How Do You Choose the Appropriate Time Horizon for an Economic Evaluation?

How Far Into the Future Should an Economic Evaluation Measure Costs and Outcomes?

A research team evaluates a digital educational intervention over one semester. The intervention requires substantial initial investment in software configuration and instructor training, but some resources may be reused in subsequent academic terms.

If the economic evaluation ends after one semester, the intervention may appear expensive because the initial investment is concentrated within a short period. If the evaluation extends over several years, those costs may be considered alongside recurring expenditure and benefits for later student cohorts.

However, extending the time horizon introduces uncertainty. Software may become obsolete, staff may require additional training, and future learning benefits may differ from those observed during the trial.

How should researchers decide how long an economic evaluation should extend? Should it follow the trial duration, the intervention's expected useful life, or the period over which all important consequences may occur?

Choose a Horizon Long Enough to Capture Important Differences Between Alternatives

Selecting an Economic Time Horizon That Matches the Intervention and Decision

What Is a Time Horizon in Economic Evaluation?

The time horizon is the period over which relevant costs and consequences are considered in an economic analysis.

For example, an economic evaluation may consider costs and outcomes over six months, five years, or the remaining lifetime of the affected population.

The appropriate duration depends on the intervention, comparator, and decision context.

A short horizon may be sufficient for an intervention whose costs and effects occur entirely within a brief period. A longer horizon may be required when consequences emerge or persist after the intervention ends.

The horizon should be justified by the economic question rather than selected solely because it is convenient for the research team.

How Is the Economic Time Horizon Different From Trial Follow-Up?

Trial follow-up is the period during which participants are observed and outcomes are collected under the research protocol.

The economic time horizon is the period over which the evaluation considers relevant costs and consequences.

These periods may coincide, but they do not have to.

For example, a clinical trial may observe outcomes for two years while a model-based economic evaluation estimates consequences over participants' remaining lifetimes.

Similarly, a semester-long educational trial may provide evidence for an economic model examining subsequent cohorts, provided the assumptions supporting future costs and benefits are defensible.

Researchers should distinguish observed evidence from modeled extrapolation.

What Factors Should Determine the Time Horizon?

Researchers should consider these factors together. A longer horizon is not automatically better if the intervention's consequences cannot be estimated credibly.

When Is a Short Time Horizon Appropriate?

A short horizon may be appropriate when the important differences between alternatives occur within a limited period.

For example, an intervention intended to improve immediate performance on a specific task may have a narrowly defined economic objective limited to the intervention period.

Similarly, a budget analysis may focus on expenditure within an institution's annual planning cycle.

However, researchers should not assume that short intervention duration means short-lived consequences.

An intervention delivered over several weeks may influence outcomes or resource use for years afterward.

When Is a Longer Time Horizon Necessary?

A longer horizon may be necessary when meaningful costs or consequences extend beyond the intervention period.

For example, a preventive healthcare intervention may affect disease outcomes and healthcare utilization over many years.

An educational intervention may require initial instructor training but be used with multiple student cohorts.

Longer horizons may capture both delayed benefits and recurring expenses.

Researchers should examine whether short-term trials miss important long-term economic consequences before restricting the analysis to observed follow-up.

When Is a Lifetime Horizon Appropriate?

A lifetime horizon may be appropriate when an intervention influences outcomes, survival, or resource use throughout the remaining life of the affected population.

It is common in some health economic evaluations, particularly where interventions affect mortality or long-term disease progression.

However, a lifetime horizon is not universally necessary.

For example, a short-term organizational training intervention may not require lifetime modeling when the relevant decision concerns operational costs and outcomes over a limited period.

Researchers should consult applicable economic evaluation guidance and justify why the selected horizon captures the consequential differences between alternatives.

Does the Time Horizon Depend on the Economic Perspective?

Yes. Different perspectives may emphasize consequences occurring over different periods.

An institution may be concerned with expenditure during its budget cycle, while a broader economic evaluation may consider costs and benefits extending beyond that period.

For example, a university may assess whether it can afford software licensing over the next three academic years. A broader evaluation may also consider longer-term consequences for students.

