What Health Economics Asks

Health economics asks how choices are made when needs exceed available time, staff, and beds. It asks the same when needs exceed medicines, equipment, and money. It studies health itself, the production and distribution of care, the behavior created by incentives, and the consequences of choosing one option instead of another.

The field does not begin with the claim that cheaper is better. It begins with scarcity. Even a wealthy system cannot provide every potentially beneficial service immediately to everyone. A transparent analysis identifies alternatives, outcomes, costs, uncertainty, and who bears each consequence.

Scarcity creates opportunity cost#

Using an operating room for one procedure means it is unavailable for another during that period, and funding a new medicine may require new money, lower spending elsewhere, or longer waits. A clinician's hour spent on low-value paperwork cannot be spent on patient care, supervision, or improvement.

Opportunity cost is the value of the best alternative forgone. It can be measured in health, time, services, or other outcomes, not only dollars. The budget price of an intervention and its opportunity cost can differ.

This concept prevents a common error: treating a funded program as free because no new invoice appears. Staff, facilities, and attention still have alternative uses. It also prevents the opposite error of assuming every cost is waste. A costly service may produce benefits worth more than the alternatives it displaces. Whichever way the argument runs, the alternative has to be named: comparing a new technology with “nothing” is misleading when usual care is already consuming resources and producing outcomes.

Positive and normative questions#

Positive analysis describes what happens. How does a copayment change medicine use? Does a bundled payment alter hospital behavior? How much staff time does a screening program require? These questions can be investigated empirically.

Normative analysis concerns what should happen. How should benefits be distributed? Should severity receive extra weight? Which costs count? These questions involve values as well as evidence.

The boundary is not perfectly clean. Measuring health requires choices, and model structure embeds assumptions. Good analysis separates observations from judgments and states where values enter.

A result such as “the program costs $20,000 per quality-adjusted life year compared with current care” does not itself tell you whether to fund the program. The decision rule, the uncertainty, the budget, equity, and the alternatives all remain to be considered.

Health is produced beyond healthcare#

Medical care is one input into health. Housing, education, and food also matter. So do work, environment, and social conditions. So do behavior, genetics, and public health. Spending more on healthcare does not guarantee proportional health improvement.

Health economics examines both demand for care and production of health. Insurance can protect people from catastrophic costs while changing the price they face when deciding to seek care. Clinicians and institutions influence demand because patients rely on their expertise. Information is unequal, and outcomes are uncertain.

These features make healthcare different from a simple consumer market. A person with chest pain cannot compare emergency departments as if choosing an ordinary product. Professional duties, regulation, risk pooling, and public financing shape the system.

Market power also matters. Prices can reflect negotiation and concentration rather than production cost or clinical benefit. Economic analysis should not confuse a high price with high value.

Four forms of economic evaluation#

Cost-minimization analysis compares costs only after credible evidence establishes equivalent outcomes. That equivalence is a demanding condition and should not be assumed from a nonsignificant difference.

Cost-effectiveness analysis reports outcomes in natural units, such as strokes prevented, symptom-free days, or life-years. This keeps the outcome concrete but makes comparison across disease areas harder.

Cost-utility analysis often uses quality-adjusted life years, or QALYs, combining length of life and health-related quality into one measure; it supports comparison across programs but depends on how health states are valued and raises equity concerns.

Cost-benefit analysis expresses both costs and benefits in monetary terms. It can compare health programs with other sectors but requires contentious valuation of health and life.

A cost-consequence analysis presents several costs and outcomes separately. This can preserve detail, leaving decision makers to weigh components, though it does not produce one summary ranking.

The incremental question#

Economic evaluation compares differences. If a new option costs $12,000 and produces 0.2 more QALYs than current care, its incremental cost-effectiveness ratio, or ICER, is $60,000 per QALY.

That arithmetic is only the start. The estimate may be uncertain. The 0.2 gain may depend on extrapolation beyond trial follow-up. Costs may differ across settings. Adverse effects and monitoring may be incomplete.

An option that costs less and produces better outcomes dominates its comparator. An option that costs more and produces worse outcomes is dominated; when it costs more and helps more, the decision weighs the incremental gain against opportunity cost and other criteria. Ratios behave poorly near zero effect and are hard to summarize statistically, which is why you will often learn more from a cost-effectiveness plane or a net-benefit curve than from the ratio itself.

Perspective determines what counts#

A health-system perspective may count treatment, monitoring, hospital, and other healthcare costs. A payer perspective follows the specific payer's obligations. A societal perspective can include patient time, unpaid caregiving, transportation, productivity, and effects outside healthcare.

The Second Panel on Cost-Effectiveness in Health and Medicine recommended reporting both healthcare-sector and societal reference cases in the United States, with an impact inventory showing which consequences are included.

Perspective can change conclusions. A telehealth program may add equipment costs to a clinic while saving travel and time for patients. A rehabilitation service may increase healthcare spending but improve return to work.

Distribution matters too. A program can save the system money while shifting substantial cost to families. Reporting only the total hides who wins and who pays.

Time horizon and discounting#

The time horizon should be long enough to capture important differences in costs and outcomes. A vaccine may cost money now and prevent illness years later. A screening program may create immediate tests and procedures while benefits emerge over decades.

Short horizons can miss delayed benefit or harm. Very long horizons require stronger assumptions about survival, adherence, prices, practice, and technology.

Discounting gives less present value to future costs and outcomes. Rates vary by jurisdiction and guidance. Because preventive programs often spend now for later benefit, discount choices can affect conclusions materially, so an analysis will tell you its horizon, its discount rate, and why it chose them, then show what happens under reasonable alternatives.

