Intervention

Intervention: What Counts as One and How It Is Specified for Economic Evaluation

In health economics the word intervention names a broad class: anything done to protect, restore or improve health, or to organise care, from a single blood test to a national screening programme. Within an economic evaluation an intervention also takes on a role, as the option being assessed against a comparator, which is usually another intervention. This page sets out what the class includes, how an intervention is used in a comparison and in a NICE scope, what a usable description specifies, and why strategies and continuation rules count as separate interventions. An illustrative example shows one medicine, defined in two ways, giving two different cost-effectiveness results.

What counts as an intervention: health technology, treatment and complex interventions

The class is wide, and sources draw its edges in slightly different places. The World Health Organization's International Classification of Health Interventions defines a health intervention as an act performed for, with or on behalf of a person or a population whose purpose is to assess, improve, maintain, promote or modify health, functioning or health conditions. Acts aimed at whole populations, such as screening programmes, fall within it as readily as a single prescription.

The international definition of health technology assessment published in 2020 defines a health technology as an intervention developed to prevent, diagnose or treat medical conditions, promote health, provide rehabilitation or organise healthcare delivery, and notes that it can be a test, device, medicine, vaccine, procedure, programme or system. On that definition every health technology is an intervention, and NICE's word technology plays the part that intervention plays in a PICO question. Treatment is narrower: it covers interventions that act on an existing condition, whereas prevention, diagnosis and service organisation are interventions without being treatments.

Some interventions are complex, in the sense used by the framework commissioned by the Medical Research Council and the National Institute for Health Research: they have several components, target a range of behaviours or several groups, settings or levels, need particular skills to deliver and receive, or allow flexibility. As a result, what is delivered in routine care may differ from what was tested.

How an intervention is used in a comparison

Economic evaluation is defined in the standard textbook as the comparative analysis of alternative courses of action in terms of both their costs and their consequences. An intervention therefore has an economic result only in relation to something else: the analysis estimates what changes, in cost and in health, if it is used instead of the comparator. The intervention's own expected cost and expected health effect depend on what it is taken to include, so a change in its definition changes the incremental cost and the incremental effect, and with them the incremental cost-effectiveness ratio.

Being the intervention or the comparator is a role in a particular comparison, not a property of the thing itself. A medicine that is the intervention in one appraisal can be the comparator in the next, once it has become established practice. The Cochrane Handbook calls the comparator a control intervention, which may be inactive (placebo, no treatment, standard care or a waiting list) or active (a different drug, therapy or variant of the same intervention).

When a decision problem has more than two options, they are compared by fully incremental analysis. The options are ranked by cost, options subject to dominance or extended dominance are removed, and each remaining option is compared with the next less costly one. Comparing net monetary benefit across all options at a stated threshold gives the same choice.

Where the intervention sits in PICO and in a NICE scope

The intervention is one of the four elements of a structured question, alongside the population, the comparator and the outcomes (the PICO framework). Health technology assessment uses the same structure, and the question has to be fixed before evidence is gathered or a health economic model is built. The intervention anchors the other elements of a decision problem: the population it is offered to, the resources it consumes, the evidence about its effects and the model parameters that describe them.

NICE's manual for technology appraisal and highly specialised technologies (PMG36) uses the word technology for the intervention. Its scope defines the issues for consideration and sets the boundaries of the evaluation. The areas it lists include the population for whom the technology is evaluated, the technology itself and where it will be used (for example in a hospital inpatient or outpatient setting, or in the community), the care pathway, the potential comparators, the principal outcome measures, the costs and the time horizon. Comparators are the potentially relevant options that are established practice in the NHS. For evaluations of HealthTech the manual allows interventions to be defined as a group or class of technologies with shared features or functions, such as laboratory tests for a particular genetic marker.

Item 7 of the CHEERS 2022 reporting statement likewise asks authors to describe the interventions or strategies being compared and why they were chosen.

What a usable description of an intervention specifies

A label such as "drug X" or "a falls prevention service" is not enough to cost or to model. The Cochrane Handbook advises that drug interventions are defined by preparation, route of administration, dose, duration and frequency, and that multi-component interventions are defined by their common or core features. The TIDieR checklist, a 12-item template for describing interventions, adds who provides the intervention, how and where it is delivered, the number of sessions with their schedule, duration, intensity or dose, and any planned tailoring.

Each detail has an economic counterpart. NICE's manual notes that implementing a new technology affects the units of the evaluated and comparator technologies provided, such as doses of drugs or theatre hours, and can have knock-on effects on other resources, including staff numbers and hours, training and support services. The cost items follow from the description, so a vague description produces a vague costing.

For diagnostic technologies the manual asks for the technology to be described precisely because many variants may exist, for example different test thresholds, and notes that these variants may need to be evaluated separately or in different sequences. A test is rarely the whole intervention; its value comes through the treatment decisions that follow, which is why the scope covers the care pathway.

Strategies, sequences and continuation rules

In modelling, the unit being compared is often a strategy rather than a single product: a policy for what is done, to whom and when, including what happens after a response or a failure. Decision tree and Markov model structures represent each strategy as a separate arm, and a treatment sequence model represents an ordered series of treatments as one option.

