Biostatistics & Population Health

ICER and willingness-to-pay thresholds

Clinical Overview and When to Suspect Cost-Effectiveness Decision-Making

— Numerator: incremental cost in dollars (direct medical, sometimes indirect)

— Denominator: incremental effectiveness, typically quality-adjusted life-years (QALYs) gained, sometimes life-years or cases averted

— $50,000/QALY — historical benchmark (dialysis-era)

— $100,000–$150,000/QALY — ICER (Institute for Clinical and Economic Review) contemporary range

— WHO global: 1–3× GDP per capita per QALY

— Vignette compares two screening strategies, drugs, or devices with both cost and outcome data

— Asks "which is most cost-effective" or "should the health system adopt"

— Population-health, value-based care, or formulary committee framing

— Mentions QALYs, life-years gained, or budget impact

Board pearl: ICER is always incremental — never compute cost/QALY of a single strategy in isolation. You need two strategies, and the comparator matters as much as the intervention itself. If a question gives you absolute costs and outcomes for two options, your first move is to subtract, not divide.

Key distinction: Cost-effectiveness ≠ cost-saving. A cost-effective intervention may still raise total spending; it just buys enough health to be worth the price.

Incremental cost-effectiveness ratio (ICER) is the central metric of cost-effectiveness analysis (CEA), used to decide whether a new intervention provides enough additional health benefit to justify its additional cost compared with the current standard.

Formula: ICER = (Cost_new − Cost_standard) / (Effect_new − Effect_standard)

Units: dollars per QALY gained ($/QALY). A lower ICER = better value.

Willingness-to-pay (WTP) threshold: the maximum $/QALY society or a payer will pay. In the US, commonly cited thresholds:

Decision rule: If ICER < WTP threshold → intervention is cost-effective and generally favored for adoption/coverage.

When to suspect a CEA question on Step 3:

Presentation Patterns and Key History

— A hospital or ACO is choosing between two competing interventions

— A formulary or P&T committee is deciding drug coverage

— A public health department compares screening intervals or modalities

— A payer asks whether to reimburse a new technology

— Two clearly labeled strategies (A vs B, or "new" vs "standard")

— Cost per patient or per intervention for each

— Effectiveness outcome: QALYs, life-years, cases detected, deaths averted

— Sometimes a stated WTP threshold ("the health system uses $100,000/QALY")

— Time horizon (lifetime, 10-year, etc.) and perspective (payer, societal, healthcare sector)

— Payer perspective — direct medical costs to insurer only

— Healthcare sector — all medical costs regardless of payer

— Societal perspective — adds productivity losses, caregiver time, transportation; recommended by the Second Panel on Cost-Effectiveness in Health and Medicine

Step 3 management: When the stem hands you raw numbers, work in this order — (1) identify the comparator, (2) compute ΔCost and ΔEffect, (3) divide for ICER, (4) compare to stated WTP. If no WTP is given, default to $100,000/QALY for US-based questions.

Board pearl: If the new intervention is both cheaper and more effective, it dominates — no ICER calculation needed; adopt it. If it is more expensive and less effective, it is dominated — reject it. ICER math only matters in the "more costly AND more effective" quadrant.

Step 3 vignettes on ICER typically arrive as a population health, quality improvement, or value-based purchasing scenario rather than a bedside case. Recognize the framing:

Key history elements in the stem:

Perspectives matter:

Discounting: Future costs and benefits are discounted (typically 3% annually in US analyses) to reflect time preference. Both costs and QALYs are discounted at the same rate.

Physical Exam Findings (and Hemodynamic Assessment when relevant)

— NE quadrant (more costly, more effective) → calculate ICER, compare to WTP

— SE quadrant (less costly, more effective) → dominant — always adopt

— NW quadrant (more costly, less effective) → dominated — always reject

— SW quadrant (less costly, less effective) → calculate ICER; may adopt if savings outweigh modest health loss (rare on exams)

— One-way sensitivity analysis: vary one parameter (e.g., drug price) and see if ICER crosses WTP

— Tornado diagram: ranks parameters by impact on ICER; widest bars = most influential drivers

— Probabilistic sensitivity analysis (PSA): Monte Carlo simulation; output is a cost-effectiveness acceptability curve (CEAC) showing probability the intervention is cost-effective across WTP thresholds

Key distinction: Simple dominance = one strategy beats another outright. Extended dominance = a strategy is beaten by a combination of two others on the efficiency frontier. Both must be removed before ranking remaining options by sequential ICER.

