makes that sensitivity visible.

Suppose you can realistically approach 500 eligible people and need 150 participants:

The useful question is not which percentage you prefer. It is which percentages are genuinely plausible and what the study would do under each scenario.

If only the optimistic scenario succeeds, recruitment uncertainty is consequential. If conservative scenarios still succeed, the study is more robust.

Calculate the recruitment pace the study requires

Population size alone does not tell you whether recruitment can finish on time. A study may have access to enough people eventually but still be unable to enroll them quickly enough.

This is particularly important for student projects and other time-limited studies. An enrollment target that might be attainable over three years may be entirely unrealistic within one semester.

Consider whether the uncertainty can be reduced before committing

Some uncertainties are relatively inexpensive to investigate. You may be able to obtain aggregate counts from a site, clarify recruitment procedures with a gatekeeper, examine previous response rates, or verify how many potentially eligible people pass through a service each month.

Other uncertainties may require more formal feasibility work. When recruitment is central to whether the study can proceed and existing evidence is weak, it may be appropriate to test the recruitment pathway before finalizing the study.

Recruitment-feasibility frameworks recommend specifying recruitment goals, tracking the process systematically, and evaluating whether those goals are met rather than simply reporting how many participants happened to enroll. Such tracking can also identify where potential participants are lost and which recruitment methods might need modification.

Prespecified progression criteria can help, but they should not be arbitrary

In pilot and feasibility research, investigators may specify progression criteria before collecting data. These criteria connect observed feasibility outcomes with decisions about whether to proceed, modify the study, or reconsider progression.

A traffic-light structure is often used:

The thresholds should correspond to what the future study actually requires. Methodological guidance on pilot trials recommends prespecifying feasibility parameters and meaningful goal and minimum thresholds rather than choosing cutoffs after seeing the results.

At the same time, missing a threshold need not always produce a mechanical stop decision. Recruitment tracking may reveal a modifiable reason for the shortfall. For example, investigators may identify excessive participant burden or a weak recruitment channel that can credibly be changed.

Ask whether you have a realistic recovery path

The same uncertainty is less dangerous when the study has viable corrective options.

If recruitment is slower than expected, can you extend recruitment without compromising the project? Can additional appropriate sites be added? Is there another legitimate channel to the same population? Can unnecessary participant burden be reduced? Are there scientifically defensible eligibility changes?

A backup plan counts only when it is operationally and scientifically plausible. “We will find another site” is not much protection if no alternative site has been identified and each new site requires months of approval.

Likewise, replacing the target population with whoever is easiest to recruit may solve the logistics while changing the study. Before treating population substitution as a contingency, consider whether a more accessible population would change the scientific question.

Some uncertainty deserves more caution in student research

A student may have less ability to absorb recruitment delays than a large funded project. Academic deadlines, graduation schedules, ethics-review cycles, semester calendars, and limited staffing can make recovery difficult.

Dependence on a single unconfirmed gatekeeper is especially consequential when losing that gatekeeper would eliminate the only viable participant pool. In such cases, consider whether the research question depends too heavily on access controlled by one powerful gatekeeper.

This does not mean student research should avoid every difficult population. It means the acceptable uncertainty should reflect how much time and flexibility the project actually has.

When a plausible recruitment plan becomes fragile after conservative assumptions are applied

Recruitment uncertainty is not just about whether your estimate is accurate

Ask whether your study survives a reasonably bad recruitment outcome

A useful feasibility assessment does not require you to predict recruitment perfectly. It asks whether the study remains viable when reality is somewhat less favorable than your preferred forecast.

What counts as “too much” therefore depends on fragility. A study with uncertain recruitment but substantial reserve capacity may be reasonable. A study with apparently precise estimates but no margin for error may be much riskier.

Before accepting recruitment uncertainty, stress-test the plan

Questions about recruitment uncertainty before a study begins

Recruitment uncertainty is too much when the study has no room to be wrong

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.