03 · What You Need to Know
Turn Uncertainty Into Answerable Feasibility Questions
Begin With the Uncertainties in the Future Study
A feasibility study should be designed backward from the decision that researchers will eventually need to make. Before choosing outcomes, identify what is genuinely uncertain about the proposed main study.
Suppose you intend to recruit participants from several universities. The existence of a large student population does not establish that recruitment is feasible. You may need to know how many students satisfy the eligibility criteria, how many can actually be approached, what proportion consent, and how quickly recruitment occurs.
If the main uncertainty concerns a complex intervention, the relevant questions may be different. Can it be delivered as planned? Do participants engage with it? Can staff implement it consistently? Does it fit within the intended setting?
There is therefore no universal set of feasibility outcomes that every study must collect. The study should investigate the uncertainties that could materially affect the future research.
Can You Recruit the Participants You Need?
Recruitment is one of the most common feasibility concerns because an otherwise excellent study cannot answer its research question if it cannot enroll an adequate sample.
Useful recruitment information may include the number of potentially eligible participants, the proportion who satisfy eligibility criteria, the proportion approached, the proportion consenting, recruitment per site or per unit of time, and reasons people decline when those data can appropriately be collected.
The most useful measure depends on the bottleneck. A low final recruitment rate could reflect a small eligible population, poor access to that population, low willingness to participate, restrictive eligibility criteria, or an ineffective recruitment strategy. Knowing only the number recruited may therefore conceal the actual problem.
When recruitment is a major uncertainty, investigate whether the proposed recruitment strategy can realistically produce the required sample before committing to the main study.
Will Participants Stay in the Study and Complete What Is Required?
Recruitment gets participants into a study. Retention and adherence determine whether many of them remain long enough to provide the information the design requires.
A feasibility study may therefore examine withdrawal, follow-up completion, missed assessments, adherence to intervention or study procedures, and the amount and pattern of missing data. These measures become particularly important in longitudinal research, repeated-measures designs, and studies that impose substantial participant burden.
A high recruitment rate can coexist with poor retention. If participants readily enroll but routinely fail to complete later assessments, increasing recruitment alone may not solve the problem.
Are the Study Procedures Acceptable to Participants and Other Stakeholders?
A procedure can be technically possible yet unacceptable in practice. Participants may regard an assessment as too intrusive, burdensome, confusing, or time-consuming. Staff may consider an intervention incompatible with ordinary workflow. Institutions may be unwilling to support particular procedures.
Acceptability can sometimes be assessed quantitatively, but qualitative evidence may explain why a procedure succeeds or fails. Interviews, observations, open-ended responses, or process evaluation can reveal problems that a completion percentage alone cannot explain.
For participant-facing procedures, it may also be necessary to establish whether participants understand what the study asks them to do. Apparent non-adherence may sometimes reflect unclear instructions rather than unwillingness to participate.
Can the Intervention or Procedure Be Delivered as Intended?
When a future study evaluates an intervention, feasibility may depend on whether the intervention can actually be implemented according to the protocol. Relevant questions might concern delivery, fidelity, dose, attendance, adherence, staff training, equipment, scheduling, or variation across sites.
This distinction matters because an intervention can be acceptable in principle while proving difficult to deliver consistently. Conversely, researchers may implement it exactly as planned while participants engage with it poorly.
A feasibility study should therefore distinguish problems with the intervention itself from problems with the research processes used to evaluate it.
Can You Collect the Data the Main Study Will Require?
A protocol may specify what variables will be collected without establishing that collecting them is practical.
Researchers may need to examine whether assessments can be completed within the intended time, whether instruments produce excessive missing data, whether measurements can be performed consistently, whether data from different systems can be linked correctly, or whether scheduled follow-up occurs at the intended intervals.
The question is not merely whether data can be collected at all. It is whether the planned procedures can produce sufficiently complete and usable data under realistic study conditions. Where this remains uncertain, examining whether the proposed data-collection procedures are practical can prevent serious problems later.
Are the Required Resources and Study Processes Realistic?
Feasibility can also depend on resources that are easy to underestimate during protocol development. These may include staff time, training requirements, equipment availability, space, site capacity, administrative processes, scheduling, data management, and the time required to complete individual procedures.
A process that works when a senior researcher personally supervises every case may not scale to a larger study. Likewise, a procedure that takes 20 minutes longer than anticipated may appear trivial during preliminary testing but become a major staffing problem when repeated hundreds of times.
The feasibility question should therefore consider not only whether something can be done, but whether it can be done under conditions sufficiently similar to those planned for the main study.
Feasibility Findings Should Lead to a Progression Decision
Collecting feasibility outcomes is only half of the task. Researchers also need to decide what those outcomes mean for the future study.
For some studies, this can be formalized through progression criteria specified before the feasibility data are collected. Criteria may define circumstances supporting progression, progression with modifications, additional feasibility work, or a decision not to proceed with the proposed design.
Feasibility outcome
The evidence collected about an uncertainty, such as recruitment rate, follow-up completion, intervention fidelity, or acceptability.
Progression criterion
A prespecified basis for interpreting feasibility evidence when deciding whether and how the research should move forward.
These criteria should be justified for the particular study. A recruitment rate acceptable for one design may be inadequate for another because sample requirements, recruitment periods, available sites, and resources differ.
Progression decisions also need not be reduced mechanically to one number. Quantitative results, uncertainty around estimates, qualitative evidence, safety considerations, and the feasibility of proposed modifications may all inform the final judgment.
A Feasibility Study Can Legitimately Tell You Not to Proceed
Researchers naturally hope that preliminary work will support the main study. That expectation can create a subtle bias toward interpreting almost any result as “feasible with minor modifications.”
But a well-designed feasibility study should be capable of producing an unfavorable answer. It may show that recruitment is insufficient, participant burden is unacceptable, the intervention cannot be delivered reliably, or required resources exceed what is available.
That does not make the feasibility study unsuccessful. It may instead prevent a larger study from consuming substantially more resources while remaining unable to answer its research question.
Watch Out
Do not define feasibility retrospectively so that whatever happened appears acceptable. Identify the important uncertainties and, where appropriate, progression criteria before seeing the results. Otherwise, the decision to proceed can become a post hoc justification rather than an evidence-informed judgment.