Board pearl: When Step 3 shows you a graph with strategies plotted on the cost-effectiveness plane, the efficiency frontier is the lower-right convex hull. Strategies off the frontier are dominated. The slope between adjacent frontier points = the ICER for moving up to the next strategy.

There is no physical exam in CEA — the "exam" is inspection of the cost-effectiveness plane, a 2×2 conceptual grid Step 3 may render as a graph or table.

Cost-effectiveness plane (ΔCost on y-axis, ΔEffect on x-axis):

"Hemodynamic" equivalents — sensitivity analyses that probe stability of the conclusion:

Inspect for "extended dominance" when comparing ≥3 strategies: a strategy is extendedly dominated if a linear combination of two others produces more QALYs at lower cost. Exclude it before computing sequential ICERs.

Diagnostic Workup — Initial Labs / Imaging / ECG / Biomarkers

— Drug acquisition cost (WAC, AWP, or net of rebates)

— Procedure cost (CMS fee schedule, DRG reimbursement)

— Downstream costs: hospitalizations averted or incurred, adverse event management

— Indirect costs (societal perspective only): lost wages, caregiver time

— Clinical outcomes from RCTs or meta-analyses (mortality, event rates)

— Utility weights (0 = death, 1 = perfect health) from standard-gamble, time-trade-off, or instruments like EQ-5D, SF-6D, HUI

— QALYs = Σ (time in health state × utility weight), discounted

— Disease prevalence/incidence (for screening models)

— Natural history transition probabilities (often Markov models for chronic disease)

Step 3 management: When asked which input most influences the ICER, look at the tornado diagram if provided, or reason qualitatively — drug price, comparator effectiveness, and utility weights typically dominate. Discount rate and time horizon also swing results.

Board pearl: Utility ≠ probability. A utility of 0.6 means a year in that health state is valued like 0.6 years of perfect health, not a 60% chance of survival. Confusing these on a vignette is a classic distractor.

The "labs" of CEA are the input parameters that populate the model. Step 3 expects you to recognize what data are needed and where they come from.

Cost inputs:

Effectiveness inputs:

Epidemiologic inputs:

Quick QALY math example: A patient lives 10 years at utility 0.8 = 8 QALYs. If a new drug extends this to 12 years at 0.85 = 10.2 QALYs. Incremental QALYs = 2.2. If the drug adds $200,000 in lifetime cost, ICER = $200,000 / 2.2 ≈ $91,000/QALY — cost-effective at $100K WTP.

Diagnostic Workup — Advanced or Confirmatory Studies

— Each input parameter is assigned a probability distribution (e.g., beta for utilities, gamma for costs, lognormal for relative risks)

— Monte Carlo draws (often 10,000 iterations) produce a cloud of ICERs on the cost-effectiveness plane

— Output: cost-effectiveness acceptability curve (CEAC) plotting P(cost-effective) vs WTP threshold

— Expected value of perfect information (EVPI): maximum a decision-maker should pay to eliminate all uncertainty

— Justifies funding additional research when EVPI is high

— Complementary to CEA — answers "can we afford it?" rather than "is it worth it?"

— Estimates total spending change over 1–5 years given expected uptake

— Critical for payers even when ICER is favorable

— NMB = (Effect × WTP) − Cost

— Positive NMB → intervention is cost-effective at that WTP

— Useful when comparing >2 strategies; highest NMB wins

— ISPOR good-practice guidelines for model transparency

— CHEERS reporting checklist (analog of CONSORT for CEA)

— ICER (the organization) publishes US value assessments using $100K–$150K/QALY benchmarks

Key distinction: CEA tells you value per dollar; BIA tells you total dollars. A drug with a great ICER ($60K/QALY) can still be unaffordable if it treats 2 million patients at $50K each. Step 3 may test this with a stem where the committee rejects a cost-effective drug for budgetary reasons — that is not a contradiction.

Board pearl: A CEAC crossing 50% at a WTP of $100K means there is a 50/50 chance the intervention is cost-effective at that threshold — useful for conveying uncertainty without a single point estimate.

