Feasibility Assessment

How a feasibility assessment tests whether a plan can work

A feasibility assessment examines whether a proposed study, intervention, service or implementation plan can be delivered under the conditions that are expected to apply. This page explains the questions a feasibility assessment should address, the evidence it uses, and how its findings support a decision to proceed, modify, test further or stop.

Feasibility is not the same as desirability or effectiveness. A proposal may be clinically valuable but impossible to deliver with available staff, infrastructure, time or funding, while an easily delivered proposal may provide too little benefit to justify implementation.

Defining what is being assessed

The assessment should begin with a clear description of the proposal, intended population, setting, delivery model and decision that will follow. Without a defined plan, feasibility becomes a vague judgement about whether an idea seems possible.

The specification should identify:

• The study, intervention, service or programme being proposed.
• The population expected to participate or receive the intervention.
• The organisations and locations involved.
• The activities, technology and workforce required.
• The expected scale and duration.
• The outcomes or outputs the proposal is intended to produce.
• The constraints that cannot be changed.
• The aspects of the proposal that can be adapted.
• The criteria that will determine whether the proposal proceeds.

Feasibility should be assessed for the version of the proposal that could actually be implemented. Evidence about an idealised service may not apply to a reduced or locally adapted version.

The main dimensions of feasibility

Feasibility is multidimensional because a plan can be workable in one respect and impractical in another. The dimensions should be tailored to the proposal rather than treated as a checklist that must always contain the same items.

Common dimensions include:

• Technical feasibility: The required technology, equipment, data and specialist capabilities exist and can perform as intended.
• Operational feasibility: The proposal can fit into workflows, referral pathways, governance arrangements and routine service delivery.
• Workforce feasibility: Sufficient staff with the required skills and time can be recruited, trained and retained.
• Population feasibility: Enough eligible participants or patients can be identified, reached and retained.
• Financial feasibility: Funding is available for implementation, operation, maintenance and evaluation.
• Time feasibility: The work can be completed within the decision, funding or research timetable.
• Acceptability: Patients, staff, organisations and other affected groups are willing to participate or change practice.
• Ethical and legal feasibility: Consent, privacy, safety, procurement and regulatory requirements can be satisfied.
• Data feasibility: Required data can be collected with sufficient completeness, validity and timeliness.
• Scalability: Delivery can expand beyond an initial site without losing essential quality or creating unaffordable resource demands.

A single critical failure may make the plan infeasible even when most other dimensions appear favourable. The assessment should therefore identify essential conditions separately from features that are merely desirable.

Feasibility of a research study

A study feasibility assessment examines whether the proposed design can recruit, deliver, measure and retain enough participants to answer the research question. It should test the assumptions on which the full study depends rather than attempt to estimate treatment effectiveness prematurely.

Important questions include:

• Can eligible participants be identified accurately?
• Are enough participants available within the recruitment period?
• Will patients and clinicians accept recruitment and allocation procedures?
• Can the intervention and comparator be delivered consistently?
• Can contamination between study groups be controlled?
• Can required outcomes be measured at the planned times?
• Are follow-up and retention rates likely to be adequate?
• Can sites complete the data and governance requirements?
• Can the study team operate the randomisation, data and safety processes?
• Are the time and budget assumptions realistic?

These questions should be linked to explicit progression criteria. A study should not be declared feasible solely because some participants were recruited.

Feasibility of an intervention or service

An intervention feasibility assessment examines whether the proposed activity can be delivered to the intended population within a real service environment. It should consider both the intervention itself and the system changes required to support it.

The assessment may examine:

• Demand and eligible population size.
• Referral and patient-identification processes.
• Physical space, equipment and digital infrastructure.
• Workforce roles, training and supervision.
• Procurement and supply chains.
• Compatibility with existing clinical pathways.
• Patient travel, time and access requirements.
• Implementation support and change management.
• Ongoing maintenance and quality assurance.
• Effects on other services and capacity.

Delivering an intervention once under close project supervision does not establish that it can operate routinely. The assessment should distinguish temporary project resources from resources that would be available after implementation.

