Cost-Effective Is Not Affordable, and Efficient Is Not Fair: Budget Impact and Equity

อ่านฉบับภาษาไทย (Thai version)

Abstract

An incremental cost-effectiveness ratio (ICER) compares additional spending with additional health gained. It does not show whether a payer can fund widespread access. Budget impact analysis tracks annual spending, while equity analysis asks who gains and who loses. We build an illustrative three-year budget using an eligible population, rising uptake and incremental treatment costs. Then we subtract health displaced elsewhere to distinguish gross gains from net health benefit. This conversion needs matching costs and horizons, plus an explicit assumption about health lost when resources move between services. A policy threshold alone cannot establish that assumption. We also examine treatment access, disease severity, protection from financial hardship and limitations of quality-adjusted life years. The equity impact plane makes trade-offs visible without deciding their ethical weight. Thai benefit-package discussions need evidence on value, affordability, feasibility and fairness together. The practical task is to present these questions separately, then explain the judgment that connects them.

🔍 Click to enlarge Click anywhere to close

Start with the payer and the budget calendar

Budget impact analysis (BIA) estimates how adoption changes a payer's annual spending. The payer perspective counts costs borne by the organization funding care. An incremental cost-effectiveness ratio (ICER) instead compares additional cost with additional health gained.

Quality-adjusted life years (QALYs) combine survival with health-related quality of life. A favorable cost per QALY cannot tell us whether next year's budget can fund access.

Report undiscounted annual spending over the payer's planning horizon [1]. Undiscounted means keeping payments in their budget years without converting them to present value. State expected price changes explicitly.

Count eligible people, uptake and displaced treatment

The eligible population includes existing cases plus new cases that meet the treatment criteria. Remove deaths, recovery and other exits before carrying people into later years. Do not add all prevalent cases again when they already belong to a continuing cohort.

Uptake is the share of eligible people who receive the intervention. Build its path from referral capacity, staffing, supply and patient preferences. Eligibility alone does not guarantee access.

Compare total payer spending with adoption against spending without adoption. Displaced treatments are existing services replaced by the new intervention. A cost offset is a saving elsewhere, such as fewer admissions, attributable to that intervention.

Avoid counting replaced treatment costs twice, both in the cost difference and as an offset. Match costs to actual treatment duration for people entering during a year. Read [[costing-perspective-inflation-discounting]] for perspective and price-year choices.

Turn uptake into annual spending

$$BI_t=N_t s_t(c_{new}-c_{old})-O_t$$

Here, $N_t$ is the eligible population and $s_t$ is uptake in year $t$. The costs $c_{new}$ and $c_{old}$ are payer costs per treated patient-year. Their difference excludes the additional offsets recorded in $O_t$.

The term $O_t$ is the total annual offset, not a per-person saving. Subtract it once. Add separate implementation costs when the per-patient difference omits them.

Use scenarios, meaning coherent alternative assumptions, for negotiated price and uptake. Lower prices can reduce spending without reducing access if clinical outcomes stay unchanged. Slower uptake lowers near-term spending but can delay health gains and widen access gaps.

Subtract the health lost elsewhere

Health opportunity cost is the health lost when resources leave other services. Let $k$ denote spending required to produce an additional QALY in those displaced services. A policy threshold is not automatically an estimate of this quantity. [3]

Net health benefit is health gained minus health displaced elsewhere. With health gain $\Delta E$ and incremental spending $\Delta C$, calculate it as follows.

$$NHB=\Delta E-\frac{\Delta C}{k}$$

Converting budget impact into QALYs requires matching the populations, cost definitions, timing and horizons. An ICER often uses lifetime discounted costs, while BIA uses short-term undiscounted payments. Dividing one by the other without reconciliation can mislead.

Calculate annual spending

# Illustrative annual budget. Writes: stdout only; no files.
eligible <- 10000
uptake <- c(0.2, 0.4, 0.6)
extra_cost <- 50000
annual <- eligible * uptake * extra_cost / 1e6
cat("Eligible patients per year:", format(eligible, big.mark = ","), "\n")
cat("Extra THB per patient-year:", format(extra_cost, big.mark = ","), "\n")
for (i in 1:3)
  cat("Uptake, year", i, ":", formatC(100 * uptake[i], format = "f", digits = 2), "%\n")
for (i in 1:3) cat("Budget, year", i, "(million THB):", annual[i], "\n")
cat("Total budget (million THB):", sum(annual), "\n")
cat("Annual offsets (THB):", 0, "\n")

Eligible patients per year: 10,000
Extra THB per patient-year: 50,000
Uptake, year 1 : 20.00 %
Uptake, year 2 : 40.00 %
Uptake, year 3 : 60.00 %
Budget, year 1 (million THB): 100
Budget, year 2 (million THB): 200
Budget, year 3 (million THB): 300
Total budget (million THB): 600
Annual offsets (THB): 0

