03 · What You Need to Know
Planning Economic Evaluation Before Intervention Research Begins
What Is Prospective Economic Evaluation Planning?
Prospective economic evaluation planning involves specifying the economic objectives, required information, and analytical approach before the relevant research data are collected.
It may be incorporated into a clinical trial, educational intervention study, implementation evaluation, or another appropriate research design.
For example, a university evaluating a new instructional support program may plan to measure learning outcomes alongside staff time, software expenditure, technical support, and the resources used by the comparison program.
The economic component may be integrated into the main protocol or described in a separate prespecified economic evaluation plan.
Prospective planning should be distinguished from prospective data collection. A prospectively planned evaluation may appropriately use existing records or modeling, while a study that collects some costs prospectively may still lack a coherent economic analysis plan.
Why Is Prospective Planning Important?
Economic evaluation requires information that may not be captured by the primary effectiveness study.
Relevant resource use may include staff preparation, equipment, training, service utilization, participant expenses, and subsequent costs influenced by the intervention.
If these quantities are not recorded, reconstructing them after the study may be difficult or impossible.
Prospective planning also encourages researchers to align the economic evaluation with the intervention comparison and intended decision.
However, planning does not eliminate uncertainty or guarantee high-quality economic evidence. The study must still collect suitable information and use methods consistent with its objectives.
What Should Be Specified Before the Study Begins?
A prospective economic evaluation plan should define the economic question and the methods necessary to address it.
| Planning element |
Question to resolve |
Why it matters |
| Economic objective |
What economic decision or comparison will the study inform? |
Determines the required evidence |
| Perspective |
Whose costs and consequences are relevant? |
Determines which resources belong in the analysis |
| Comparator |
Which alternative will the intervention be compared against? |
Establishes the incremental economic question |
| Outcome |
What consequence or benefit will be evaluated? |
Connects economic analysis to meaningful results |
| Time horizon |
How long should costs and consequences be considered? |
Determines whether delayed effects are included |
| Resource categories |
What goods, services, and time inputs should be measured? |
Guides data collection |
| Valuation |
How will resource quantities receive monetary values? |
Supports transparent costing |
| Missing data |
How will incomplete economic observations be handled? |
Supports defensible estimation |
| Uncertainty |
How will statistical and structural uncertainty be assessed? |
Supports appropriate interpretation |
| Analysis |
What economic estimates and comparisons will be reported? |
Reduces ambiguity and selective analytical decisions |
The level of detail should be proportionate to the economic question. A descriptive costing study may require a simpler plan than a model-based cost-utility analysis.
How Should the Economic Research Question Be Defined?
The economic question should identify the alternatives being compared, the relevant decision-makers, and the consequences that matter.
For example, a university may want to know whether a new digital learning intervention provides additional learning benefits at an acceptable incremental cost compared with existing instructional support.
This differs from asking how much the intervention costs to operate.
The former is a comparative economic question, while the latter may be answered through descriptive costing.
Researchers should avoid treating all questions involving money as equivalent.
Why Must the Comparator Be Specified Prospectively?
Economic evaluation concerns choices between alternatives. The relevant comparator is therefore fundamental to the analysis.
For example, a new educational intervention may be compared with existing teaching practice, another support program, or a different delivery model.
Each comparator may have different costs and outcomes.
Researchers should ensure that the economic comparison is consistent with the intervention evaluation and the decision being informed.
Changing comparators after examining results can alter the economic question and may introduce selective interpretation.
How Should the Economic Perspective Be Selected?
The perspective determines which costs and consequences are relevant.
For example, an institutional perspective may focus on university expenditure, while a broader perspective may include participant time, connectivity costs, and other resource consequences.
The appropriate perspective depends on the decision context and applicable methodological guidance.
Researchers should not assume that institutional financial records capture all relevant economic consequences.
Where participant costs matter, the protocol should specify whether participant expenses will be included alongside institutional costs.
What Economic Outcomes Should Be Selected?
Economic outcomes should correspond to the study's intended decision and methodological approach.
Cost-effectiveness analysis may express consequences in natural units, such as additional successful course completions or another appropriately defined outcome.
Cost-utility analysis commonly uses measures such as quality-adjusted life-years in health economic evaluation.
Cost-benefit analysis expresses relevant consequences in monetary terms under specified valuation assumptions.
Researchers should not assume that every educational or social intervention can be meaningfully evaluated using health-related utility measures.
Outcome selection should reflect the intervention's purpose and the evidence required by the decision-maker.
When Should Quality-of-Life Measures Be Included?
