Health Economics Series Guide: Read the Evidence Before the Costs

อ่านฉบับภาษาไทย (Thai version)

Abstract

A clinical epidemiologist reads an economic evaluation by checking the evidence before the price tag. Start with the population, the outcome, and the time over which events were counted. A wrong denominator changes the health gain before costs enter the calculation. This guide maps a reading path from those clinical inputs to economic decisions. Part I covers occurrence, study design, treatment effects, survival, bias, evidence certainty, and diagnostic screening. Part II adds costs, health preferences, comparisons, decision models, and budget impact. A short illustrative example connects baseline risk with an absolute treatment effect. A separate calculation shows how assumed health gains and extra costs produce a value estimate. These calculations reveal the assumptions that need scrutiny; they do not validate those assumptions. Use the R boxes to reproduce the arithmetic and the browser calculators to explore alternatives. Read the articles in order, then keep value, affordability, and fairness as separate decision questions.

🔍 Click to enlarge Click anywhere to close

Why evidence comes first

A model cannot repair an unclear clinical question. Using a prevalence count as an incidence rate distorts how many new events the model predicts. Treating an association as a causal effect can then distort the benefit attributed to treatment.

Part I: Establish the clinical inputs

Part I has 7 articles. Start with occurrence and its denominator, then define the estimand: the exact effect the study aims to estimate. Design and analysis determine whether that question can be answered.

Next, connect relative effects to absolute risks and survival over a stated period. Examine bias, then judge certainty: confidence in the evidence for an outcome. Finish with screening, where predictive values and downstream care matter alongside sensitivity and specificity.

Part II: Evaluate the consequences

Part II has 5 articles. Define whose costs count, their price year, and how future costs and outcomes are discounted. Then examine utility, a preference weight used to adjust survival for health quality. A quality-adjusted life-year (QALY) combines time with those weights; a disability-adjusted life-year (DALY) measures health loss.

Compare extra costs with extra health gains before building models over time. A decision tree follows branching events; a Markov model tracks movement between health states. Probabilistic sensitivity analysis repeats calculations with sampled inputs to describe parameter uncertainty. End with budget impact and who receives the gains.

Reading map and clinical arithmetic

part_i <- 7; part_ii <- 5 # Reading map and illustrative inputs. Writes: stdout only.
cat("Epidemiology articles:", part_i, "\n")
cat("Economics articles:", part_ii, "\n")
cat("Articles after the guide:", part_i + part_ii, "\n")
incidence_rate <- 2 / 1000
cat("Incidence per 1000 person-years:", incidence_rate * 1000, "\n")
baseline_risk <- 0.217; relative_risk <- 0.56
cat("Control risk (%):", formatC(100 * baseline_risk, format = "f", digits = 2), "\n")
cat("Risk ratio:", formatC(relative_risk, format = "f", digits = 3), "\n")
risk_difference <- baseline_risk * (relative_risk - 1)
cat("Risk difference (pp):", formatC(100 * risk_difference, format = "f", digits = 2), "\n")
cat("Number needed to treat:", ceiling(1 / abs(risk_difference)), "\n")

Epidemiology articles: 7
Economics articles: 5
Articles after the guide: 12
Incidence per 1000 person-years: 2
Control risk (%): 21.70
Risk ratio: 0.560
Risk difference (pp): -9.55
Number needed to treat: 11

From extra costs and health to value

extra_cost <- 450000 # Illustrative comparison. Writes: stdout only.
extra_qalys <- 3; threshold <- 160000
icer <- extra_cost / extra_qalys
net_benefit <- threshold * extra_qalys - extra_cost
cat("Extra cost (THB):", format(extra_cost, big.mark = ","), "\n")
cat("Extra QALYs:", formatC(extra_qalys, format = "f", digits = 1), "\n")
cat("Threshold (THB/QALY):", format(threshold, big.mark = ","), "\n")
cat("ICER (THB/QALY):", formatC(icer, format = "f", digits = 3, big.mark = ","), "\n")
cat("Incremental net monetary benefit (THB):", format(net_benefit, big.mark = ","), "\n")

Extra cost (THB): 450,000
Extra QALYs: 3.0
Threshold (THB/QALY): 160,000
ICER (THB/QALY): 150,000.000
Incremental net monetary benefit (THB): 30,000

Use the calculators to test assumptions

First reproduce the R results. Then use econclinicalepidem to explore how clinical inputs change. Use economic-evaluation for cost and health comparisons, economic-decision-tree for branching events, and markov-model for repeated health-state transitions.

Change one assumption at a time and keep units and time horizons aligned. A matching calculation checks arithmetic; the evidence must still justify the inputs.

The series, in learning order

1. Health Economics Series Guide: Read the Evidence Before the Costs (this article)
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