Cost-Consequence Analysis

Cost-consequence analysis is a form of economic evaluation that presents the costs and several outcomes of competing alternatives side by side, each in its own unit, rather than combining them into one measure. Every result needs a unit, preferred direction, time horizon and measure of uncertainty, and decision-makers interpret the evidence without adding unlike outcomes together or creating an unsupported overall score.

Cost-consequence analysis is commonly shortened to CCA. It is useful when a healthcare intervention affects several important health and non-health outcomes that cannot be combined credibly into one measure. By keeping the results separate, CCA makes the pattern of benefits, harms, costs and uncertainty visible.

What cost-consequence analysis shows

CCA presents the costs and multiple consequences of competing alternatives in their original or most meaningful units. It does not convert every outcome into money, combine all outcomes into QALYs or calculate one cost-effectiveness ratio.

The result is a structured evidence presentation rather than an automatic recommendation. Decision-makers must examine the pattern of costs, outcomes, uncertainty and distribution before reaching a conclusion.

When cost-consequence analysis is useful

CCA is particularly useful when an intervention produces several important outcomes that should remain visible. It can also support decisions involving non-health or cross-sector consequences that would be obscured by a single health-outcome measure.

CCA may be useful in the following situations.

• An intervention affects several distinct clinical outcomes.
• Benefits and harms should remain visible separately.
• Patient, caregiver, provider or system outcomes all matter.
• Important consequences use different measurement units.
• Non-health or cross-sector consequences are relevant.
• Stakeholders may value outcomes differently.
• A single summary measure would conceal important trade-offs.
• Transparent deliberation is preferable to an imposed weighting rule.

CCA's flexibility is also a limitation. Different decision-makers may reach different conclusions from the same evidence because the method does not supply a universal rule for combining the results.

Which choices define a cost-consequence analysis

A CCA is shaped by choices about the decision, outcomes, costs, evidence and presentation. These choices determine whether the evidence provides a balanced account or a selective picture of the alternatives.

1. Define the decision problem. Specify the population, setting, alternatives, comparator, perspective, time horizon and decision context.
2. Identify relevant consequences. Include the health and non-health outcomes needed for the decision.
3. Measure resource use and costs. Keep quantities and unit costs distinguishable where possible.
4. Measure each consequence. Preserve the natural or relevant unit for every outcome.
5. Calculate incremental results. Show the difference between each intervention and its comparator.
6. Align direction and timing. Make clear whether higher or lower values are favourable and ensure results use compatible time horizons.
7. Assess uncertainty. Report uncertainty separately for costs and each consequence.
8. Present the results clearly. Keep materially different outcomes disaggregated and traceable.
9. Support deliberation. Explain trade-offs without imposing an unjustified overall score or weighting rule.
10. Report limitations. Identify missing outcomes, measurement weaknesses and effects that remain uncertain.

Choosing which consequences to include

Consequences should be included because they are material to the healthcare decision, not merely because data are available. Selecting only outcomes with convenient or favourable evidence can distort the apparent balance between the alternatives.

Relevant domains may include the following.

• Mortality.
• Symptoms.
• Clinical events.
• Functional outcomes.
• Quality of life.
• Adverse events.
• Treatment burden.
• Service use.
• Patient experience.
• Caregiver effects.
• Productivity.
• Access.
• Equity.
• Environmental effects.
• Consequences outside healthcare.

The analysis should explain why each consequence matters. It should also identify important consequences that could not be measured.

The outcome set should avoid unnecessary duplication. Multiple measures of the same underlying effect may be useful, but their overlap must be disclosed.

Measuring resource use and costs

The analytical perspective determines which resources and costs belong in the CCA. A healthcare-system perspective may include treatment, administration, monitoring, adverse events and downstream healthcare. A broader societal perspective may also include patient time, caregiver time, travel or productivity effects.

For each resource, the cost is the quantity multiplied by the unit cost.

Resource Cost = Resource Quantity × Unit Cost

Total cost is the sum of these resource costs.