Calculate health gained and displaced

# Illustrative matched-cost conversion. Writes: stdout only; no files.
spending <- sum(10000 * c(0.2, 0.4, 0.6) * 50000)
icer <- 150000
cost_per_qaly <- 160000
gained <- spending / icer
displaced <- spending / cost_per_qaly
cat("Matched spending (million THB):", spending / 1e6, "\n")
cat("Assumed ICER (THB/QALY):", format(icer, nsmall = 3, big.mark = ","), "\n")
cat("Opportunity cost (THB/QALY):", format(cost_per_qaly, nsmall = 3, big.mark = ","), "\n")
cat("QALYs gained:", format(gained, big.mark = ","), "\n")
cat("QALYs displaced:", format(displaced, big.mark = ","), "\n")
cat("Net QALYs:", gained - displaced, "\n")

Matched spending (million THB): 600
Assumed ICER (THB/QALY): 150,000.000
Opportunity cost (THB/QALY): 160,000.000
QALYs gained: 4,000
QALYs displaced: 3,750
Net QALYs: 250

Bring value, affordability and feasibility to Thai coverage decisions

Thai health technology assessment guidance places budget impact alongside economic evaluation [4]. A coverage proposal should show annual funding needs, clinical benefit and delivery capacity.

Price negotiation can improve value and reduce spending if outcomes stay unchanged. A lower price does not create staff or reliable follow-up. Identify who funds implementation and who might lose access elsewhere.

Read [[icer-dominance-ce-plane-and-thresholds]] for value comparisons and [[decision-trees-markov-models-and-psa]] for uncertainty.

Ask who gains and who gives up health

Distributional cost-effectiveness analysis (distributional CEA) examines how health gains and opportunity costs differ between population groups. Define the groups before comparing them, such as income groups or geographic areas. Count displaced health across groups as well as treatment benefits. [2]

An equity weight gives different social importance to health gains in specified groups. Such weights express value judgments, not biological facts. Report the unweighted results alongside the weighted results, and show whose gains receive priority.

Severity concerns how badly people are affected. Proportional shortfall is the share of expected remaining health lost because of disease under the comparator. State its reference health and time horizon; it is not simply a diagnosis label.

The rule of rescue describes a moral pull to help identifiable people facing grave harm. Rare diseases raise difficult questions about evidence, access and treatment development. Rarity alone does not establish severity, benefit or fairness.

A positive total net benefit cannot locate this example on the inequality axis. We have no group-specific data on access, benefit or displaced care. A phased rollout should therefore be assessed for who waits and who is left out.

Include protection against financial hardship

Extended cost-effectiveness analysis (extended CEA) reports health and financial consequences across groups, often by income. Financial risk protection means protection against financial hardship caused by paying for care. Public coverage can shift spending from households to the payer. [2]

Catastrophic expenditure means health payments exceed a stated share of household resources. Define that denominator and threshold before counting cases averted. Also examine impoverishment and forgone care; low spending may reflect inability to obtain treatment.

The budget example supplies no household data, so it cannot estimate catastrophic expenditure averted. Collect payment, income and access data before making that claim. Report the distribution alongside health gains and payer spending.

Treat cost per QALY as evidence for deliberation

QALYs make unlike health outcomes comparable, but they do not contain every social value. Age can affect remaining time to benefit. Disability and the utility scale can affect measured capacity to gain health. [5]

Those features can influence rankings even when people have equal claims to care. A health-state value must not be read as a person's worth. Equal QALY gains can also carry different meanings when severity and disadvantage differ.

Keep the aggregate calculation visible, then state any equity adjustment and its justification. Examine access, caregiver effects and financial hardship separately when the model omits them. Avoid counting the same severity concern again through overlapping weights.

The series, in learning order

1. Health Economics Series Guide: Read the Evidence Before the Costs
2. Incidence, Prevalence and the Right Denominator: The Epidemiology Beneath Cost-of-Illness and Markov Models [[incidence-prevalence-and-the-right-denominator]]
3. Design, Model and Estimand: What Each Study Design Can Actually Estimate [[study-design-estimand-and-effect-measures]]
4. Risk Ratio, Odds Ratio, Risk Difference and NNT: One 2×2 Table, Four Answers
5. The Hazard Ratio Is Not a Risk Ratio: Proportional Hazards, Restricted Mean Survival Time and Model Inputs
6. Chance, Bias and Confounding: Why an Association Is Not Yet a Treatment Effect
7. Beyond the Evidence Pyramid: GRADE Certainty and Choosing Effectiveness Inputs for Economic Models
8. From Sensitivity to Cost per Case Detected: Predictive Values, Prevalence and the Economics of Screening
9. Whose Cost, in Which Year? Perspective, Opportunity Cost, Inflation and Discounting
10. QALY, DALY and the Utilities Behind Them: What Is Actually Being Measured
11. ICER, Dominance and the Cost-Effectiveness Plane: Reading the Thai Threshold of 160,000 Baht per QALY
12. Decision Trees, Markov Cohorts and Probabilistic Sensitivity Analysis: How a Model Turns Evidence into a Decision
13. Cost-Effective Is Not Affordable, and Efficient Is Not Fair: Budget Impact and Equity (this article)