Beyond a single ICER point estimate, sophisticated analyses include:

Probabilistic sensitivity analysis (PSA):

Value of information (VOI) analysis:

Budget impact analysis (BIA):

Net monetary benefit (NMB):

Confirmatory frameworks:

Risk Stratification or First-Line Management Logic

— Step 1: Identify all strategies and their (Cost, Effect) pairs

— Step 2: Rank by effectiveness (ascending)

— Step 3: Eliminate strongly dominated strategies (higher cost AND lower effect than another)

— Step 4: Eliminate extendedly dominated strategies (ICER higher than next more effective strategy)

— Step 5: Compute sequential ICERs along the remaining efficiency frontier

— Step 6: Adopt the strategy with the highest effectiveness whose sequential ICER is below WTP

— A: $1,000, 5.0 QALYs

— B: $3,000, 5.5 QALYs → ICER B vs A = $2,000/0.5 = $4,000/QALY

— C: $10,000, 5.7 QALYs → ICER C vs B = $7,000/0.2 = $35,000/QALY

— At WTP $50,000/QALY → choose C (most effective whose marginal ICER < WTP)

— At WTP $20,000/QALY → choose B

— At WTP $1,000/QALY → choose A

— Computing ICER of C vs A instead of C vs B (must use sequential comparator)

— Forgetting to discount future costs and QALYs

— Mixing perspectives mid-analysis

Step 3 management: When a vignette gives you a table of strategies, always rank by effectiveness first, then compute sequential ICERs. Compare each ICER to the WTP and choose the most effective strategy that still "passes."

Board pearl: The "right" strategy depends entirely on the WTP threshold the question specifies — read it twice. Changing WTP from $50K to $150K can flip the answer, and Step 3 loves to embed the threshold in a single sentence buried in the stem.

Decision algorithm for a CEA Step 3 stem:

Worked logic — three screening strategies:

Common traps:

Pharmacotherapy — First-Line Drug Regimen

— Childhood vaccinations (MMR, DTaP) — often dominant

— Tobacco cessation counseling + pharmacotherapy (~$2,000–$6,000/QALY)

— Statins for secondary prevention (<$10,000/QALY)

— Antihypertensives in established hypertension

— Colonoscopy screening at guideline-recommended intervals (~$20,000/QALY)

— Many novel oncology agents (immune checkpoint inhibitors often $100K–$200K/QALY)

— PCSK9 inhibitors at original pricing (>$300K/QALY → after price cuts ~$100K/QALY)

— Implantable cardioverter-defibrillators in primary prevention

— Some orphan drugs and gene therapies (though "ultra-rare" exceptions exist)

— Screening for very-low-prevalence conditions in average-risk populations

— Generic ACE inhibitor vs brand ARB (generic typically dominates)

— DOAC vs warfarin (DOAC often cost-effective ~$50K–$100K/QALY despite higher drug cost, due to fewer strokes/bleeds and no INR monitoring)

— SGLT2 inhibitor vs sulfonylurea (favorable ICER in HF/CKD subgroups)

Key distinction: Drug acquisition cost ≠ total cost. A more expensive drug can have a lower ICER if it prevents hospitalizations, avoids monitoring (warfarin INRs), or improves adherence. Always look at the total downstream cost captured in the model.

Board pearl: ICER (the organization) publishes value-based price benchmarks — when a drug's launch price exceeds the $100K–$150K/QALY benchmark, expect payer pushback, prior auth, or step therapy requirements regardless of FDA approval.

Applied to drug coverage decisions, the ICER framework drives formulary tier placement and prior authorization rules.

Classic high-value (low ICER) interventions — usually <$50K/QALY, often cost-saving:

Borderline value ($50K–$150K/QALY):

Low value (>$150K/QALY or higher):

Step 3 pharmacoeconomic vignettes often pit:

Procedures / Revascularization / Invasive Management (or expanded pharmacology if non-procedural)

— Number needed to screen (NNS) is often very large → small absolute QALY gains

— Lead-time bias and length-time bias can artificially inflate effectiveness if not modeled correctly

— Overdiagnosis lowers utility (treatment harms with no mortality benefit) and worsens ICER

— Example: low-dose CT lung cancer screening in NLST-eligible smokers ~ $50K–$80K/QALY (cost-effective); same screening in low-risk never-smokers → ICER balloons above WTP