Total Cost = Σ(Resource Quantityᵢ × Unit Costᵢ)

Resource quantities and unit costs should be reported separately wherever possible. Each important cost input should identify the following.

• The resource being measured.
• The quantity and unit of measurement.
• The applicable population and setting.
• The source of the resource-use estimate.
• The unit cost and valuation source.
• The currency and price year.
• Any inflation or currency adjustment.
• The applicable analytical perspective.
• The time horizon.
• Any missing-data or extrapolation assumptions.
• The uncertainty assigned to the estimate.

Charges, reimbursement amounts and acquisition prices should not be treated as automatically equal to complete economic costs. Administration, monitoring, implementation, adverse-event and downstream resource use should be included when they differ materially between alternatives.

Building the cost-consequence results presentation

Each cost or consequence should be understandable without requiring the reader to reconstruct the analysis. Every result needs a clear name, unit, preferred direction, result for each alternative, incremental difference, uncertainty, time horizon and evidence source.

The following illustrative example uses hypothetical per-patient figures to compare a new programme with current care over an assumed one-year horizon; uncertainty ranges are illustrative 95% intervals.

Cost per patient

• Lower values are preferred.
• The new programme costs £1,250 per patient.
• Current care costs £1,100 per patient.
• The increment is £150 per patient.
• The uncertainty range is £60 to £245 per patient.
• The new programme is expected to cost more per patient.

Hospital admissions per patient

• Lower values are preferred.
• The new programme has 0.12 admissions per patient.
• Current care has 0.18 admissions per patient.
• The increment is −0.06 admissions per patient.
• The uncertainty range is −0.10 to −0.02 admissions per patient.
• The new programme is expected to reduce admissions.

Symptom-free days

• Higher values are preferred.
• The new programme gives 245 symptom-free days.
• Current care gives 220 symptom-free days.
• The increment is 25 days.
• The uncertainty range is 12 to 38 days.
• The new programme is expected to increase symptom-free days.

Quality-adjusted life years

• Higher values are preferred.
• The new programme produces 0.82 QALYs.
• Current care produces 0.79 QALYs.
• The increment is 0.03 QALYs.
• The uncertainty range is −0.01 to 0.07 QALYs.
• The direction of the QALY difference remains uncertain because the interval crosses zero.

Caregiver time

• Lower values are preferred.
• The new programme requires 32 hours.
• Current care requires 45 hours.
• The increment is −13 hours.
• The uncertainty range is −20 to −6 hours.
• The new programme is expected to reduce caregiver time.

Patient satisfaction

• Higher values are preferred.
• Satisfaction with the new programme is 84%.
• Satisfaction with current care is 72%.
• The increment is 12 percentage points.
• The uncertainty range is 5 to 19 percentage points.
• The new programme is expected to improve patient satisfaction.

The programme costs more but may reduce admissions, increase symptom-free days, reduce caregiver time and improve satisfaction. The QALY difference remains uncertain. CCA deliberately leaves these results disaggregated and does not automatically determine adoption.

Calculating and interpreting each increment

An increment shows how the intervention differs from its comparator. The calculation is the same for costs and outcomes, but the meaning of its sign depends on whether higher or lower values are preferred.

For every result, the increment is calculated in the same way.

Increment = Intervention Result − Comparator Result

For outcomes where higher values are desirable, a positive increment favours the intervention. For outcomes where lower values are desirable, such as costs, admissions or adverse events, a negative increment favours the intervention.

A result of zero indicates no estimated difference on that measure. Statistical or decision uncertainty should still be considered before interpreting a result as evidence that the alternatives are equivalent.

Why absolute and incremental results are both needed

Absolute results show the underlying level of cost or outcome for each alternative. Incremental results show the difference associated with moving from the comparator to the intervention.

Reporting only increments can conceal whether both options perform well or poorly. Reporting only absolute results makes direct comparison harder. A complete CCA should therefore report both.

