Incremental Effectiveness

This page explains how the difference in health outcomes between an intervention and its comparator is calculated and interpreted. It covers outcome units, comparison direction, timing, uncertainty and the role of the result in cost-effectiveness analysis.

What the result tells us

Incremental effectiveness shows how much more or less health outcome one alternative produces than another. It is a comparative result rather than the total outcome produced by either alternative.

• Incremental effectiveness compares an intervention with a stated comparator.
• Incremental effectiveness can be expressed in QALYs, life-years, cases prevented, events avoided or another defined outcome unit.
• Incremental effectiveness does not include cost and does not by itself determine whether an intervention is cost-effective.

How the difference is calculated

The calculation subtracts the comparator outcome from the intervention outcome. The alternatives, outcome unit, population and time horizon must be consistent for the subtraction to be meaningful.

Incremental effectiveness = Intervention outcome − Comparator outcome

For example, suppose an intervention produces 4.80 QALYs per person and its comparator produces 4.50 QALYs per person:

Incremental effectiveness = 4.80 − 4.50 = 0.30 QALYs per person

The intervention is therefore estimated to produce 0.30 additional QALYs per person relative to the stated comparator. This is an expected difference across the evaluated population; it does not mean that every person gains exactly 0.30 QALYs.

Why the outcome unit matters

An incremental result cannot be interpreted without knowing what was measured. A gain of 0.30 QALYs, 0.30 life-years and 0.30 clinical-response units describes three different types of outcome.

• A QALY combines length of life and health-related quality of life.
• A life-year measures survival without adjusting for health-related quality of life.
• A natural clinical unit may measure events avoided, cases detected, symptom-free days or another condition-specific outcome.
• Outcomes expressed in different units cannot be combined in one incremental ratio without a valid conversion or common measure.

Why the direction of the outcome matters

A positive numerical difference is favourable only when a higher outcome value represents improvement. Some outcomes, including deaths, adverse events and hospital admissions, are normally better when their values are lower.

For a harmful outcome, analysts can report a reduction directly or redefine the measure as events avoided. The analysis must state the convention clearly so that readers and machines do not reverse the interpretation.

• Positive incremental QALYs normally indicate a health gain.
• Negative incremental deaths may indicate fewer deaths and therefore a favourable result.
• Reporting events avoided reverses the direction used for the original harmful-event count.
• Switching the intervention and comparator reverses the sign of incremental effectiveness.

How timing and modelling affect the result

Incremental effectiveness depends on the period over which outcomes are measured or modelled. Trial follow-up may capture only part of the eventual difference when benefits, harms, recurrence or survival continue beyond the observed evidence.

Time horizon, extrapolation, treatment-effect duration and discounting can therefore change the estimated result. These choices must be justified and tested rather than selected because they produce a preferred conclusion.

• Time horizon determines which health consequences enter the calculation.
• Extrapolation estimates outcomes occurring beyond the observed evidence.
• Treatment-effect assumptions determine how long a difference between alternatives is maintained.
• Discounting changes the present value of health outcomes occurring at different times when required by the applicable reference case.

How the result enters cost-effectiveness analysis

Incremental effectiveness becomes the outcome denominator in an incremental cost-effectiveness ratio when ratio interpretation is appropriate. It also contributes to incremental net monetary benefit and other net-benefit calculations.

ICER = Incremental cost ÷ Incremental effectiveness

Incremental net monetary benefit = Threshold × Incremental effectiveness − Incremental cost

Incremental cost and incremental effectiveness must compare the same alternatives in the same direction. Their signs must be examined together before an ICER is interpreted.

An ICER is undefined when incremental effectiveness equals zero. A negative ICER is also ambiguous when reported alone because it can indicate either that the intervention is dominant or that it is dominated.

How to interpret positive, negative and zero results

The sign of incremental effectiveness describes the direction of the outcome difference under the stated convention. Interpretation must also consider the outcome measure, comparator, uncertainty and whether higher or lower values are desirable.

• Positive incremental effectiveness is favourable when higher outcome values represent better health.
• Negative incremental effectiveness is unfavourable when higher outcome values represent better health.
• Zero incremental effectiveness means that no outcome difference is estimated at the reported precision.
• A small non-zero difference may be uncertain or clinically unimportant even when it changes an economic calculation.

Why statistical significance is not the whole answer

An estimated outcome difference has uncertainty because it is based on evidence, assumptions or both. Statistical significance, clinical importance and economic relevance answer different questions and should not be treated as interchangeable.

A difference can be statistically uncertain but still affect expected cost-effectiveness, particularly when uncertainty is represented probabilistically. Conversely, a statistically significant difference may be too small to matter clinically or economically.

The point estimate should therefore be reported with an appropriate measure of uncertainty and interpreted alongside the clinical evidence and economic decision framework.

Worked interpretation example

Suppose an intervention costs £2,500 more per person and produces 0.10 additional QALYs relative to current care:

ICER = £2,500 ÷ 0.10 QALYs = £25,000 per QALY gained

The incremental effectiveness remains the separate estimate of 0.10 additional QALYs per person. The ICER combines that outcome difference with the corresponding incremental cost.

The economic conclusion depends on the decision threshold and uncertainty, while the clinical interpretation depends on what produced the QALY difference. Affordability, equity, feasibility and implementation remain wider decision considerations.

Common mistakes and safeguards

The calculation is straightforward, but missing context can reverse the interpretation or make the comparison invalid. Consistent alternatives, outcome definitions and analytical assumptions are essential.

• Do not subtract the intervention and comparator in inconsistent directions.
• Do not combine outcomes measured in different units.
• Do not treat a positive number as beneficial without checking the outcome direction.
• Do not treat a point estimate as certain.
• Do not calculate an ICER when incremental effectiveness equals zero.
• Do not interpret a negative ICER without examining the signs of both incremental cost and incremental effectiveness.
• Do not treat incremental effectiveness as evidence of cost-effectiveness without considering incremental cost and the applicable decision rule.

Further learning

The following resources explain how incremental outcomes are measured, modelled and combined with costs:

• Methods for the Economic Evaluation of Health Care Programmes, Drummond and colleagues — LIB-000001
• NICE Health Technology Evaluations: The Manual — LIB-000099
• Guidelines for the Economic Evaluation of Health Technologies: Canada, 4th Edition — LIB-000117

The verified “View all resources for Incremental Effectiveness” destination should use the exact concept slug incremental-effectiveness. The website link should not be added until its permanent Library URL structure has been confirmed.