How Do You Manage Differences in Recruitment Capacity Across Research Sites?

What Should You Do When Some Research Sites Can Recruit More Participants Than Others?

You are planning a multisite study involving several universities, hospitals, schools, or community organizations. One institution has thousands of potentially eligible participants, while another serves a much smaller population. Some sites have experienced recruitment personnel and established communication systems; others must build their recruitment processes almost from scratch.

Should every site receive the same enrollment target? Should high-capacity institutions compensate when smaller sites fall behind? And what happens if one site contributes most of the participants?

These decisions affect more than recruitment timelines. Unequal enrollment can change the composition of the sample, the precision of statistical estimates, and the extent to which the study represents different institutional contexts.

Plan Recruitment According to Capacity and Scientific Requirements

Understanding and Managing Recruitment Capacity in Multisite Studies

What Does Recruitment Capacity Actually Mean?

Recruitment capacity is a site's practical ability to identify, approach, screen, and enroll eligible participants within a specified period while following the approved research protocol.

It is not simply the number of people registered at an institution or the size of its service population.

A university may have 10,000 students, but only a fraction may meet the study's eligibility criteria. Among those eligible, some may never receive an invitation, decline participation, or be unavailable during the recruitment period.

Similarly, a hospital may report treating hundreds of patients with a particular condition, but historical patient counts may include repeat visits, individuals who do not meet the study criteria, or patients who cannot reasonably participate.

Recruitment capacity therefore depends on several distinct factors:

• The number of potentially eligible individuals.
• The proportion meeting the complete eligibility criteria.
• The site's ability to identify and approach those individuals.
• The expected willingness and ability of eligible individuals to participate.
• The availability of trained personnel and operational resources.
• The recruitment period and competing institutional activities.

Researchers should assess these factors separately rather than treating an institution's population size as its expected enrollment.

Why Do Recruitment Capacities Differ Across Sites?

Some differences arise from the underlying population. A specialist hospital may encounter more eligible patients than a general hospital. A large university may offer more sections of a particular course than a smaller institution.

Other differences are operational. Sites may have different staffing arrangements, recruitment permissions, administrative processes, communication systems, and levels of research experience.

Institutional calendars can also affect recruitment. Examination periods, academic breaks, clinical service pressures, and organizational transitions may reduce the availability of eligible participants or research personnel.

Differences in participant trust, language, accessibility, and perceived burden may further influence enrollment.

These factors should not automatically be interpreted as evidence that one site is performing better than another. A site with a smaller eligible population may achieve excellent recruitment coverage while enrolling fewer participants overall.

How Should Recruitment Capacity Be Estimated Before the Study?

Recruitment planning should begin with site-specific feasibility information rather than identical targets assigned for administrative convenience.

Where possible, researchers should obtain historical records, screening logs, relevant institutional statistics, or pilot recruitment data.

For each site, estimate the number of potentially eligible individuals expected during the recruitment period, the proportion likely to satisfy eligibility criteria, the proportion that can realistically be approached, and the anticipated enrollment rate among those approached and eligible.

These estimates should reflect the actual protocol. A site's experience recruiting participants for a less demanding study may not accurately predict enrollment for a study requiring repeated assessments or substantial participant commitment.

Where reliable historical information is unavailable, researchers can develop a range of plausible scenarios rather than present a single optimistic estimate as certain.

How Can You Calculate a Preliminary Recruitment Projection?

A simple recruitment projection can help distinguish population availability from operational enrollment capacity.

This calculation is a planning approximation, not a validated universal recruitment model. Its usefulness depends on the definitions, assumptions, and quality of the underlying information.

The conditional probabilities must match the actual recruitment sequence. If the study screens people only after contact, for example, investigators may need to express the calculation using the probability of contact followed by eligibility among those screened.

Repeated contacts, overlapping recruitment lists, changes in eligibility, and recruitment interruptions can make simple projections inaccurate.

Researchers should update projections as actual screening and enrollment information becomes available.

Should Every Research Site Have the Same Recruitment Target?

Not necessarily. Equal recruitment targets may be useful when the study requires balanced site-specific comparisons, but they may be unrealistic when institutions have substantially different eligible populations.