Quality-of-life measures may be relevant when an intervention is expected to influence health-related quality of life and a future economic evaluation requires such outcomes.
For example, a clinical intervention trial may need suitable health utility data to estimate quality-adjusted life-years.
However, collecting quality-of-life measures in every intervention study would create unnecessary burden when those outcomes are unrelated to the research question.
Researchers should consider whether quality-of-life data should be collected for future cost-effectiveness analysis before finalizing the outcome schedule.
Instrument selection should reflect the target population, intended economic framework, and available validity evidence.
How Should Resource-Use Data Collection Be Planned?
Researchers should identify the relevant resource categories, measurement periods, data sources, and responsible personnel.
Some information may be available from institutional financial systems, administrative records, or service utilization databases.
Other information, such as participant expenses or informal support, may require direct reporting.
Researchers should distinguish resource quantities from monetary valuations and establish consistent definitions across sites.
Detailed planning of resource-use data collected during intervention studies can reduce omissions and improve comparability.
Should Economic Data Be Collected at Every Research Visit?
Not necessarily. Economic measurement frequency should reflect how resource use occurs, the accuracy of available records, and the burden of repeated reporting.
Frequent measurement may reduce recall difficulties for some events but increase participant demands.
Longer recall periods may be feasible for memorable or well-documented expenditures but less suitable for frequent minor resource use.
Researchers should select assessment intervals based on the economic question and the characteristics of the data.
The objective is to obtain sufficiently complete and accurate information without unnecessarily expanding the participation schedule.
How Does Prospective Planning Affect the Time Horizon?
The time horizon should be long enough to capture the important differences in costs and consequences between alternatives.
A short intervention trial may observe initial implementation costs but miss later maintenance expenses or downstream benefits.
For example, software development and instructor training may create substantial first-year expenditure, while subsequent cohorts may use the intervention with different recurring costs.
Researchers should determine the appropriate economic time horizon before selecting data collection periods.
When the necessary horizon exceeds the study's follow-up, modeling may be considered with explicit assumptions and uncertainty.
Can an Economic Evaluation Extend Beyond the Trial?
Yes. Trial-based economic evaluation may be supplemented by decision-analytic modeling when important costs or consequences occur beyond the observed follow-up period.
For example, a short clinical trial may provide information about immediate outcomes while longer-term economic consequences are estimated using additional evidence and modeling assumptions.
However, modeling does not convert uncertain future effects into established facts.
Researchers should explain the evidence supporting extrapolation and evaluate uncertainty associated with model structure, parameters, and assumptions.
Prospective planning helps identify which trial observations are necessary to support later modeling.
Should an Economic Analysis Plan Be Prespecified?
Prespecifying the economic analysis can improve transparency and reduce opportunities for selective analytical decisions.
The plan may describe the perspective, outcomes, cost categories, time horizon, valuation methods, missing-data procedures, and uncertainty analyses.
It should also clarify how economic estimates relate to the main study's estimand and comparison design.
Not every analytical detail can be anticipated. Legitimate changes may become necessary when new information emerges.
Researchers should distinguish prespecified analyses from subsequent modifications and explain consequential departures from the original plan.
How Should Missing Economic Data Be Handled?
Economic data may be incomplete because participants miss follow-up assessments, institutions fail to record resource use, or certain expenses are difficult to observe.
Missingness may be related to participant characteristics, intervention exposure, or outcomes.
Researchers should establish procedures for monitoring data completeness and consider appropriate analytical methods.
Multiple imputation or other approaches may be suitable under defensible assumptions, but no statistical method automatically recovers unobserved resource use.
Sensitivity analyses may be needed when conclusions depend on uncertain missing-data assumptions.
Prospective planning allows researchers to collect auxiliary information that may support more credible missing-data analyses.
How Can Economic Evaluation Account for Uncertainty?
Economic evaluations involve uncertainty about costs, outcomes, model parameters, and sometimes the structure of the economic model.
Trial-based analyses may use appropriate statistical methods to estimate uncertainty in incremental costs and effects.
Model-based analyses may use deterministic or probabilistic sensitivity analyses to examine how conclusions depend on uncertain inputs.
Researchers should distinguish sampling uncertainty from uncertainty about assumptions or model structure.
A single cost-effectiveness ratio without an appropriate assessment of uncertainty may provide an incomplete basis for decision-making.
What Is an Incremental Cost-Effectiveness Ratio?
The incremental cost-effectiveness ratio, commonly abbreviated ICER, compares the difference in costs between alternatives with the difference in their effects.