These questions should not be conflated.

The horizon should be selected in conjunction with the perspective, comparators, and intended decision.

How Should Initial and Recurring Costs Be Treated?

Initial costs may include equipment acquisition, software development, installation, or training.

Recurring costs may include subscriptions, maintenance, staff support, consumables, and replacement activities.

Researchers should identify when these resources are consumed and avoid assuming that initial expenditure is repeated every period.

Likewise, recurring costs should not be omitted merely because they occur after the original trial.

Where capital resources are used over multiple periods, their treatment should follow the selected economic costing conventions and the relevant decision context.

Can a Longer Horizon Change Which Intervention Appears Preferable?

Yes. Interventions may differ in the timing of their costs and benefits.

One intervention may require substantial initial expenditure but generate lower operating costs later. Another may have lower setup costs but higher recurring expenses.

Similarly, some interventions produce immediate benefits, while others may produce delayed effects.

Consequently, the preferred alternative may depend on the horizon selected.

Researchers should avoid choosing a horizon after examining which duration produces the most favorable result.

How Can Researchers Compare Alternative Time Horizons?

Where uncertainty about the appropriate duration is consequential, researchers may examine results under several justified horizons.

For example, a study might report costs over one year and explore longer-term scenarios over three or five years.

These comparisons can show how economic conclusions depend on assumptions about maintenance, intervention durability, and recurring expenditure.

However, alternative horizons should not be treated as interchangeable estimates of the same decision problem.

The primary horizon should be specified and justified, while supplementary scenarios should be clearly identified.

What Is Discounting in Economic Evaluation?

Discounting adjusts future costs and consequences to reflect their value at a common reference point, commonly the present.

Economic evaluations extending across multiple years may require discounting according to the relevant methodological guidance.

The appropriate discount rate varies by jurisdiction, evaluation framework, and sometimes the type of consequence.

Researchers should not assume that one rate applies universally.

For example, NICE guidance in the United Kingdom specifies discounting conventions for relevant health technology evaluations, but those conventions should not automatically be applied to educational interventions in other jurisdictions.

How Is Present Value Calculated?

A common discrete-time present-value formula discounts a future cost or consequence by the number of periods separating it from the reference point.

The 3% rate is used solely to demonstrate the calculation. Researchers should select rates according to applicable guidance and justify the choice.

Discounting should also be distinguished from adjusting monetary values for inflation. These operations address different economic concepts.

Should Both Costs and Outcomes Be Discounted?

Many economic evaluation frameworks require future costs and consequences to be discounted, although the prescribed rates and conventions may differ.

For example, health economic guidelines may specify whether the same rate should apply to costs and health outcomes.

Researchers should follow the methodological requirements relevant to their jurisdiction and decision context.

Where alternative rates could materially change results, sensitivity analyses may be appropriate.

Discounting assumptions should be reported transparently.

How Does Inflation Differ From Discounting?

Inflation adjustment expresses monetary values in a consistent price year, while discounting converts future values to a common reference time.

For example, costs observed in different calendar years may need adjustment to a common price level before economic comparison.

Future costs may then require discounting, depending on the analytical framework.

Researchers should distinguish nominal and real monetary values and ensure that discount rates are applied consistently.

Combining real costs with nominal discount rates without appropriate adjustment can produce misleading estimates.

How Can Economic Models Extend the Time Horizon?

Decision-analytic models may be used when relevant consequences extend beyond directly observed study data.

Depending on the research question, approaches may include decision trees, state-transition models, or other suitable modeling structures.

For example, a clinical intervention study may observe short-term outcomes while a model estimates later health events and costs.

In education, a model might examine the cost of maintaining a program across subsequent cohorts using assumptions about enrollment, licensing, and staff requirements.

However, modeling should not be used to invent unsupported long-term benefits. Its structure and parameters require defensible evidence.

What Assumptions Are Needed for Long-Term Extrapolation?

Extrapolation may require assumptions about the persistence of intervention effects, changes in resource use, population characteristics, and future implementation conditions.

For example, an educational intervention that improves immediate assessment performance may not necessarily improve long-term employment outcomes.

