Two-option incremental cost-effectiveness ratio

Divides the difference in expected cost by the difference in expected health effect, both taken as the intervention minus the comparator. The result reads as cost per unit of effect gained only when the effect difference is positive.

Signature

ICER = (C_1 - C_0) / (E_1 - E_0)
Inputs
InputsDefinitionUnit
C_1Expected total relevant cost of the evaluated interventioncurrency per defined population or person
C_0Expected total relevant cost of the comparatorthe same currency, price year and population basis as C_1
E_1Expected health effect of the evaluated intervention, for example QALYs or life yearsthe chosen health-outcome unit per defined population or person
E_0Expected health effect of the comparatorthe same health-outcome unit as E_1
Output
ICERAdditional expected cost per additional unit of expected health effect when the intervention replaces the comparatorcurrency per unit of health effect, for example £ per QALY gained

Function

Pairwise incremental cost-effectiveness function

Maps the expected costs and expected health effects of an intervention and one comparator to the additional cost per additional unit of health effect.

Try this function

Implementations

  • Excel

    Calculate a pairwise ICER

    Excel divides incremental cost by incremental effect using named ranges and returns #N/A when the two expected effects are equal. The sign of the effect difference is checked separately before the result is interpreted.

    =IF(InterventionEffect<>ComparatorEffect,(InterventionCost-ComparatorCost)/(InterventionEffect-ComparatorEffect),NA())
  • Excel

    Label the cost-effectiveness-plane position

    Excel labels the position of the comparison from incremental cost and incremental effect before the ratio is read. Results with one difference equal to zero are classified as dominance, consistent with the inclusive dominance rule.

    =IF(AND(IncrementalCost=0,IncrementalEffect=0),"No difference",IF(AND(IncrementalCost>0,IncrementalEffect>0),"North-east: compare ICER with threshold",IF(AND(IncrementalCost<0,IncrementalEffect<0),"South-west: saving per unit of effect forgone",IF(AND(IncrementalCost<=0,IncrementalEffect>=0),"Intervention dominates","Intervention is dominated"))))

Assumptions

  • One comparator and two mutually exclusive options

    The formula compares the evaluated intervention with one clearly identified comparator. When three or more mutually exclusive options are relevant, dominated options are removed and sequential ICERs are calculated between adjacent options on the frontier, as set out on the cost-effectiveness analysis page.

  • Consistent analytical basis for the ICER

    Costs and effects of both options refer to the same population, perspective, time horizon, price year and discounting approach, and the numerator and denominator both take the intervention minus the comparator.

  • Ratio of expected values

    The ratio is formed from expected incremental cost and expected incremental effect, as NICE requires, and not by averaging ratios calculated for individual simulations or patients.

  • Positive incremental effect for threshold comparison

    The ICER is compared with a threshold in the usual direction, as cost per unit of effect gained, only when E_1 minus E_0 is greater than zero. When the difference is negative the ratio is the saving per unit of effect forgone and the threshold comparison reverses, and when it equals zero no ratio is calculated.

  • Quadrant identified before interpretation

    The signs of incremental cost and incremental effect are examined before the ratio is read. An intervention that is less costly and more effective dominates, one that is more costly and less effective is dominated, and in neither case is the ratio compared with a threshold.

Worked examples

  • North-east quadrant: cost per QALY gained

    A new medicine costs £18,000 per patient and produces 4.5 QALYs, while current care costs £12,000 and produces 4.2 QALYs. Incremental cost is £6,000 and incremental effect is 0.3 QALYs, so the ICER is £20,000 per QALY gained. Because the incremental effect is positive the ratio can be compared with a threshold, and it lies below the lower end of the NICE range of £25,000 to £35,000 per QALY gained. The figures are illustrative.

    C_1 = 18000; C_0 = 12000; E_1 = 4.5; E_0 = 4.2; ICER = 20000
  • South-west quadrant: saving per QALY forgone

    A cheaper option costs £9,000 and produces 4.0 QALYs, against current care at £12,000 and 4.2 QALYs. Incremental cost is minus £3,000 and incremental effect is minus 0.2 QALYs, so the ratio is £15,000. Because the incremental effect is negative, the ratio is the saving per QALY forgone. At any threshold from £25,000 to £35,000 per QALY the saving is too small to justify the health lost, so a ratio below the threshold counts against the cheaper option. The figures are illustrative.

    C_1 = 9000; C_0 = 12000; E_1 = 4.0; E_0 = 4.2; ICER = 15000

Common errors

  • Reading a south-west ratio as cost per unit gained

    When incremental cost and incremental effect are both negative the ICER is positive, but it measures the saving per unit of effect forgone. A value below the threshold then favours the comparator, so applying the usual rule reverses the correct conclusion.

  • Comparing ICERs of non-adjacent options

    With several mutually exclusive options, ICERs calculated against one common baseline, or between options that are not adjacent on the frontier, can make a dominated or extendedly dominated option look acceptable. The pairwise formula applies to adjacent options only after dominated options have been removed, which is the sequential ICER on the cost-effectiveness analysis page.

  • Averaging ICERs across simulations

    The mean of ICERs from probabilistic simulations is not the ratio of mean incremental cost to mean incremental effect, and individual ratios become unstable when incremental effects are close to zero. The ICER is formed from expected values, and expected net benefit is used to compare options under uncertainty.

  • Ranking negative ICERs by size

    A negative ICER arises both when the intervention dominates and when it is dominated, and its magnitude carries no ranking information: larger savings push the ratio towards minus infinity while larger health gains pull it towards zero.

Sources

  • NICE manual on ICERs and decision rules

    National Institute for Health and Care Excellence. NICE technology appraisal and highly specialised technologies guidance: the manual (PMG36). Published 31 January 2022, last updated 31 March 2026. Chapter 4 Economic evaluation, section 4.2.16 (ICERs as the ratio of expected additional total cost to expected additional QALYs, with standard decision rules reflecting dominance and extended dominance).

    View source →

  • NICE manual on the cost-effectiveness threshold

    National Institute for Health and Care Excellence. NICE technology appraisal and highly specialised technologies guidance: the manual (PMG36). Published 31 January 2022, last updated 31 March 2026. Chapter 6 Committee recommendations, sections 6.3.4, 6.3.7 and 6.3.8 (ICERs of £25,000 and £35,000 per QALY gained) and 6.3.9 (the usual levels of £25,000 to £35,000 per QALY apply in the south-west quadrant).

    View source →

  • Cost-effectiveness plane

    Black WC. The CE plane: a graphic representation of cost-effectiveness. Medical Decision Making. 1990;10(3):212-214.

    View source →

  • Foundations of cost-effectiveness ratios

    Weinstein MC, Stason WB. Foundations of cost-effectiveness analysis for health and medical practices. New England Journal of Medicine. 1977;296(13):716-721.

    View source →

Canonical Identity

Stable URI · Machine-readable · Resolvable · CC BY 4.0

Two-option incremental cost-effectiveness ratio | HealthEconomics.wiki