Site-specific targets can reflect recruitment capacity while preserving the overall sample-size and analysis requirements.

For example, a study with four institutions might plan enrollment targets of 180, 140, 100, and 80 participants rather than requiring 125 from every site.

Whether this distribution is appropriate depends on the research question and statistical design.

If the primary objective is to estimate a participant-level association across the combined population, unequal contributions may be manageable with suitable design and analysis.

If the study aims to compare institutional outcomes or estimate site-level variation, very small site samples may limit precision and make the planned comparisons difficult.

The question of whether sample sizes should be balanced across research sites should therefore be resolved through statistical planning rather than an assumption that equality is always preferable.

What Happens When One Site Recruits Most of the Participants?

Unequal site contributions can influence the interpretation of pooled results.

Suppose one university contributes 70% of the participants in a four-university study. A participant-weighted analysis may largely reflect the characteristics and experiences of students from that institution.

This may be appropriate if the target estimand is defined accordingly. However, it may be inconsistent with an intended conclusion about an equally weighted set of institutional contexts.

Researchers should distinguish between estimating an average across participants and estimating an average across sites. These quantities need not be identical.

Site imbalance may also affect the precision of site-specific estimates and the estimation of between-site variation.

Appropriate statistical methods can account for some consequences of unequal site sizes, but they cannot automatically recover missing information about underrepresented institutions or populations.

How Should Recruitment Capacity Influence Site Selection?

Recruitment capacity is an important selection criterion, but it should not be the only one.

Sites should also be evaluated for their relevance to the target population, institutional characteristics, ability to implement the protocol, and contribution to the study's intended comparisons.

Excluding every low-capacity institution may make recruitment easier while systematically omitting smaller, rural, under-resourced, or specialized settings.

Conversely, including sites with negligible realistic enrollment may create substantial operational demands without providing enough information for the intended analyses.

The scientific benefits of adding research sites should be considered alongside enrollment feasibility.

Where contextual representation matters, investigators may deliberately include a lower-capacity site because it represents a setting that would otherwise be absent.

How Can Researchers Monitor Recruitment Performance Fairly?

Monitoring should distinguish absolute enrollment from performance relative to realistic opportunity.

A site enrolling 20 participants from 25 eligible individuals who were appropriately approached may be performing differently from a site enrolling 40 participants from 400 comparable opportunities.

Useful monitoring indicators include:

• Number of distinct potentially eligible individuals identified.
• Number approached or invited.
• Number screened and determined eligible.
• Number consenting and enrolled.
• Enrollment relative to the site's approved target.
• Enrollment over time compared with the planned recruitment trajectory.
• Documented reasons for nonparticipation, where ethically and operationally appropriate.
• Recruitment-related protocol deviations and operational interruptions.

Comparisons require consistent definitions. A site that counts every email sent as an invitation cannot be compared directly with a site counting only confirmed contacts without acknowledging the difference.

Monitoring should also account for staggered site activation. A newly activated institution should not be judged against another site that has been recruiting for several additional months.

What Should You Do When a Site Falls Behind Its Recruitment Target?

Begin with diagnosis rather than immediately increasing recruitment pressure.

A shortfall may arise because the eligible population was overestimated, screening criteria are more restrictive than anticipated, invitations are not reaching candidates, or participants consider the study too burdensome.

Operational problems such as inadequate staffing, delayed approvals, and competing responsibilities may also contribute.

Appropriate corrective actions depend on the cause. They might include improving approved communication procedures, extending recruitment within an authorized period, providing administrative support, or revising unrealistic site projections.

Changes should preserve the approved eligibility criteria, informed consent requirements, and protection against coercion.

Where recruitment methods must change, investigators should consider whether the adaptation affects selection mechanisms and obtain necessary approvals.

This requires attention to recruitment consistency across sites, even when local operational flexibility is permitted.

Should High-Capacity Sites Compensate for Underperforming Sites?

Sometimes, but redistribution should not be automatic.

If the primary objective is an overall participant-level estimate, additional recruitment at a high-capacity site may help achieve the planned total sample size, subject to the sampling design and analysis assumptions.

