Partial Economic Evaluation

Partial Economic Evaluation in Health Care

A partial economic evaluation is a health economic study that falls short of a full evaluation, because it looks at a single option on its own or compares options on costs alone or on outcomes alone. Hospital costing exercises and cost-of-illness estimates belong to this group, and so do clinical trials that report effects without costs. Such studies supply many of the unit costs and treatment effects that cost-effectiveness models depend on, yet none of them can show on its own whether an option is worth what it costs. The type of partial study tells an analyst which questions it can answer and what extra evidence a decision needs.

Five kinds of partial study

The usual classification comes from the textbook by Drummond, Sculpher, Claxton, Stoddart and Torrance (fourth edition, 2015). Malburg and colleagues reproduce an adapted version in a 2025 scoping review that names five partial types alongside the full form. A study is partial when it fails either of two tests, comparing two or more alternatives and examining costs and consequences together. The whole grid is set out under full economic evaluation, so the table below concentrates on the question each partial type can answer.

The third column is this page's gloss, not part of the source table. Malburg and colleagues add that a cost description may use micro-costing or macro-costing designs, which they equate with bottom-up and top-down costing. A cost analysis and a cost-of-illness study each have their own page, and Turner and colleagues place cost-of-illness work among analyses that examine only the costs of an intervention or a disease.

Why definitions of partial evaluation differ

The broad classification above is not the only one in use, and the differences matter when a systematic review decides what to include. The CRD guidance on systematic reviews (2009) describes a partial evaluation as one that makes no comparison and simply describes a single intervention or service through its costs or its consequences alone. Yet the same guidance goes on to treat cost analyses, which do compare alternatives, as partial evaluations that a review may sometimes include, so even one document mixes the two senses.

The Cochrane Handbook chapter on economic evidence (Aluko and colleagues, last updated August 2023) takes almost the opposite emphasis. It defines partial economic evaluations as studies that compare only costs or effects, and states that under this definition a trial comparing the effects and harms of an intervention is itself a form of partial economic evaluation. Turner and colleagues count both studies that examine costs or consequences independently and cost-outcome studies of a single course of action, and note that the literature varies on whether partial evaluations count as economic evaluation at all.

The label alone therefore does not fix the scope of a review, and a protocol that includes partial evaluations needs to say which cells it means.

What each partial form measures

Each partial form yields a different number. A cost description builds a total from quantities of resources and their unit costs. For one patient or one programme the calculation is

$$C_j = \sum_{k=1}^{K} q_{jk} \, p_k$$

where $C_j$ is the total cost for patient or programme $j$, $q_{jk}$ is the quantity of resource $k$ used, $p_k$ is the unit cost of resource $k$ and $K$ is the number of resource items counted. The choice of perspective decides which resources enter the sum, and the time horizon decides how long they are counted.

A cost analysis takes the same calculation for two alternatives and reports the difference.

$$\Delta C = \bar{C}_1 - \bar{C}_0$$

where $\Delta C$ is the incremental cost, $\bar{C}_1$ is the mean cost per patient under the option of interest and $\bar{C}_0$ is the mean cost per patient under the comparator. An effectiveness evaluation gives the matching health difference, $\Delta E = \bar{E}_1 - \bar{E}_0$, where $\bar{E}_1$ and $\bar{E}_0$ are the mean health effects under each option. Each comparative partial study therefore supplies one half of an ICER, and neither half alone settles the decision.

A cost analysis can still be taken one step further. Given a value $\lambda$ placed on a unit of health, it can state the smallest health gain that would make the extra spending worthwhile.

$$\Delta E^{*} = \frac{\Delta C}{\lambda}$$

where $\Delta E^{*}$ is the break-even health gain, $\Delta C$ is the incremental cost from the cost analysis and $\lambda$ is the value of one unit of health effect, for example GBP per QALY. The expression applies when $\Delta C$ is positive. This simple threshold analysis turns a cost-only result into a statement of what the missing effect evidence would have to show. It does not replace that evidence, because the true $\Delta E$ may be smaller, larger or even negative.

Where partial evidence is useful

Partial studies have legitimate jobs of their own. LeFevre and colleagues, writing on mHealth programmes, note that for programmes at demonstration stage with no comparator a partial evaluation may be all that is feasible, though without comparative evidence on value for money expansion remains difficult to justify. Cost descriptions also feed budget setting and service planning, and a cost-of-illness study conveys the scale of the burden a condition places on a health system.

The second job is as input to full evaluations. In a model-based evaluation the relative treatment effect often comes from an effectiveness evaluation and the unit costs from cost descriptions. The partial studies are then components, and the model supplies the comparison that each of them lacks.

Systematic reviews treat partial studies selectively. The CRD guidance says cost analyses may be relevant when the review question concerns incremental costs. The Cochrane Handbook notes that partial evaluations comparing only effects would already be in a Cochrane review, and that including cost-only analyses is not currently recommended, with methodological research needed on their value.

In some fields partial evidence is all there is. Malburg and colleagues, reviewing the costs of maternal and perinatal death surveillance and response in low and middle income countries, identified five studies with partial evaluations of these costs and reported that no full economic evaluation had been published.

Worked example with a pulmonary rehabilitation programme

All numbers in this example are illustrative. An NHS commissioner in England is considering a pulmonary rehabilitation programme for people with chronic obstructive pulmonary disease, compared with usual care, over one year with no discounting.