Estimating recruitment feasibility

Recruitment feasibility depends on the number of potentially eligible people, the proportion who meet the final criteria, the proportion approached and the proportion who consent. Optimistic assumptions can make an otherwise sound study impossible to complete.

Expected recruitment can be represented as:

$$ N_R = N_P \times p_E \times p_A \times p_C $$

where:

• (N_R) is the expected number recruited.
• (N_P) is the number of potentially eligible people.
• (p_E) is the proportion meeting the eligibility criteria.
• (p_A) is the proportion who can be approached.
• (p_C) is the proportion who consent.

If 1,000 people are potentially eligible, 60% meet the final criteria, 80% can be approached and 50% consent, expected recruitment is:

$$ N_R = 1{,}000 \times 0.60 \times 0.80 \times 0.50 = 240 $$

Each input should be supported by local data, prior studies or a justified range. The calculation should also allow for site activation delays, competing studies, seasonal variation and recruitment that changes over time.

Assessing retention and follow-up

Recruitment alone is not sufficient when the decision requires complete follow-up. Attrition reduces the amount of analysable information and may introduce bias when withdrawal is related to treatment, health or participant burden.

Expected completed follow-up can be estimated as:

$$ N_F = N_R \times (1 - p_L) $$

where:

• (N_F) is the expected number with the required follow-up.
• (N_R) is the number recruited.
• (p_L) is the expected proportion lost to follow-up.

The assessment should investigate why participants might withdraw, which outcomes are most vulnerable to missingness and whether retention procedures are acceptable. Increasing the initial recruitment target does not correct bias caused by differential loss to follow-up.

Testing data feasibility

Data feasibility concerns whether the information needed for delivery, monitoring or evaluation can be obtained with adequate quality. A field appearing in a database does not prove that it is complete, valid or available at the required time.

The assessment should examine:

• Data definitions and coding consistency.
• Coverage of the intended population.
• Completeness of required fields.
• Timing and frequency of data availability.
• Linkage between relevant systems.
• Legal authority and consent for data use.
• Burden of collecting new information.
• Ability to distinguish intervention and comparator pathways.
• Availability of outcomes, costs and important confounders.
• Processes for validation, correction and audit.

A small sample extraction or data-quality study may be necessary before committing to a full evaluation. Missing essential data should be treated as a feasibility finding rather than hidden within the later analytical plan.

Workforce and organisational capacity

Workforce feasibility depends on more than the total number of staff. The proposal may require particular competencies, protected time, clinical supervision, training or coverage outside normal hours.

Questions should include:

• Which tasks are new and who will perform them?
• Do staff have the authority and skills required?
• How much time will each task require?
• Can existing duties be reduced or reassigned?
• Is training available and can competence be maintained?
• What happens during absence, turnover or peak demand?
• Which organisation owns each responsibility?
• Are leaders and frontline staff willing to support the change?

Using staff time for the proposal creates an opportunity cost even when no additional staff are hired. The assessment should identify which existing activities may be displaced.

Financial and economic feasibility

Financial feasibility tests whether the proposal can be funded and sustained. It should include the full resource requirement rather than only the purchase price or initial project grant.

Relevant costs may include:

• Planning and project management.
• Equipment, software and licensing.
• Recruitment and training.
• Clinical and administrative staff time.
• Data collection, monitoring and evaluation.
• Facilities and infrastructure.
• Maintenance and technical support.
• Patient travel or participation support.
• Governance, contracting and regulatory work.
• Replacement, scaling and long-term operation.

Budget feasibility asks whether the responsible organisation can meet these costs at the required times. Economic value asks whether the benefits justify the resources used. A proposal can be cost effective but unaffordable, or affordable but poor value.

Acceptability and behavioural feasibility

A technically workable intervention may fail when patients, clinicians or organisations are unwilling to use it. Acceptability should be examined through direct engagement rather than inferred from expert opinion alone.

Relevant issues include:

• Perceived benefit and burden.
• Trust and privacy concerns.
• Compatibility with patient preferences and cultural expectations.
• Effects on professional roles and autonomy.
• Ease of use and training needs.
• Incentives and unintended behavioural responses.
• Stigma or fear associated with participation.
• Equity of access and participation.