— TAVR vs SAVR in intermediate-risk aortic stenosis ~$50K–$80K/QALY → cost-effective

— Left atrial appendage occlusion vs warfarin in AF ~$50K/QALY in high-bleed-risk patients

— CAR-T therapy ~$100K–$200K/QALY — sits near or above WTP, drives novel payment models (outcomes-based contracts, annuities)

— Decision trees — short time horizon, discrete events

— Markov models — chronic disease with recurring health states (well, sick, dead); cycle length matched to clinical event tempo

— Microsimulation — individual-level heterogeneity, used for cancer screening models (CISNET)

Step 3 management: For a screening question, expect the ICER to be most sensitive to disease prevalence and test specificity. Screening high-prevalence populations almost always improves ICER; screening low-prevalence populations rarely passes WTP.

Board pearl: Many cardiac devices (ICDs, TAVR) and screening programs are cost-effective in high-risk subgroups but not in average-risk populations — a recurring Step 3 theme that mirrors guideline-directed patient selection.

CEA applied to procedures, devices, and screening programs introduces extra wrinkles Step 3 may probe.

Screening interventions — special considerations:

Device and procedure CEAs:

Modeling structures:

Equity-informative CEA (distributional CEA) — weights QALYs by socioeconomic disadvantage; emerging in policy.

Special Populations — Elderly and Renal/Hepatic Impairment

— Shorter remaining life expectancy → fewer potential QALYs gained → ICER tends to be higher (worse) for interventions in older adults

— This can systematically disadvantage elderly populations in coverage decisions — a recognized age discrimination concern

— UK NICE explicitly forbids age-based discrimination; the US Affordable Care Act (Section 1182) prohibits CMS from using cost-per-QALY thresholds that discount older or disabled life-years

— Report life-years gained alongside QALYs

— Use equal value of life-years gained (evLYG) — ICER's alternative metric that weights all life-years equally regardless of baseline utility — designed to mitigate disability bias

— Proportional shortfall weighting (used in Netherlands) gives priority to those with greater lifetime health deficits

— Often excluded from RCTs → effectiveness inputs uncertain → wider PSA distributions, lower confidence in ICER

— Drug clearance issues raise adverse event rates and downstream costs

— Example: DOACs in advanced CKD — limited efficacy data, dose adjustments needed, ICER less favorable than in normal renal function

Key distinction: QALY captures both quantity and quality of life — useful but can penalize disabled or elderly populations. evLYG equalizes life-year value to address this. Step 3 may ask which metric is more "equity-sensitive."

Board pearl: US Medicare cannot legally use a hard $/QALY threshold to deny coverage — but private payers, PBMs, and state Medicaid programs can and do, often citing ICER reports.

Elderly patients and CEA — methodologic and ethical tensions:

Workarounds:

Renal/hepatic impairment populations:

Dialysis itself — the original WTP benchmark ($50K/QALY) was reverse-engineered from Medicare's 1972 ESRD program coverage decision. Modern dialysis ICER ~$80K–$130K/QALY.

Special Populations — Pregnancy, Pediatrics, or Other Demographic Subgroups

— Long remaining life expectancy → small per-year benefits compound into large QALY gains → pediatric vaccines and screening are often dominant or near-dominant (e.g., HPV vaccine ~$3,000–$30,000/QALY; rotavirus vaccine cost-saving in many models)

— Utility elicitation in children is methodologically harder (parent proxy, child-friendly EQ-5D-Y)

— Discounting at 3% over 70+ year horizons heavily attenuates future benefits → some argue for lower pediatric discount rates (1.5%)

— Dual-patient framing — interventions affect maternal and fetal QALYs

— Universal prenatal screening (HIV, syphilis, GBS, gestational diabetes) is typically cost-effective or cost-saving due to high downstream costs of untreated cases

— Aspirin for preeclampsia prevention in high-risk pregnancies → highly cost-effective

— Clear comparator (vaccinate vs no vaccinate)

— High effectiveness, low marginal cost

— Often dominant strategy — no ICER math required

— Race, ethnicity, and socioeconomic factors influence both baseline risk and access — analyses ignoring these can perpetuate disparities

— Distributional CEA explicitly models equity trade-offs

Step 3 management: When a vignette asks about cost-effectiveness of a routine childhood vaccine or universal prenatal screening, the answer is almost always "cost-saving / dominant" — the intervention prevents catastrophically expensive downstream disease in a long-lifetime population.