1. Cost description. Each patient attends 16 supervised sessions costing GBP 40 each and two assessments costing GBP 80 each. The programme cost per patient is 16 × 40 = 640 plus 2 × 80 = 160, so 640 + 160 = 800, or GBP 800. For 250 patients a year the budget line is 250 × 800 = 200,000, or GBP 200,000.

2. Cost analysis. A second study follows patients for a year and counts all their care. Programme patients cost GBP 800 for rehabilitation plus GBP 1,650 for other care, so 800 + 1,650 = 2,450. Usual care patients cost GBP 1,950. The incremental cost is 2,450 − 1,950 = 500, or GBP 500 per patient.

3. Break-even health gain. The NICE manual (PMG36) refers to a range of GBP 25,000 to GBP 35,000 per QALY gained. At GBP 25,000 the programme needs a gain of 500 / 25,000 = 0.02 QALYs per patient to break even. At GBP 35,000 it needs 500 / 35,000 = 0.0143 QALYs. Over one year a gain of 0.02 QALYs is the same as 0.02 × 365 = 7.3 days in full health, a concrete measure of the benefit required.

4. Adding effectiveness evidence. An effectiveness evaluation of the same programme against the same usual care reports a mean gain of 0.03 QALYs per patient over the year. Combining the two gives an ICER of 500 / 0.03 = 16,667, or about GBP 16,667 per QALY gained. The net health benefit at GBP 25,000 per QALY is 0.03 − 0.02 = 0.01 QALYs, and at GBP 35,000 per QALY it is 0.03 − 0.0143 = 0.0157 QALYs, both positive.

5. Interpretation. Neither partial study could support an adoption decision on its own. The cost description gave a budget and the cost analysis set a bar for the effect evidence to clear, and only the pairing produced a value-for-money result. That pairing is valid only if the two studies share the population, comparator, perspective and time horizon, which is why such evidence is normally combined in a decision model with its uncertainty analysed.

Boundaries with nearby study types

Two designs sit close to the partial cells but count as full. Cost-consequence analysis compares two or more alternatives on costs and several outcomes without combining them into one measure, and Turner and colleagues list it among the five main types of full evaluation. A cost-outcome description reports costs and outcomes for one option only, so the presence of a comparator is what separates them.

Cost-minimisation analysis also reports only costs, yet Turner and colleagues count it as full because the alternatives compared are assumed to have equivalent health effects. A plain cost analysis makes no claim of equivalence.

Reading a partial study without overreaching

Labels in the published literature are not a reliable guide to scope. In a 1997 review, Zarnke, Levine and O'Brien examined 95 health care studies labelled as cost-benefit analyses and reclassified 53% of them as cost comparisons because health outcomes were not appraised, judging only 32% to be bona fide cost-benefit analyses. They concluded that studies with that label often offer only partial programme evaluation.

Several misreadings recur. A cost-outcome description often reports a cost per unit of outcome for one option, an average cost-effectiveness ratio that implicitly compares the option with an alternative that costs nothing and produces no health. A cost analysis that finds a new option cheaper is often read as a saving, but if the option is also less effective the NICE manual (PMG36, section 6.3.9) still asks for the usual range of GBP 25,000 to GBP 35,000 per QALY to be considered. A large burden estimate from a cost-of-illness study says nothing about whether any particular intervention would reduce that burden at an acceptable cost.

The partial label describes the design and says nothing about rigour. A careful cost description can serve a modeller better than a weak full evaluation against an irrelevant comparator.

Sources

• Aluko P, Graybill E, Craig D, Henderson C, Drummond M, Wilson ECF, Robalino S, Vale L. Chapter 20: Economic evidence (last updated August 2023). In: Higgins JPT, Thomas J, Chandler J, et al, editors. Cochrane Handbook for Systematic Reviews of Interventions version 6.5. Cochrane; 2024. https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-20
• Centre for Reviews and Dissemination. Systematic Reviews: CRD's guidance for undertaking reviews in health care. University of York; 2009. Chapter 5. https://www.york.ac.uk/media/crd/Systematic_Reviews.pdf
• 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.
• LeFevre AE, Shillcutt SD, Broomhead S, Labrique AB, Jones T. Defining a staged-based process for economic and financial evaluations of mHealth programs. Cost Effectiveness and Resource Allocation. 2017;15:5. https://doi.org/10.1186/s12962-017-0067-6
• Malburg C, Willcox M, Sach T, Kinney M, Akweongo P, Biswas A, Serbanescu F. Cost of implementation and maintenance of maternal and perinatal death surveillance and response: a scoping review. BMC Pregnancy and Childbirth. 2025;25(Suppl 1):1016. https://doi.org/10.1186/s12884-025-08181-z
• National Institute for Health and Care Excellence. NICE technology appraisal and highly specialised technologies guidance: the manual (PMG36). Published 31 January 2022, last updated 13 October 2026. Sections 6.3.4 to 6.3.7 and 6.3.9. https://www.nice.org.uk/process/pmg36
• Turner HC, Archer RA, Downey LE, Isaranuwatchai W, Chalkidou K, Jit M, Teerawattananon Y. An introduction to the main types of economic evaluations used for informing priority setting and resource allocation in healthcare: key features, uses, and limitations. Frontiers in Public Health. 2021;9:722927. https://doi.org/10.3389/fpubh.2021.722927
• Zarnke KB, Levine MA, O'Brien BJ. Cost-benefit analyses in the health-care literature: don't judge a study by its label. Journal of Clinical Epidemiology. 1997;50(7):813-822. https://doi.org/10.1016/s0895-4356(97)00064-4