Concept Architecture
Concept
Theoretically, Extended Dominance Analysis is the formal analytical procedure used in cost-effectiveness analysis to identify and eliminate interventions subject to extended dominance. It is based on the principles of economic efficiency and opportunity cost, ensuring that only interventions lying on the efficient cost-effectiveness frontier remain under consideration. The analysis determines whether an intervention is inefficient because a combination of adjacent alternatives can achieve additional health benefit at a lower incremental cost-effectiveness ratio.
Mathematically, Extended Dominance Analysis is performed by ordering interventions according to increasing effectiveness, calculating sequential incremental cost-effectiveness ratios (ICERs) and comparing adjacent ICERs. If the ICER of one intervention exceeds the ICER of the next more effective intervention, that intervention is extendedly dominated, removed from the analysis and the ICERs are recalculated iteratively until no further violations remain.
In practice, Extended Dominance Analysis is a standard component of fully incremental cost-effectiveness analysis in health technology assessment. Analysts rank interventions by effectiveness, eliminate strongly dominated alternatives, identify interventions subject to extended dominance and reconstruct the efficient cost-effectiveness frontier before comparing the remaining ICERs with the relevant willingness-to-pay threshold.
Purpose
Used to identify and remove interventions subject to extended dominance, thereby constructing the efficient cost-effectiveness frontier and ensuring that only economically efficient alternatives are considered in decision-making.
Mathematical Formulae
Primary Formula
Extended dominance is identified when:
ICER???,? > ICER?,???
where:
- ICER???,? = incremental cost-effectiveness ratio between interventions i?1 and i
- ICER?,??? = incremental cost-effectiveness ratio between interventions i and i+1
If this condition is satisfied, intervention i is eliminated and the ICERs are recalculated.
Supporting Formulae
Incremental Cost-Effectiveness Ratio:
ICER = ?C / ?E
where:
- ?C = incremental cost
- ?E = incremental effectiveness
Related Mathematical Methods
- Fully incremental analysis
- Incremental cost-effectiveness analysis
- Dominance analysis
- Cost-effectiveness frontier construction
- Decision-analytic modelling
Example
Four interventions are ranked by increasing effectiveness.
| Intervention | Cost (�) | QALYs | ICER (�/QALY) |
|---|---|---|---|
| A | 10,000 | 4.0 | ? |
| B | 14,000 | 4.3 | 13,333 |
| C | 18,000 | 4.5 | 20,000 |
| D | 20,000 | 4.8 | 10,000 |
Extended Dominance Analysis compares the sequential ICERs. Because the ICER for C (�20,000/QALY) exceeds the ICER for the more effective intervention D (�10,000/QALY), intervention C is removed. The remaining interventions are then re-analysed to produce the efficient cost-effectiveness frontier.
Excel Implementation
| Function | Example Formula | Health Economics Application |
|---|---|---|
| SORT | =SORT(A2:D6,3,1) | Orders interventions by effectiveness before incremental analysis. |
| IF | =IF(E3>E4,""Extendedly Dominated"",""Efficient"") | Identifies interventions meeting the extended dominance criterion. |
| FILTER | =FILTER(A2:E6,E2:E6<>""Extendedly Dominated"") | Removes extendedly dominated interventions before recalculating ICERs. |
| INDEX / MATCH | =INDEX(E:E,MATCH(MAX(D:D),D:D,0)) | Retrieves ICERs for sequential comparison during the analysis. |
VBA (Optional)
Automate fully incremental cost-effectiveness analysis by iteratively identifying extendedly dominated interventions, removing them and recalculating ICERs until the efficient frontier is established.
Sources
- Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes. 4th ed. Oxford University Press; 2015.
- Briggs A, Claxton K, Sculpher M. Decision Modelling for Health Economic Evaluation. Oxford University Press; 2006.
- Gold MR, Siegel JE, Russell LB, Weinstein MC, eds. Cost-Effectiveness in Health and Medicine. Oxford University Press; 1996.
- Fenwick E, Claxton K, Sculpher M. Representing uncertainty: the role of cost-effectiveness acceptability curves. Health Economics. 2001;10(8):779?787.
- NICE. Health Technology Evaluation Manual. Latest edition.
Related Concepts (2)
Frequently Asked Questions (6)
What is extended dominance analysis?
The process of identifying interventions subject to extended dominance by comparing each option's ICER against that of the next most effective alternative.
Source: Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes. 4th ed. Oxford University Press; 2015.
What steps does extended dominance analysis follow?
Options are ordered by increasing effect, simply dominated options are removed, and incremental ratios are calculated between each remaining option and the next less effective one. Any option whose ratio falls below that of the option beneath it is removed. The ratios are then recalculated for the reduced set, since removing an option changes the comparators around it, and the process repeats until the ratios rise consistently across the survivors.
Source: Drummond et al. 2015
Why must extended dominance analysis be repeated?
Because each removal changes the neighbours of the options that remain, and a recalculated ratio can reveal an option that has become extendedly dominated only after the earlier removal. A single pass therefore does not produce a clean frontier. The termination condition is that incremental ratios increase consistently across the remaining options, which indicates no further removals are possible. Where the analysis is presented as a table, showing the ratios at each iteration makes the sequence inspectable rather than leaving only the final result.
Source: Drummond et al. 2015
What does extended dominance analysis produce?
A frontier of surviving options, each with an incremental ratio against its immediate predecessor, together with a list of the options removed and the reason for each. The frontier is then compared against the threshold to identify the most effective option whose ratio remains acceptable. Reporting the removals matters, since a reader cannot otherwise distinguish a deliberate exclusion from an omission. Recording which options were removed at which stage also allows the procedure to be checked, since an error in one iteration propagates to all that follow.
Source: Drummond et al. 2015
What errors occur in extended dominance analysis?
Omitting the step entirely is the most common, which leaves options on the frontier that a mixture of others would beat. Calculating ratios against a single fixed baseline rather than against the next surviving option is the second, and produces figures that cannot be compared with a threshold. Applying the procedure to options that are not mutually exclusive is the third, since independent options should be ranked individually and funded downward.
Source: Drummond et al. 2015
When is extended dominance analysis unnecessary?
Where only two options are compared, since extended dominance requires at least three. Where the analysis is fully probabilistic and options are compared on expected net benefit, the sequential removal is not performed at all. And where options are independent rather than mutually exclusive, the frontier construction does not apply and each option is assessed against the threshold on its own. It is also unnecessary where every option has already been shown to be simply dominated except one, since no combination can then beat the survivor.
Source: Briggs, Claxton & Sculpher 2006
Trust Record
Verified by Dr Darrin Baines
British health economist
Professional identity: darrinbaines.org
Verification date: 6 Aug 2025
Content version: 1.0.0
Canonical Identity
- Term code
- HE-EE-CEA-030
Stable URI · Machine-readable · Resolvable · CC BY 4.0