Models connect evidence that trials cannot contain#

Trials rarely observe every relevant outcome for a lifetime. Decision trees, state-transition models, discrete-event simulations, and other structures combine trial results with epidemiology and utilities. They also combine them with costs and assumptions.

A model is a conditional argument. If the inputs and structural assumptions hold, the outputs follow. Validation asks whether coding is correct, outputs behave as expected, and predictions align with external evidence where possible.

Structural uncertainty can matter more than a precise parameter. Whether disease states allow recovery, whether treatment effect wanes, and how adverse events recur can change results. Testing only confidence intervals around fixed assumptions understates uncertainty. Transparency includes model structure, parameter sources, and code availability where possible. It includes funding, conflicts, and scenario analyses. CHEERS 2022 guides reporting; it is not a guarantee that the chosen model is appropriate.

QALYs are useful and contested#

One QALY represents one year in a health state assigned a utility of one, or an equivalent combination of time and utility, and a year valued at 0.5 contributes 0.5 QALYs under the conventional calculation.

This summary allows programs with different outcomes to be compared. It can incorporate both survival and health-related quality. Yet it compresses domains and depends on valuation methods. It may not capture caregiver effects, dignity, or hope. It may not capture convenience or all disability experiences.

Equity concerns arise when the same health gain produces different QALY increments because of baseline disability or life expectancy. Jurisdictions respond differently through modifiers, deliberation, alternative measures, or legal constraints.

The correct stance is neither to treat QALYs as a complete moral calculus nor to dismiss the information they provide; report what the measure captures, what it omits, and how the decision process addresses those limits.

Cost-effective does not mean affordable#

Cost-effectiveness relates incremental cost to incremental outcome. Budget impact estimates how adoption changes actual spending over a defined short-to-medium period for a payer or system.

A treatment for a rare condition can have a large cost per patient but modest total budget impact, and a low-cost intervention offered to millions can be cost-effective yet require substantial new spending.

Budget-impact analysis needs the eligible population, uptake, and displacement. It needs capacity, price, offsets, and timing. It should show scenarios because adoption rarely follows one fixed path.

Affordability is not a permanent property. Prices change, patents end, delivery capacity expands, and competing needs shift. Coverage conditions and price negotiation can alter both value and budget impact.

Incentives change behavior#

Payment methods influence what organizations provide. Fee-for-service can reward volume. Capitation can support prevention and coordination but may create pressure to limit costly care. Budgets constrain total spending but can produce waits if capacity and demand are misaligned.

No payment design removes incentives. Blended systems try to balance access, quality, complexity, and stewardship. Measures can be gamed when payment depends on narrow targets.

Patients respond to prices, but not all use is discretionary. Cost sharing can reduce low-value and high-value care together, especially when people cannot distinguish them. Financial protection is therefore a health objective as well as a financing mechanism. All of that behavior belongs inside the model: a price change modeled on the assumption that prescribing, adherence, and market entry stay put may misestimate both the cost and the outcome.

Equity and distribution#

An average ICER can hide who benefits. Distributional cost-effectiveness analysis examines how health gains and opportunity costs fall across groups. Equity impact analysis can consider access, financial risk, geography, socioeconomic position, and other dimensions.

Giving priority to severity, rarity, or children is a value judgment that should be explicit. So is giving priority to caregivers or underserved groups. Some systems apply formal modifiers; others use deliberation.

Equal treatment can preserve unequal access when distance, language, disability, or time costs differ. Implementation costs needed for equitable reach should be included rather than treated as inefficiency, and small subgroup samples create real uncertainty, so a distributional claim should not outrun its data, and privacy must be protected.

Reading an economic evaluation#

Start with the decision problem: population, intervention, and comparator. Then take perspective, horizon, and setting. Then ask whether the clinical-effect evidence is credible and applies to the people you have in mind, because economic precision cannot repair a biased efficacy estimate.

Trace resource quantities separately from prices. Ask when the costs were valued and whether currency conversion and inflation are reported. Then check the adverse events, the monitoring, the patient's own time, and the downstream care.

Inspect extrapolation, discounting, missing data, and model validation. Look for deterministic sensitivity analysis, probabilistic uncertainty, scenario analysis, and expected-value-of-information methods where relevant.

Check funding and conflicts without assuming that sponsorship determines validity. Transparency, methods, and replication remain the basis for judgment.

The health-technology assessment guide shows how economic evidence joins clinical, organizational, ethical, and social information. The cost-effectiveness threshold guide develops the decision context in more detail. The site's research overview connects these methods to evidence appraisal.

What the analysis can responsibly conclude#

A strong economic analysis can show which assumptions drive value, where uncertainty lies, what adoption may displace, and who bears consequences, and it can compare coherent alternatives and reveal when additional research would be valuable.

It cannot determine society's priorities by formula. The final decision still requires legitimate rules, public accountability, equity judgment, and a feasible implementation plan. Health economics improves that decision when it makes tradeoffs visible and testable.

References#

1. WHO Guide to Cost-Effectiveness Analysis
2. WHO economic evaluation publications
3. CHEERS 2022 reporting guidance
4. Second Panel on Cost-Effectiveness in Health and Medicine
5. NICE health technology evaluations manual
6. ISPOR budget-impact analysis good-practice guidance

This article is educational, not legal, financial, reimbursement, or health-policy advice. Requirements and decision rules vary by jurisdiction and payer.