However, redistribution may be inappropriate when the study requires minimum site-specific samples, balanced institutional representation, or sufficient numbers of independent clusters.

Increasing enrollment at one existing site cannot necessarily compensate for the loss of another site, particularly when the number of sites is central to statistical power or contextual inference.

Before reallocating targets, researchers should revisit the estimand, planned comparisons, clustering assumptions, and intended population.

Any consequential changes to the recruitment or statistical analysis plan should be documented and approved where required.

When Should an Additional Research Site Be Activated?

Activating a reserve site may be appropriate when recruitment projections indicate that the existing sites are unlikely to achieve the necessary enrollment within the available timeframe.

However, additional sites introduce activation delays, training needs, coordination requirements, and potentially new contextual variation.

Researchers should consider whether the proposed site can contribute meaningful enrollment soon enough to justify these costs.

They should also examine whether the site's population and procedures are compatible with the study design.

Where reserve sites are anticipated, it is preferable to establish objective activation criteria during planning rather than wait until recruitment difficulties become severe.

Can Recruitment Differences Introduce Bias?

Yes. Unequal recruitment capacity can influence who enters the study, especially when sites differ in their underlying populations or recruitment mechanisms.

If high-capacity institutions recruit disproportionately from particular demographic or organizational groups, the combined sample may underrepresent other relevant populations.

Similarly, differences in participation rates may create selection bias when enrollment is associated with the exposures and outcomes being studied.

Researchers should distinguish between site-size imbalance and selection bias. Unequal sample sizes do not automatically create bias, but they can change the population represented by an estimate and interact with existing selection mechanisms.

Site characteristics and participant recruitment patterns may also contribute to heterogeneity in multisite findings.

How Should Recruitment Capacity Be Incorporated Into Statistical Planning?

Statistical planning should reflect the anticipated distribution of participants across sites, not merely the total number expected.

Depending on the design, investigators may need to consider unequal cluster sizes, intracluster correlation, site-level covariates, minimum site-specific sample sizes, and the number of sites available for analysis.

For cluster-randomized trials, unequal cluster sizes can reduce efficiency under common design assumptions. The consequences depend on the correlation structure, allocation, and analysis method.

For individually randomized multicenter studies, the implications differ because randomization may occur within sites and site adjustment may be incorporated into the analysis.

Researchers should therefore account for site effects during study design and examine whether realistic recruitment projections still support the intended precision and power.

When Should Recruitment Targets Be Revised?

Targets may need revision when new information reveals that the original assumptions were inaccurate or when circumstances materially change.

Revisions should be based on documented recruitment evidence and the study's scientific requirements, not simply on a desire to produce favorable results.

Investigators should assess whether the revised targets affect the target population, statistical power, planned comparisons, or ethical and operational arrangements.

When applicable, the protocol, statistical analysis plan, trial registration, funding arrangements, and institutional approvals may require updating.

Recruitment changes should be distinguished from outcome-driven design changes. Decisions informed by unblinded outcome patterns can introduce additional risks of bias and may require independent oversight or specialized adaptive-design procedures.

Managing Unequal Recruitment Across Four Universities

At the scheduled review, the study has enrolled 420 participants against the cumulative planned target of 500. The remaining gap is 80 participants.

The example illustrates why recruitment management requires more than transferring enrollment targets from a slower site to a faster one. A redistribution that improves total enrollment may simultaneously reduce the study's ability to examine particular institutional conditions.

Common Mistakes When Research Sites Recruit at Different Rates

How to Manage Unequal Recruitment Without Compromising the Study

Recruitment management should begin before site activation. Establish realistic projections, identify the minimum contributions required for the intended analyses, and define how enrollment will be monitored.

When recruitment differs from expectations, distinguish operational problems from limitations in the eligible population. Corrective action should address the actual cause while preserving the study's scientific purpose.

A well-managed multisite study does not require identical enrollment performance. It requires recruitment decisions that remain consistent with the research question, sampling strategy, ethical requirements, and statistical analysis plan.

Before Finalizing Multisite Recruitment Targets

Frequently Asked Questions About Recruitment Capacity Across Sites

Manage Recruitment According to What Each Site Can Contribute Scientifically

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.