Board pearl: ACIP recommendations are explicitly informed by CEA, but vaccines also receive non-economic priority (herd immunity, equity, public trust) — Step 3 may frame this as a multi-criteria decision rather than pure ICER.

Pediatric CEA — unique features:

Pregnancy and maternal-fetal CEA:

Pediatric vaccine vignettes are a Step 3 favorite for cost-effectiveness because:

Equity considerations across subgroups:

Complications and Adverse Outcomes

— Effectiveness inputs from underpowered or short-horizon trials extrapolated over a lifetime

— Industry-sponsored CEAs systematically produce more favorable ICERs (publication and sponsorship bias)

— Comparing to placebo instead of current standard of care inflates incremental benefit

— Always demand the clinically relevant comparator

— Short horizons miss long-term benefits (e.g., 1-year horizon for a curative therapy understates value)

— Long horizons require uncertain extrapolation

— Failure to capture treatment toxicities in cost or utility terms biases ICER downward

— Choosing 0% vs 3% vs 5% can swing ICERs substantially, especially for prevention

— A drug with NNT=20 to prevent one event is not automatically cost-effective; depends on cost and event severity

— Treating WTP as a bright line ignores opportunity cost, budget impact, and disease severity

— Rigid ICER thresholds may underfund care for rare diseases, disabled populations, or end-of-life conditions → ethical pushback

Key distinction: Statistical significance ≠ clinical significance ≠ cost-effectiveness. A drug can have a p<0.001 mortality benefit, a clinically meaningful HR of 0.7, and still have an ICER of $500,000/QALY that fails any reasonable WTP. Step 3 may serially test these three concepts in one stem.

Board pearl: When a CEA shows an implausibly low ICER (e.g., $5,000/QALY for a brand-new oncology drug), suspect a flawed comparator, optimistic extrapolation, or industry sponsorship — and check the sensitivity analyses.

Pitfalls and "complications" of CEA methodology that Step 3 may probe:

Garbage-in-garbage-out:

Inappropriate comparator:

Time horizon truncation:

Ignoring adverse events:

Discount rate manipulation:

Confusing ARR/NNT with cost-effectiveness:

Threshold rigidity:

Adverse social consequences:

When to Escalate Care — ICU, Consult, or Inpatient Triage

— Single-input one-way analyses show the ICER straddles the WTP threshold

— Multiple parameters have wide credible intervals

— Decision is high-stakes (large population, irreversible coverage)

— Eligible population is large (>100,000 patients)

— Annual per-patient cost is high

— Payer cash-flow constraints matter (Medicaid, small employer plans)

— Disease severity, unmet need, equity, or rule-of-rescue concerns dominate

— Examples: ultra-rare pediatric disease, end-of-life cancer therapies

— Coverage denial would deny only available therapy

— Equity impacts are substantial (e.g., disease concentrated in marginalized population)

— Rule of rescue invoked (identifiable individual facing imminent death)

— Payer / P&T committee for formulary decisions

— CMS National Coverage Determination for Medicare coverage

— State Medicaid drug review boards for state-level decisions

— ICER public meetings for transparent value assessment

Step 3 management: Even when an ICER falls above the WTP threshold, the decision may legitimately be to cover the intervention if (1) it is the only option, (2) the affected population is severely disadvantaged, or (3) high uncertainty warrants more research rather than denial. Pure threshold-based denial is rarely the only "right" answer on ethically-framed Step 3 stems.

Board pearl: The rule of rescue — the moral pull to save identifiable lives at any cost — frequently overrides ICER thresholds in real coverage decisions, and is a recognized exception in value frameworks.

In CEA terms, "escalation" means when does a decision require additional analytic rigor, stakeholder input, or override of the ICER?

