What Happens If Recruitment Is Much Lower Than Expected?

You Planned for 300 Participants. What If You Can Recruit Only 120?

Your sample-size calculation says you need 300 participants. The population appears large enough. An institution has agreed to provide access. Recruitment is scheduled for four months.

Then the invitations go out.

Far fewer people respond than expected. Some who respond are ineligible. Others consent but never complete the first assessment. After several weeks, the recruitment graph is nowhere near the trajectory in your research plan.

What happens now?

Low recruitment is not merely an administrative inconvenience. Depending on the research design, it can affect statistical precision, the composition of the sample, the feasibility of planned comparisons, the project timeline, costs, and ultimately whether the study can answer its research question. The important decision is not simply how to recruit more people. It is whether the study remains scientifically defensible with the participants you can realistically obtain.

Lower Recruitment Can Change What Your Study Is Capable of Answering

Low Recruitment Is a Scientific Problem When It Changes the Evidence You Can Produce

Recruitment plans often look deceptively simple: identify the population, invite eligible participants, obtain consent, and continue until the required sample has been reached.

Real recruitment contains several transitions. People must be reachable. Those reached must be potentially eligible. Eligible individuals must be willing to participate. Those who consent must complete whatever steps are required to become analyzable participants. In longitudinal research, enough of them must then remain in the study.

A weakness at any stage can reduce the final sample substantially.

This is why some research funders treat recruitment as an explicit feasibility issue rather than a logistical detail. For example, the U.S. National Institute of Mental Health requires recruitment milestones for covered clinical research and advises investigators to consider historical participation rates, source-population size and diversity, likely willingness to participate, participant availability, start-up delays, and the time needed for follow-up and analysis when developing recruitment timelines.

Start With the Recruitment Funnel, Not the Final Number

Suppose your target is 300 completed participants. That does not mean you need access to 300 people.

Imagine that 70% of the people you contact respond, 75% of respondents meet the eligibility criteria, 80% of eligible respondents consent, and 90% of those who consent provide usable data.

The expected proportion reaching the final stage is:

This is a planning illustration, not a universal recruitment formula. Actual probabilities differ substantially among populations, designs, recruitment channels, eligibility requirements, participant burdens, and research settings.

The useful principle is that the target sample sits at the end of a funnel. Recruitment feasibility should therefore be estimated using evidence about each important stage rather than assuming that the size of the underlying population equals the number available for analysis.

A Large Population Does Not Guarantee an Adequate Sample

A university may have 20,000 students, a hospital may treat thousands of patients, or an online community may contain hundreds of thousands of members. Those numbers can create false confidence.

Your accessible population is usually much smaller.

Eligibility criteria may remove most potential participants. Permission to contact them may be restricted. Recruitment may depend on particular courses, clinics, organizations, or gatekeepers. Only a fraction of those contacted may respond. Some may decline because participation requires too much time or because the topic is sensitive.

Current NIH clinical-trial planning requirements in some funding programs explicitly ask investigators to provide evidence that participating sites have access to sufficient numbers of eligible participants and to conduct pre-study assessments of their capacity to reach recruitment targets. The underlying lesson travels well beyond clinical trials: population size should not substitute for evidence of recruitability.

Low Recruitment Usually Reduces Statistical Precision

In quantitative studies, a smaller sample will often increase uncertainty around estimated effects, all else being equal.

Suppose a study was designed to estimate an association with enough precision to distinguish a substantively important relationship from a negligible one. If recruitment reaches only half the planned sample, confidence intervals may become wider and the study may be less capable of distinguishing among scientifically different possibilities.

For hypothesis-testing designs, reduced sample size can also reduce statistical power relative to the original plan.

The appropriate response is not simply to write “the study was underpowered” after data collection. Recalculate or otherwise reassess what the achievable sample means for the planned analyses while there is still time to make a decision.

Ask what effect sizes or levels of precision remain achievable with the sample you can realistically obtain.