Researchers should not treat such downstream benefits as established without appropriate evidence.

Where assumptions are necessary, they should be explicitly stated and examined through sensitivity or scenario analyses.

The strength of economic conclusions depends partly on how well the available evidence supports the modeled future.

How Should Intervention Durability Be Considered?

Intervention durability concerns how long the effects or functions of an intervention persist.

For example, a one-time training program may produce skills that decline without reinforcement. A software intervention may require updates to remain functional.

Researchers should distinguish the duration of delivery from the duration of benefit.

Evidence about sustained effects may come from longitudinal studies, follow-up assessments, or relevant external research.

Where durability is uncertain, analyses should avoid assuming indefinite benefits.

Can Time Horizon Selection Affect Equity?

Yes. Different population groups may experience costs and benefits at different times.

For example, an intervention may impose immediate participant expenses while generating benefits that emerge later.

Individuals with limited financial resources may face barriers even when the intervention appears economically attractive over a longer horizon.

Researchers should distinguish long-term economic value from short-term affordability.

Where relevant, the distribution and timing of costs across stakeholders should be reported rather than concealed within a single aggregate estimate.

How Does Time Horizon Affect Productivity Costs?

Productivity consequences may occur during intervention delivery or continue afterward.

For example, a health intervention may initially require time away from employment but subsequently improve work capacity.

Researchers should consider whether productivity losses and time costs belong in the evaluation and over what period they should be measured.

Long-term productivity estimates may depend on assumptions about employment, wages, and the persistence of intervention effects.

These assumptions should be distinguished from directly observed changes.

Should the Economic Horizon Be Specified Prospectively?

Generally yes, when economic evaluation is an important study objective.

Prospective specification reduces ambiguity about which costs and outcomes belong in the analysis and discourages selecting a horizon because it produces favorable findings.

Researchers should consider planning economic evaluation prospectively alongside the intervention study.

However, legitimate changes may become necessary when new evidence emerges.

Material changes should be documented, justified, and distinguished from prespecified analyses.

What Happens When the Study Follow-Up Is Too Short?

When trial follow-up does not capture important economic consequences, researchers have several options.

They may extend follow-up, link suitable administrative records, use relevant external evidence, or develop an appropriate economic model.

Each option introduces different requirements and limitations.

Extended follow-up may increase participant burden, while modeling may increase dependence on assumptions.

Researchers should select the approach that best supports the intended economic inference rather than assume that more observation or more modeling is always preferable.

How Should Uncertainty About the Time Horizon Be Reported?

Reports should identify the selected horizon, its rationale, the duration of directly observed data, and any extrapolation beyond observation.

Researchers should describe assumptions about effect persistence, recurring costs, replacement, and discounting.

Where conclusions depend on the horizon, alternative scenarios should be presented.

The CHEERS 2022 reporting guidance provides relevant recommendations for describing economic evaluation methods and uncertainty.

Transparent reporting enables readers to distinguish evidence-supported conclusions from projections dependent on uncertain assumptions.

Choosing a Time Horizon for a University Digital Learning Intervention

The three-year total is larger because it includes more years of operation. This does not mean the intervention becomes less economically attractive. A longer evaluation would also need to consider the number of students served, the comparator's costs, and any benefits occurring during those years.

The three-year horizon may better reflect the university's intended decision, but it introduces assumptions not required for a one-semester descriptive cost analysis.

The researchers should therefore distinguish observed costs and outcomes from modeled projections and avoid claiming that future benefits are established simply because the intervention performed well during the initial study.

Common Misconceptions About Economic Evaluation Time Horizons

How to Select and Justify Your Economic Evaluation Time Horizon

Begin with the economic decision and identify when meaningful costs and consequences arise for each alternative.

Determine whether the study's observed follow-up captures those consequences or whether longer-term analysis is needed.

The selected horizon should be documented prospectively where feasible and interpreted in relation to the quality of the evidence supporting future consequences.

Check Whether Your Economic Evaluation Uses an Appropriate Time Horizon

Frequently Asked Questions About Economic Evaluation Time Horizons

The Appropriate Horizon Captures the Economic Consequences That Matter

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.