NICE's manual recognises this in two places. When both the technology and the comparator form part of a sequence in the care pathway, the evaluation may compare alternative sequences. A treatment continuation rule, under which treatment continues only for people who achieve a specified response within a given time, should not be analysed as a subgroup; instead the strategy with the rule is analysed as an additional treatment strategy alongside the base-case interventions and comparators, so that the costs and consequences of any extra monitoring are included.

Worked example: one medicine, two strategies

The figures below are illustrative and describe a fictional medicine, drug X, compared with standard care over a lifetime horizon, with costs and QALYs already discounted. The scope lists two versions of the intervention: drug X given to everyone who starts it, and drug X with a continuation rule that stops treatment at 12 weeks in people who have not responded.

1. The naive comparison: drug X for all against standard care, ignoring the rule.

$$ ICER_{all,SC} = \frac{20000 - 8000}{6.40 - 6.00} = \frac{12000}{0.40} = 30000 $$

where $ICER_{all,SC}$ is the cost in pounds per QALY gained by giving drug X to everyone instead of standard care.

2. Compare the continuation-rule strategy with standard care. Ranked by cost, the rule strategy is the next option after standard care:

$$ ICER_{rule,SC} = \frac{14000 - 8000}{6.30 - 6.00} = \frac{6000}{0.30} = 20000 $$

where $ICER_{rule,SC}$ is the cost in pounds per QALY gained by moving from standard care to drug X with the continuation rule.

3. Compare drug X for all with the continuation-rule strategy.

$$ ICER_{all,rule} = \frac{20000 - 14000}{6.40 - 6.30} = \frac{6000}{0.10} = 60000 $$

where $ICER_{all,rule}$ is the cost in pounds per QALY gained by continuing drug X in non-responders as well.

4. Interpret. No option is dominated or extendedly dominated, because the sequential ratios rise from £20,000 to £60,000 per QALY gained. Once the rule strategy is in the comparison, the decision-relevant ratio for treating everyone is £60,000 per QALY gained against the rule strategy, not the £30,000 of the naive comparison. At an illustrative threshold of £25,000 per QALY gained, drug X with the continuation rule is preferred. The medicine is the same throughout; only the definition of the intervention changed. The rule strategy is only a real option if response can be measured reliably at 12 weeks and the rule is followed, and the cost of that assessment belongs in its £14,000.

Where the intervention is misdescribed

The most common problem is a mismatch between the intervention in the evidence and the intervention in the model. A trial may test a dose, setting or delivery mode that differs from the one proposed for routine use, and transferring its effect without comment hides an assumption. Costing the intervention from its acquisition price alone, without the administration, staff time, training and follow-up that its description implies, is a related error. Conclusions hold only for the intervention as defined, in the population and against the comparators in the scope, and they do not carry over automatically to other doses, settings or sequences.

Sources

• Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes. 4th ed. Oxford: Oxford University Press; 2015.
• Hoffmann TC, Glasziou PP, Boutron I, et al. Better reporting of interventions: template for intervention description and replication (TIDieR) checklist and guide. BMJ. 2014;348:g1687. Checklist: https://www.equator-network.org/reporting-guidelines/tidier/
• Husereau D, Drummond M, Augustovski F, et al. Consolidated Health Economic Evaluation Reporting Standards 2022 (CHEERS 2022) statement: updated reporting guidance for health economic evaluations. BMJ. 2022;376:e067975. https://pmc.ncbi.nlm.nih.gov/articles/PMC8749494/
• McKenzie JE, Brennan SE, Ryan RE, Thomson HJ, Johnston RV, Thomas J. Chapter 3: Defining the criteria for including studies and how they will be grouped for the synthesis. In: Higgins JPT, Thomas J, Chandler J, et al, editors. Cochrane Handbook for Systematic Reviews of Interventions version 6.5. Cochrane; 2024. Section 3.2.2. https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-03
• National Institute for Health and Care Excellence. NICE technology appraisal and highly specialised technologies guidance: the manual (PMG36). London: NICE; 2022, updated March 2026. Sections 2.1, 2.2, 4.9.13 and 4.11.4. https://www.nice.org.uk/process/pmg36
• O'Rourke B, Oortwijn W, Schuller T; International Joint Task Group. The new definition of health technology assessment: a milestone in international collaboration. International Journal of Technology Assessment in Health Care. 2020;36(3):187-190. doi:10.1017/S0266462320000215. Note 1 of the definition credits the HTA Glossary for the definition of health technology.
• Skivington K, Matthews L, Simpson SA, et al. A new framework for developing and evaluating complex interventions: update of Medical Research Council guidance. BMJ. 2021;374:n2061. https://pmc.ncbi.nlm.nih.gov/articles/PMC8482308/
• World Health Organization. ICHI Reference Guide: International Classification of Health Interventions. Maintenance-platform version dated October 2025 (the guide states it is not the released version); WHO suggested citation: International Classification of Health Interventions (ICHI), Geneva: WHO, 2023. https://icd.who.int/dev11/Downloads/Download?fileName=ichi%2FICHI_Reference_Guide.pdf