Skip to content
Back to search results

Cost-Effectiveness Analysis in HTA

Introduced by
Diane Meade Drummond, Michael Gold
Year
1996
Type
Economic evaluation methodology
Family
Health economics, Decision analysis
Font size

What does it do?

Cost-Effectiveness Analysis (CEA) is an economic evaluation method that compares the cost and health benefits of alternative treatments to determine whether an intervention provides good value for money.

Within Health Technology Assessment, CEA is the primary tool for recommending reimbursement and coverage decisions.

The intuition behind the method

Healthcare resources are finite. A new treatment may extend life by 6 months at a cost of $100,000, while an alternative extends life by 3 months at $30,000. Which is better value? CEA answers this by calculating cost per unit of health benefit (e.g., cost per year of life gained or cost per quality-adjusted life year). Decision-makers compare this ratio to a threshold (e.g., $50,000 per QALY) to determine if the technology is worth funding. CEA forces explicit discussion of what health benefit is worth what cost.

History

CEA emerged from health economics in the 1970s. Gold, Siegel, Russell, and Weinstein published the foundational 'Cost-Effectiveness in Health and Medicine' in 1996, establishing reference case methods and standards. Drummond's parallel 'Methods for the Economic Evaluation' series became the methodological bible. By the 2000s, CEA was adopted globally: Australia (PBAC), Canada (CADTH), UK (NICE), and most European countries now require CEA for reimbursement. The US lags but is moving toward CEA; many commercial and Medicare reviews increasingly use cost-effectiveness.

When to use it

Use CEA when comparing two or more treatments for the same condition, when time horizons are long (need to project beyond trial data), or when health gains are measured in QALYs rather than monetary terms. CEA is standard for reimbursement decisions in Australia, Canada, the UK, and increasingly in the US. Avoid CEA if outcomes cannot be measured (subjective preferences only), if the technology has no health impact, or if comparing across very different disease areas requires aggregation across non-comparable outcomes.

Limitations and considerations

  • QALY calculation requires utility weights (preference scores) that vary by population and may not capture individual patient values
  • Time horizon and discount rate choices significantly affect results; selecting these requires judgment
  • Modeling long-term costs and outcomes (e.g., lifetime) introduces substantial uncertainty
  • Does not capture equity: a treatment may be cost-effective overall but benefit only wealthy patients
  • Thresholds (e.g., $50,000 per QALY) are politically determined and vary across countries; no universal standard

Source conditions do not replace independent scientific review or validation by an analysis engine.

Strengths

  • Directly answers the policy question: is the benefit worth the cost given our threshold?
  • Enables transparent comparison across diverse technologies and diseases using a common metric (cost per QALY)
  • Flexible framework accommodates many types of outcomes and time horizons through modeling
  • Sensitivity and scenario analysis quantify uncertainty and support robust decision-making
  • Standardized methodology allows benchmarking across countries and institutions

Applications

  • Evaluating whether to cover a new cancer drug by comparing cost per QALY to cancer treatment threshold
  • Assessing whether a preventive program (vaccination, screening) is cost-effective by projecting disease avoidance and lifetime cost savings
  • Comparing surgical techniques by analyzing operating cost, complication rates, quality of life, and long-term reoperation risks
  • Determining optimal screening strategies (age to start, frequency) by modeling false positives, overdiagnosis, and treatment sequelae
  • Supporting budget allocation across disease areas by ranking programs by cost-effectiveness and available budget

How it works

How it works

1

Identify the intervention and comparators. Collect clinical data on outcomes (survival, symptom improvement) from trials and observational studies. Collect cost data on drugs, procedures, hospitalizations, and long-term care from invoices, registries, and published sources.

2

Standardize costs to a reference year and adjust for inflation. Choose the analytical perspective (patient, provider, societal). If international comparison is intended, adjust costs for currency and purchasing power. Calculate outcomes in natural units (years of life gained) or utility units (QALYs).

3

Perform a base case analysis with best-available estimates. Calculate the incremental cost-effectiveness ratio (ICER): (Cost_new − Cost_old) / (Outcome_new − Outcome_old). For example, if a new drug costs $50,000 more and provides 1 additional QALY, ICER = $50,000 per QALY.

4

Develop a decision tree or Markov model to project costs and outcomes beyond the trial horizon (often lifetime). Use sensitivity analysis to test how results change if key assumptions vary (treatment effectiveness, cost, discount rate).

5

Plot ICER in a cost-effectiveness plane showing cost and effectiveness trade-offs. Compare ICER to published thresholds. Discuss uncertainty, subgroup analyses, and implications for adoption and coverage.

Common pitfalls

  • Using a single trial with short follow-up to estimate lifetime costs and benefits; always extrapolate with a model and sensitivity analysis
  • Omitting important cost categories (e.g., side effect management, end-of-life care) or using biased sources for costs
  • Selecting a discount rate without justification; standard is 3% for both costs and health benefits, but range 0–5% significantly affects results
  • Claiming precision beyond the data: CEA results are estimates with confidence intervals; presenting a single ICER as truth is misleading
  • Ignoring subgroup analysis; a technology may be cost-effective for elderly patients but not young; always disaggregate by clinically meaningful groups

Frequently asked questions

What is the difference between CEA and cost-benefit analysis?

CEA measures benefits in natural units (QALYs, life years). Cost-benefit analysis converts benefits to monetary value (e.g., value of a life = $5 million). CEA is preferred in healthcare because monetizing health is ethically contentious; CEA separates the technical question (cost per outcome) from the value judgment (is this threshold acceptable?).

How do I calculate a QALY?

QALY = (years of life) × (quality weight). Quality weight ranges 0 (death) to 1 (perfect health). A treatment extending life 5 years with quality weight 0.8 = 4 QALYs. Weights come from instruments like EQ-5D (patient self-report) or TTO (time trade-off), eliciting patient preferences.

What discount rate should I use?

Standard is 3% per year for both costs and health benefits (US and many other countries). UK uses 3.5%. The discount rate reflects how much we value future costs and health relative to present: 3% per year means we value a benefit 20 years from now at 55% of its present value.

What if ICER is dominated (more costly and less effective)?

The new treatment should not be funded; the old treatment dominates. If ICER is outside the threshold but provides substantial benefit to a subgroup, consider conditional coverage (e.g., restricted to patients with genetic marker) or requiring further evidence.

How do I handle uncertainty in CEA results?

Use one-way sensitivity analysis (vary one parameter at a time) to identify key drivers. Use probabilistic sensitivity analysis (randomly sample all parameters from distributions) to generate confidence intervals. Create a cost-effectiveness acceptability curve showing probability of cost-effectiveness across thresholds.

Related methods

Sources

  1. Gold, M. R., Siegel, J. E., Russell, L. B., & Weinstein, M. C. (Eds.). (1996). Cost-Effectiveness in Health and Medicine. Oxford University Press.

    ISBN: 9780195108231
    Copied.
  2. Drummond, M. F., Sculpher, M. J., Claxton, K., Stoddart, G. L., & Torrance, G. W. (2015). Methods for the Economic Evaluation of Health Care Programmes (4th ed.). Oxford University Press.

    Open publication reference
    Copied.
  3. Shiroiwa, T., Sung-Jae, I., Fukuda, T., & Sanon, M. (2016). International survey on QALYs and cost-effectiveness thresholds. Health Policy, 120(5), 504–514.

    Open publication reference
    Copied.