This simplified calculation assumes that the outcome scale is suitable for the intended economic comparison and that the reported means represent appropriately estimated costs and effects.
It does not establish whether ₱60 per additional point represents acceptable value. That judgment requires a relevant decision context, interpretation of the outcome, uncertainty assessment, and consideration of alternatives.
ICERs also require particular care when incremental effects are close to zero or when an intervention is more costly and less effective than its comparator. Researchers should not interpret the ratio mechanically in those circumstances.
How Can Prospective Planning Improve Cost-Effectiveness Analysis?
Planning before data collection can help ensure that cost and outcome measurements refer to compatible populations, alternatives, and observation periods.
For example, if learning outcomes are measured at the end of a semester but intervention costs are recorded only during the first month, the resulting comparison may omit relevant delivery expenditure.
Prospective planning can also identify resource categories that would otherwise be missed and establish consistent procedures across study sites.
However, prospective planning cannot compensate for a biased effectiveness estimate or an inappropriate economic comparator.
Economic validity depends on the quality of both the cost and outcome evidence.
Can Economic Evaluation Be Conducted Retrospectively?
Yes. Retrospective economic evaluation may be appropriate when reliable cost, resource-use, and outcome data are available.
For example, linked administrative records may permit analysis of resource utilization and outcomes after an intervention has been implemented.
However, retrospective evaluation may be limited by missing resource categories, inconsistent definitions, incomplete follow-up, or confounding in nonrandomized comparisons.
Researchers should not assume that retrospective economic studies are inherently invalid. Their credibility depends on the data, design, assumptions, and intended inference.
When prospective collection is impractical, a separate economic evaluation rather than one embedded within a trial may sometimes be more appropriate.
Should Economic Evaluation Be Embedded in Every Randomized Trial?
No. An economic evaluation should be included when it addresses an important decision and the study can provide suitable evidence.
Some early efficacy trials may not have the duration, population, or implementation conditions needed to support meaningful economic conclusions.
Other trials may provide an excellent opportunity to collect resource-use and outcome data prospectively.
The decision should consider the intervention's development stage, relevance of economic consequences, available resources, and the extent to which trial conditions reflect intended practice.
Researchers should not add an economic evaluation merely because the main trial already collects effectiveness outcomes.
How Does Trial Design Affect Economic Evaluation?
The economic analysis must reflect the study's assignment mechanism, comparison groups, and outcome structure.
For example, cluster-randomized studies may require analytical methods that account for correlation within organizations.
Pragmatic trials may provide resource-use information under routine conditions, although the representativeness of participating settings still requires evaluation.
Explanatory trials may involve research-specific support that differs from ordinary implementation.
Researchers should distinguish the economic consequences of the intervention as studied from those expected under a proposed implementation model.
Can Prospective Economic Data Collection Increase Participant Burden?
Yes. Cost diaries, utilization questionnaires, time-use surveys, and additional follow-up assessments can impose demands on participants and providers.
Researchers should determine which data are essential and whether suitable administrative sources are available.
For example, institutional records may capture software expenditure, while participant questionnaires may be needed to assess personal connectivity expenses.
Researchers should examine how economic data collection may increase participant burden before expanding the measurement schedule.
Prospective planning should improve efficiency as well as completeness.
What Ethical and Data Governance Issues Should Be Considered?
Economic data may contain sensitive information about employment, income, healthcare utilization, household expenses, or other personal circumstances.
Researchers should collect only information necessary for the stated objectives and apply appropriate consent, confidentiality, security, and governance procedures.
Linking administrative records may require institutional permissions and other safeguards.
Researchers should also consider whether repeated financial questions create discomfort or discourage participation.
The requirements depend on the jurisdiction, institutional policies, and nature of the data.
What Guidance Should Researchers Consult?
The CHEERS 2022 statement provides reporting guidance for health economic evaluations, including the description of economic methods, assumptions, and uncertainty.
The Second Panel on Cost-Effectiveness in Health and Medicine provides methodological recommendations for cost-effectiveness analysis.
The International Society for Pharmacoeconomics and Outcomes Research has published good research practice guidance for economic evaluations conducted alongside clinical trials.
These sources are particularly relevant to health economic evaluation. Researchers working in education, social policy, or other fields should adapt applicable principles to their decision context and consult relevant disciplinary guidance.
Reporting standards should not be confused with mandatory methodological requirements for every research setting.
Watch Out
Do not wait until favorable intervention results are available before deciding what economic question to ask. Retrospective economic analysis can be legitimate, but missing resource information, inconsistent observation periods, and data-dependent analytical decisions may limit the credibility of conclusions.