The Problem Is Not Only Statistical Power

Under-recruitment is often discussed as though its only consequence were a smaller number in a power calculation. That is too narrow.

Low recruitment can also affect whether planned subgroups contain enough participants, whether multivariable analyses remain stable, whether rare outcomes are observed often enough, whether qualitative sampling captures sufficient diversity for the study's aims, and whether the project can support the comparisons or longitudinal follow-up originally proposed.

For some studies, the primary analysis may remain reasonable while secondary analyses become untenable. For others, the central question itself becomes difficult to answer.

Reassess each major research aim separately rather than assuming that one percentage reduction applies equally to everything in the protocol.

Who Is Missing May Matter as Much as How Many Are Missing

A study can reach a respectable sample size while recruiting a systematically restricted group of participants.

Suppose a study about university students recruits mainly highly engaged students because invitations are distributed through optional academic workshops. The numerical target may eventually be reached, but the resulting sample may differ meaningfully from the broader population to which the researcher hopes to generalize.

Recruitment should therefore be monitored not only for total numbers but, where relevant, for the characteristics necessary to address the scientific question.

This concern is reflected in formal recruitment policies. NIMH recruitment guidance, for example, considers the size and diversity of the source population and requires attention to relevant inclusion expectations in covered clinical research.

Increasing recruitment numbers through a different channel can also change the sample. A new recruitment strategy should therefore be evaluated for what it does to the study population, not merely for how quickly it produces participants.

Low Recruitment Can Reveal That an Assumption Was Wrong

Your proposal may have assumed that 40% of invited participants would respond. Perhaps that estimate came from another study, a previous project at the institution, or little more than optimism wearing a spreadsheet.

If the actual response rate is 8%, the issue is not merely that recruitment is “slow.” One of the assumptions underlying the study has failed.

That is why it is useful to identify the assumptions on which your research idea depends before recruitment begins. Recruitment rate, eligibility rate, consent rate, retention, and access to the source population can all be design-critical assumptions.

Once an assumption proves wrong, update the feasibility assessment using observed information rather than continuing to plan around the original estimate.

Find Where Recruitment Is Failing

“Recruitment is low” is a diagnosis about as informative as “something hurts.” You need to identify the stage producing the shortfall.

The appropriate intervention depends on where the funnel is leaking. Advertising more aggressively will not solve restrictive eligibility criteria. Relaxing eligibility criteria will not solve a burdensome consent process. Increasing incentives, where permitted, will not solve a population that your team cannot legally or practically contact.

Do Not Change Eligibility Criteria Merely to Rescue the Sample

When recruitment is poor, widening inclusion criteria can appear to be an obvious solution.

Sometimes it is reasonable. Criteria may have been unnecessarily restrictive, and broader eligibility may improve both feasibility and relevance. NIMH recruitment guidance explicitly encourages researchers to consider whether inclusion criteria are so narrow that finding eligible participants becomes unusually difficult.

But eligibility criteria define the population being studied. Changing them can therefore change the research question, intervention safety considerations, comparability with previous evidence, or interpretation of the findings.

Any modification should be scientifically justified and handled through the appropriate ethical, institutional, protocol, registration, or funding procedures that apply to the study. Do not quietly alter the sample definition because the original one proved inconvenient.

Extending Recruitment Has Costs

“We will just recruit for another three months” sounds simple until those three months collide with everything that follows.

Longer recruitment may delay follow-up, analysis, graduation, reporting, funding deadlines, intervention delivery, or staff contracts. Seasonal changes can also alter who is available or the context in which participants are recruited.

NIMH specifically advises investigators to leave enough time after recruitment for follow-up, data cleaning, analysis, and dissemination. Recruitment therefore cannot automatically expand until the end of the project period.

When considering an extension, revise the entire project timeline rather than moving one date.

Adding Recruitment Sites Can Help, but It Changes the Study

Opening another site or recruiting through additional institutions may increase the available population. It can also introduce differences in participants, procedures, implementation, personnel, local context, or measurement.

For some designs, those differences are manageable or even beneficial. For others, they introduce heterogeneity that must be considered in the design and analysis.