Escalate to probabilistic sensitivity analysis when:

Escalate to budget impact analysis when:

Escalate to multi-criteria decision analysis (MCDA) when:

Escalate to ethics committee / public deliberation when:

Stakeholder consultation:

Key Differentials — Same-Category Causes

— Effectiveness in natural units (life-years, mmHg reduction, cases detected)

— Cannot compare across diseases easily

— Special case of CEA where effectiveness = QALYs

— Enables cross-disease comparison; what people usually mean when they say "CEA"

— This is what generates $/QALY ICERs

— Both costs and benefits monetized (in dollars)

— Outputs net monetary benefit or benefit-cost ratio

— Requires monetizing life — controversial (willingness-to-pay surveys, value of statistical life ~$10M in US regulatory practice)

— Used only when two interventions have equivalent effectiveness

— Just compare costs; cheaper wins

— Common pitfall: applied when effectiveness equivalence has not actually been established

— Total spending change over a defined budget horizon

— Addresses affordability, not value

— NMB = WTP × ΔEffect − ΔCost

— Reframes the same decision; positive NMB = cost-effective

— Clinical metric, not economic — no direct cost dimension

Key distinction: CUA uses QALYs and enables cross-disease prioritization; CEA in natural units cannot. A dollar/QALY figure for diabetes can be compared to a dollar/QALY figure for cancer; a dollar per mmHg cannot be compared to a dollar per case detected.

Board pearl: "Cost-effective" colloquially is often misused to mean "cheap." On exams, cost-effective specifically means ICER < WTP threshold — the intervention may cost more, save less, or do anything in between, as long as the ratio passes.

Within health economic analysis, distinguish ICER from sister metrics — Step 3 may test the differences directly:

Cost-effectiveness analysis (CEA):

Cost-utility analysis (CUA):

Cost-benefit analysis (CBA):

Cost-minimization analysis (CMA):

Budget impact analysis (BIA):

Net monetary benefit (NMB):

Number needed to treat (NNT):

Key Differentials — Other-Category Causes

— Head-to-head clinical comparison without cost dimension

— PCORI-funded studies; cannot directly inform value

— Concerns study design quality (RCT > cohort > case-control)

— Independent of cost considerations

— Plan-Do-Study-Act, root cause analysis

— Local process change rather than population-level coverage

— Umbrella term encompassing CEA, BIA, clinical effectiveness, ethical/social impact

— Done by NICE (UK), CADTH (Canada), ICER (US private nonprofit)

— Payment models tying reimbursement to outcomes (bundled payments, ACOs, MIPS)

— Operationalizes value at provider/system level rather than per-intervention

— CEA specifically applied to drugs

— DALYs (disability-adjusted life-years) — WHO standard, similar to QALYs but burden-oriented; lower DALY = better health

— DALYs combine years of life lost (YLL) + years lived with disability (YLD)

— Distributional CEA, extended CEA, health equity impact assessment

Key distinction: QALYs maximize health gained; DALYs minimize health lost. They are conceptual mirror images. WHO global burden of disease work uses DALYs; most US/UK coverage decisions use QALYs.

Board pearl: A Step 3 stem mentioning "DALYs averted" is signaling WHO/global health framing, often with WTP set at 1× GDP per capita ($60K–$80K in the US, much lower in LMICs). The threshold itself is a hint to the perspective.

Distinguish CEA from related but distinct decision frameworks Step 3 may juxtapose:

Comparative effectiveness research (CER):

Evidence-based medicine (EBM) hierarchy:

Quality improvement (QI):

Health technology assessment (HTA):

Value-based purchasing / value-based care:

Pharmacoeconomics:

Population-level metrics:

Equity-focused frameworks:

Secondary Prevention / Discharge Medications / Long-Term Plan

— Formulary placement — preferred tier, no prior auth

— Clinical decision support — EHR prompts at point of care

— Quality measures — HEDIS, MIPS metrics aligning incentives

— Patient out-of-pocket design — value-based insurance design (V-BID) lowers copays for high-value services (statins, antihypertensives) and raises them for low-value services

— Prior authorization to limit use to highest-benefit subgroups (improves population ICER)

— Step therapy requiring trial of cheaper alternatives first

— Outcomes-based contracts — manufacturer refunds if drug fails in real-world use

— Indication-based pricing — different price per FDA indication based on value

— Post-market evidence often shows worse effectiveness than RCT (efficacy-effectiveness gap)

— Real-world ICERs frequently exceed pre-launch estimates

— Registry data (e.g., NCDR, TVT Registry) feeds back into updated CEAs

— When new evidence shows low value, deimplement (e.g., routine PSA screening reductions, vertebroplasty for osteoporotic fractures)

Step 3 management: A favorable ICER does not guarantee adoption — implementation requires aligning provider incentives, patient cost-sharing, and clinical workflow. Step 3 vignettes on value-based care often test recognition that V-BID lowers copays for high-value preventive services.