Keeping units and time horizons clear

Each result should retain its original or decision-relevant unit. Costs, admissions, QALYs, percentages, caregiver hours and symptom-free days cannot be interpreted as though they measure the same thing.

Each result should state the following.

• Its measurement unit.
• Whether it is a total, mean, rate, proportion or percentage.
• Its time horizon.
• When the outcome occurs.
• Whether discounting was applied.
• Whether extrapolation was required.
• Whether the result is per patient, per episode or for a population.

Costs and consequences should be measured over a horizon long enough to capture material differences. Results with materially different horizons should not be placed beside one another without an explanation of the difference.

Preventing double counting

Two outcomes can provide different information while still reflecting part of the same underlying effect. CCA may display related outcomes separately, but they should not later be added together or described as independent benefits without checking their overlap.

Potential overlaps include the following pairs.

• Hospital admissions and inpatient days.
• Symptom improvement and quality of life.
• Adverse events and treatment discontinuation.
• Caregiver hours and the monetary value of informal care.
• Productivity losses and days absent from work.
• Mortality and life-years gained.

The analysis should explain why related outcomes are both included and how they should be interpreted. If a later analysis combines or values those outcomes, it should check explicitly for double counting.

Why unlike units cannot be added together

The results in a cost-consequence analysis do not share a common measurement scale. Adding pounds, admissions, QALYs, caregiver hours and percentages produces a number with no meaningful unit.

CCA therefore does not calculate a total consequence score. It keeps outcomes separate so the evidence and remaining value judgements stay visible.

If explicit weights are introduced, the analysis should identify the following.

• The weighting method.
• Whose preferences the weights represent.
• How the weights were elicited.
• Whether uncertainty in the weights was examined.
• Whether the resulting method should instead be described as multicriteria decision analysis or another structured approach.

An overall score should never be created merely by adding outcomes measured in unlike units.

Showing uncertainty for every important result

Each cost and consequence has its own evidence and uncertainty. A precise cost estimate does not make an uncertain health outcome more certain.

Uncertainty may be presented using several approaches.

• Confidence intervals.
• Credible intervals.
• Standard errors.
• Deterministic sensitivity analysis.
• Probabilistic sensitivity analysis.
• Scenario analysis.
• Plausible ranges.
• Alternative structural or methodological assumptions.

When an interval includes both favourable and unfavourable values, the direction of the result should be labelled uncertain. A favourable point estimate should not be presented as a certain benefit when its uncertainty includes no difference or possible harm.

Missing uncertainty should be identified rather than interpreted as certainty.

Distribution and subgroup effects

Average results may conceal differences between population groups. CCA can retain subgroup and distributional consequences explicitly because it does not require every outcome to be combined into one measure.

Relevant differences may be reported for groups defined by the following characteristics.

• Disease severity.
• Age or other clinically relevant characteristics.
• Socioeconomic position.
• Geography.
• Ethnicity or underserved status where appropriate and supported.
• Patient, caregiver, provider or payer group.
• Timing or generation.
• Access to services.

Subgroup findings should be based on a defensible decision question and credible evidence. CCA should not be used to display selectively chosen subgroups merely because their results appear favourable.

How CCA supports healthcare decisions

CCA helps decision-makers examine the complete pattern of costs, outcomes, trade-offs and uncertainty. Health technology assessment may use this evidence alongside clinical, budgetary, ethical, organisational, equity and implementation considerations.

CCA leaves judgement with decision-makers but makes the information requiring judgement more visible.

A CCA should not label an intervention cost effective or not cost effective unless a separate, justified decision rule supports that conclusion. It also does not determine whether adoption is affordable. Budget impact analysis may still be required to estimate financial consequences for a particular budget holder.

How CCA differs from other economic evaluations

Economic-evaluation methods differ in how they represent consequences and summarise results. CCA retains multiple outcomes separately, while other methods use a common health or monetary measure.

Cost-consequence analysis

• Multiple costs and outcomes remain separate in their relevant units.
• No combined summary measure is produced.
• The primary use is transparent examination of multiple consequences and trade-offs.