Stated willingness does not always predict actual behaviour. Observation, usability testing or a limited pilot may provide stronger evidence when implementation depends on repeated action.

Feasibility studies and pilot studies

A feasibility study asks whether and how the proposed future work can be done. A pilot study is a small-scale version of the planned study or intervention used to test parts of the full process.

The terms overlap but are not interchangeable. A feasibility assessment can use interviews, record review, simulation, workflow mapping or data analysis without running a pilot. A pilot should have explicit feasibility objectives and should not be interpreted as a definitive effectiveness study merely because outcome data were collected.

Setting progression criteria

Progression criteria translate feasibility findings into a decision. They should be agreed before the results are known and should correspond to the assumptions that could make the proposal succeed or fail.

Criteria may address:

• Recruitment per site or month.
• Retention and completeness of follow-up.
• Intervention delivery or adherence.
• Data completeness.
• Safety or technical performance.
• Acceptability to patients and staff.
• Cost per participant or site.
• Site activation time.
• Availability of required workforce or infrastructure.

Traffic-light categories such as proceed, amend and stop can support decisions, but thresholds should not replace judgement. A result near a threshold may require consideration of uncertainty, causes and feasible modifications.

Managing uncertainty in feasibility estimates

Feasibility evidence is often based on small samples, incomplete data and assumptions about future behaviour. Point estimates can therefore create false confidence.

The assessment should use ranges or scenarios for uncertain inputs such as recruitment, uptake, staffing time, cost and retention. It should identify which assumptions most affect the conclusion and which can be tested before a larger commitment.

Uncertainty is not itself evidence that a plan is infeasible. The decision depends on whether uncertainty can be reduced at reasonable cost and whether the organisation can tolerate the consequences if the assumptions prove wrong.

Equity and feasibility

A proposal may appear feasible because it works for the easiest population to reach while excluding people with greater needs or barriers. Recruitment, access and delivery should therefore be examined across relevant population groups.

The assessment should consider:

• Language and communication needs.
• Disability access.
• Digital access and literacy.
• Travel and time requirements.
• Direct and indirect participation costs.
• Trust and previous experience of services or research.
• Differential eligibility, uptake and retention.

Adaptations that improve inclusion may require additional resources. Those requirements are part of feasibility rather than optional additions after the main design is complete.

A simplified example

Suppose a health system proposes a remote-monitoring programme for 2,000 patients with heart failure. The technology is available, but the assessment finds that nurses would need 25 minutes each week to review alerts for every enrolled patient.

At full enrolment, weekly review time would be:

$$ 2{,}000 \times 25\ \text{minutes} = 50{,}000\ \text{minutes} $$

$$ 50{,}000 \div 60 = 833.3\ \text{hours per week} $$

The proposal is technically feasible but not operationally feasible with the planned workforce. The programme could be redesigned by improving alert targeting, narrowing eligibility, changing review frequency or adding staff, after which the revised model should be reassessed.

Common misunderstandings

Feasibility is not a general impression that a proposal looks reasonable. It is an evidence-based assessment of whether defined requirements can be met under specified conditions.

Common misunderstandings include:

• A successful small demonstration does not prove that a programme can scale.
• Available technology does not establish workforce or operational feasibility.
• Recruitment of some participants does not establish that the study can reach its target.
• A temporary grant does not establish long-term financial sustainability.
• Cost effectiveness does not prove affordability or implementability.
• Acceptability to senior leaders does not establish acceptability to patients or frontline staff.
• Collecting outcome data in a pilot does not make it a definitive effectiveness study.
• An infeasible version of a proposal does not mean that every modified version is infeasible.

Interpreting the assessment

A feasibility assessment should end with a clear decision and the evidence supporting it. Appropriate conclusions include proceeding as planned, proceeding with specified modifications, conducting further feasibility work or stopping the proposal.

The assessment is most useful when it identifies the limiting conditions, the cost of addressing them and the uncertainty that remains. Its purpose is not to guarantee success but to prevent avoidable failure and improve the design before larger resources or participants are committed.