Additional sites also require time. Agreements, ethics review arrangements, staff training, data systems, and quality-control procedures may need to be established before recruitment begins.

Adding a site is therefore a research-design decision, not merely a mailing-list expansion.

Recruitment and Retention Are Different Problems

A study can recruit its full target and still lose much of its usable evidence through attrition. Conversely, excellent retention cannot compensate for a recruitment process that never enrolls enough participants.

Plan and monitor both. NIH application instructions for covered human-subjects research explicitly ask investigators to describe recruitment activities as well as strategies for retaining participants.

At Some Point, Recruitment Failure Becomes a Stop Decision

There is no universal percentage below the recruitment target at which every study should stop. The consequences depend on the design, question, available time, expected precision, participant characteristics, and realistic prospects for recovery.

But indefinitely extending an under-recruiting study is not automatically the ethical or scientifically responsible option.

Current NIH policies provide a useful illustration of the principle. Some institutes monitor actual enrollment against predetermined milestones so that recruitment problems can trigger early corrective action. NCCIH explicitly notes that when clinical studies cannot achieve sufficient accrual to address their aims, participants may have been exposed to research burdens or risks without generating the intended scientific benefit.

The broader lesson is straightforward: if continued recruitment is unlikely to produce a sample capable of answering the research question, the decision to continue deserves as much justification as the decision to stop.

When a Recruitment Target Starts Slipping Away

The key decision occurs before the project reaches the end of its recruitment period. Observed recruitment becomes new feasibility evidence, and the study is updated accordingly.

Common Mistakes When Recruitment Falls Behind

Decide What You Will Do Before Recruitment Becomes a Crisis

Recruitment planning should include more than a final target. Establish intermediate expectations and decide what would trigger review.

If the target is 300 participants over six months, for example, do not wait until month six to discover that only 90 have enrolled. Monitor the trajectory from the beginning. The exact milestones should reflect the study because recruitment is rarely perfectly linear. Start-up delays, academic calendars, clinic schedules, seasonal effects, and recruitment campaigns may create predictable variation.

The important principle is to compare actual progress with a realistic plan often enough that corrective action remains possible.

Use Observed Recruitment to Update Your Forecast

Once recruitment begins, you have information that was unavailable when the proposal was written.

Suppose you predicted 15 enrollments per week but observe an average of five after the start-up period. Continuing to forecast using 15 because that number appears in the proposal does not make the study more feasible.

Use the observed funnel to estimate what happens under realistic conditions. How many people can still be approached? What proportion are eligible? What proportion enroll? How long does enrollment take? What attrition has appeared so far?

The forecast should become progressively more evidence-based as the study proceeds.

Protect the Research Question When Modifying Recruitment

When a project is under pressure, almost any source of additional participants can start to look attractive.

Before changing recruitment channels, populations, sites, incentives, eligibility, or procedures, ask whether the modification changes who is being studied or what participation means. Then follow the applicable ethics, institutional, funding, registration, and protocol requirements before implementation.

The objective is not merely to reach the number printed in the sample-size section. It is to recruit a sample capable of answering the intended question.

Know When the Recruitment Problem Is Actually a Study Problem

Sometimes no recruitment strategy can solve the underlying issue.

The population may simply be too small. The eligibility criteria may be scientifically necessary but leave too few candidates. The intervention may impose a burden that participants reasonably decline. The timeline may be incompatible with the number required. The accessible sites may not contain the population assumed during planning.

At that point, ask whether infeasible recruitment has become the strongest argument against the proposed study.

If the answer is yes, protecting the original plan can become counterproductive. A different design may answer the question more efficiently, or the question itself may need to be narrowed. In some circumstances, the defensible decision is to stop investing in the current research idea rather than treating sunk effort as evidence that the project should continue.

Check Recruitment Feasibility Before and During the Study

Questions About Low Participant Recruitment

Do Not Wait Until Recruitment Ends to Discover That the Study Was Infeasible

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.