Cost-effectiveness analysis

• Outcomes are expressed using a shared natural health unit.
• Incremental cost per unit of outcome may be calculated.
• The primary use is comparison when alternatives affect a common measurable outcome.

Cost-utility analysis

• Health outcomes are represented using a preference-weighted measure, usually QALYs.
• Incremental cost per QALY or net benefit may be calculated.
• The primary use is comparison of health gains across diseases or interventions using a common outcome measure.

Cost-benefit analysis

• Costs and benefits are expressed in monetary terms.
• Net present value or a benefit-cost ratio may be calculated.
• The primary use is assessment of monetised net social value across programmes or sectors.

CCA may complement another evaluation when decision-makers need to see consequences that its primary summary measure does not capture.

Implementing the evidence presentation in Excel

A transparent workbook should use one row per cost or consequence and retain its original unit. The result name, unit, preferred direction, comparator value, intervention value, increment, uncertainty, time horizon, source and interpretation should be kept in separate fields.

Useful formulas include the following.

• Incremental cost is calculated with =InterventionCost-ComparatorCost
• An incremental consequence is calculated with =InterventionOutcome-ComparatorOutcome
• The preferred direction is labelled with =IF(Direction="Higher",IF(Increment>0,"Favours intervention",IF(Increment<0,"Favours comparator","No difference")),IF(Increment<0,"Favours intervention",IF(Increment>0,"Favours comparator","No difference")))
• An interval crossing zero is flagged with =IF(AND(LowerBound<=0,UpperBound>=0),"Direction uncertain","Direction consistent")
• A cost scenario is tested with =BaseCost*(1+ScenarioChange)
• A missing unit is caught with =IF(Unit="","Unit required","Pass")
• A missing time horizon is caught with =IF(TimeHorizon="","Time horizon required","Pass")
• A missing evidence source is caught with =IF(Source="","Source required","Pass")

A total score should not be calculated across outcomes measured in unlike units.

Interpretation safeguards

CCA requires several safeguards because its flexible presentation can otherwise create a selective or misleading picture.

• CCA exposes the information requiring judgement but leaves the judgement itself to decision-makers.
• Results in unlike units must not be added together.
• A favourable sign depends on whether higher or lower values are preferred.
• Absolute results should accompany increments.
• Outcomes with different time horizons require clear labelling.
• Related consequences may overlap and should not be double counted.
• A statistically uncertain result should not be presented as a certain benefit or harm.
• A CCA does not establish cost effectiveness without a separate decision rule.
• An overall weighted score requires an explicitly justified method.
• Selective outcome reporting can distort the apparent balance of consequences.
• Missing uncertainty should not be interpreted as certainty.
• A favourable pattern of per-patient results does not establish affordability.

CCA conclusions should remain tied to the decision problem, alternatives, population, comparator, perspective, time horizon, outcome set, evidence, uncertainty and deliberative process.

What should be reported

A CCA should report the decision problem, alternatives, comparator, population, setting, perspective, time horizon, costing year, currency, discounting, resource-use and unit-cost sources, outcome-selection rationale, units, preferred directions, absolute and incremental costs and consequences, uncertainty, evidence sources, missing outcomes, overlapping measures, subgroup or distributional effects, assumptions, limitations and the process by which decision-makers interpret the results.

Sources

• Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes, 4th edition. Oxford University Press. 2015.
• Mauskopf JA, Paul JE, Grant DM, Stergachis A. The role of cost-consequence analysis in healthcare decision-making. PharmacoEconomics. 1998;13(3):277-288.
• National Institute for Health and Care Excellence. Developing NICE guidelines: the manual (PMG20). Published 2014; updated 2025.
• Thokala P, Devlin N, Marsh K, Baltussen R, Boysen M, Kalo Z, et al. Multiple criteria decision analysis for health care decision making: an introduction. Report 1 of the ISPOR MCDA Emerging Good Practices Task Force. Value in Health. 2016;19(1):1-13.