Board pearl: Value-based insurance design lowering copays for evidence-based services (statins post-MI, controllers for asthma) measurably improves adherence and outcomes — a Step 3-favored quality improvement intervention.

Once a CEA-informed decision is made, the long-term plan involves operationalizing the choice across the care system.

Implementation pathways for a cost-effective intervention:

Implementation pathways when an intervention is NOT cost-effective:

Monitoring real-world value:

Disinvestment:

Follow-Up, Monitoring Parameters, and Rehab/Counseling

— New comparator enters market (older drug going generic flips ICER dramatically)

— Long-term effectiveness data accrue (durability of response, late adverse events)

— Price changes (manufacturer rebates, biosimilar entry)

— Population shifts (broader indication, real-world heterogeneity)

— Adherence rates (lower than RCT → worse real-world ICER)

— Adverse event rates in routine practice

— Utilization in unintended populations (indication creep)

— Choosing Wisely campaign — discipline-specific lists of low-value services

— Shared decision-making aids that include cost information

— Transparency tools showing patient out-of-pocket costs

— High-value care = best outcomes at sustainable cost

— Generic substitution rarely sacrifices effectiveness

— More expensive ≠ better (imaging, brand-name drugs)

— Total cost of care per attributed life

— Hospitalizations per 1,000 enrollees

— HEDIS quality scores alongside cost

— MIPS cost category in Medicare Quality Payment Program

— Imaging for uncomplicated low back pain <6 weeks

— Routine annual EKGs in asymptomatic adults

— Antibiotics for viral URIs

— PSA screening without shared decision-making

Step 3 management: When a stem describes a clinician ordering low-yield, expensive tests (head CT for uncomplicated migraine, MRI for acute low back pain), the right answer involves deferring the test and counseling the patient — the cost-effective choice is also the guideline-concordant clinical choice.

Board pearl: Choosing Wisely lists are explicitly built on cost-effectiveness logic — recognizing them on Step 3 ties biostatistics to ambulatory management.

Ongoing monitoring of cost-effectiveness decisions parallels clinical follow-up:

Re-evaluate ICER when:

Monitoring real-world performance:

Counseling clinicians and patients about value:

Educational counseling content for patients:

Health system metrics to track value:

Avoiding low-value care (a Step 3 ambulatory theme):

Ethical, Legal, and Patient Safety Considerations

— Utilitarianism vs equity — maximizing aggregate QALYs may disadvantage rare diseases, elderly, disabled

— Rule of rescue — moral pull to save identifiable lives can override efficiency

— Procedural justice — decisions must be transparent, evidence-based, and appealable (Daniels & Sabin's "accountability for reasonableness" framework: publicity, relevance, revision, enforcement)

— Disability discrimination — QALY weights below 1.0 implicitly value disabled life-years less; addressed via evLYG

— ACA Section 1182 prohibits CMS from using QALY-based thresholds to determine Medicare coverage

— Inflation Reduction Act (2022) allows CMS price negotiation but limits use of QALYs

— State-level: some states (e.g., New York) restrict QALY use in Medicaid

— NICE in UK uses explicit £20,000–£30,000/QALY threshold — politically and legally permitted

— Low-value care causes direct harm (radiation from unnecessary CT, antibiotic resistance, anticoagulant bleeding in low-risk patients)

— Overdiagnosis from low-value screening leads to overtreatment

— Transition-of-care risk — formulary changes at hospital discharge can cause medication errors; verify post-discharge access and affordability before sending the patient home

— Financial toxicity is a recognized adverse event in oncology

— Patients have a right to cost information for shared decision-making

— Failure to discuss expected out-of-pocket costs is increasingly viewed as a consent deficiency

Step 3 management: Before discharging a patient on a new high-cost medication (DOAC, biologic, oncology agent), confirm insurance coverage, prior auth status, copay assistance enrollment, and pharmacy fill — unaffordable scripts are a major cause of post-discharge nonadherence and readmission.

Board pearl: "Financial toxicity" is now a documented patient-safety issue; ignoring cost in discharge planning is a transition-of-care failure Step 3 may explicitly test.

CEA raises distinctive ethical, legal, and patient safety issues central to Step 3 testing.

Ethical tensions:

Legal framework in the US:

Patient safety connection:

Informed consent and cost:

High-Yield Associations and Rapid-Fire Clinical Facts

— Historical US WTP: $50,000/QALY (1970s dialysis benchmark)

— Contemporary US WTP: $100,000–$150,000/QALY (ICER standard)

— WHO: 1–3× GDP per capita per QALY/DALY

— UK NICE: £20,000–£30,000/QALY

— Standard discount rate: 3% annually for both costs and QALYs

— Strongly dominated → eliminate

— Extendedly dominated → eliminate

— Compute sequential ICERs along efficiency frontier

— SE (cheaper + better) = dominant, adopt

— NW (costlier + worse) = dominated, reject

— NE = calculate ICER, compare to WTP

— SW = calculate ICER, may adopt if savings large

— Childhood vaccines, smoking cessation, statins (secondary prevention), antihypertensives in HTN, aspirin in established CV disease

— Lung cancer screening (NLST-eligible), TAVR, DOACs vs warfarin, PCSK9 inhibitors (post-price cuts)

— PSA screening in older men, routine imaging for low back pain, brand-name when generic equivalent exists

— ICER, QALY, DALY, NMB, CEA, CUA, CBA, CMA, BIA, PSA, CEAC, EVPI, HTA, evLYG, V-BID

Step 3 management: If forced to memorize a single threshold, use $100,000/QALY — it is the most defensible US default and the ICER organization's standard reference point.

Board pearl: Sequential ICERs along the efficiency frontier is the single most-tested computational skill — when given 3+ strategies, never compare each to the cheapest; always compare to the next-most-effective non-dominated strategy.

Memorize these numbers:

Dominance rules:

Cost-effectiveness plane quadrants:

Highest-value interventions (often dominant):

Borderline-value interventions:

Often-low-value interventions:

Key acronyms:

Board Question Stem Patterns

— Stem gives cost and QALY for two strategies, often with stated WTP

— Task: compute ICER, compare to WTP, choose adopt/reject

— Trap: forgetting to subtract (incremental, not absolute)

— One strategy is cheaper AND more effective

— Answer: adopt the dominant strategy without calculating an ICER

— Trap: distractor offers an ICER value for a dominant comparison

— Table of strategies sorted (or not) by cost or effectiveness

— Task: eliminate dominated options, compute sequential ICERs, select strategy at given WTP

— Tornado diagram or CEAC presented

— Task: identify the most influential parameter or the probability of cost-effectiveness at a stated WTP

— Industry-sponsored study with implausibly favorable ICER

— Task: identify methodologic flaw (wrong comparator, short horizon, missing adverse events)

— Intervention exceeds WTP but is the only therapy for a rare pediatric disease

— Task: recognize that rigid threshold application is not the only "right" answer

— Distinguishes CEA from BIA, comparative effectiveness, or value-based care

— Lowering copays for high-value services improves outcomes

— Patient cannot afford new prescription → readmission risk

— Choosing Wisely targets in ambulatory practice

Step 3 management: Read the stem twice for (1) the WTP threshold and (2) the comparator — both are decisive and frequently buried.

Board pearl: When the stem narrates an "expensive new drug" with great efficacy, your default first computation is ICER vs current standard, not vs placebo.

Pattern 1 — Two-strategy ICER calculation:

Pattern 2 — Dominance recognition:

Pattern 3 — Three+ strategy ranking:

Pattern 4 — Sensitivity analysis interpretation:

Pattern 5 — Methodology critique:

Pattern 6 — Ethics overlay:

Pattern 7 — Health policy framing:

Pattern 8 — Implementation/V-BID:

Pattern 9 — Discharge / transition-of-care:

Pattern 10 — Low-value care recognition:

One-Line Recap

The incremental cost-effectiveness ratio (ICER = ΔCost / ΔEffectiveness, usually in $/QALY) is judged cost-effective when it falls below a willingness-to-pay threshold — historically $50,000/QALY in the US, now commonly $100,000–$150,000/QALY — with dominated strategies discarded and sequential ICERs computed along the efficiency frontier.

Board pearl: When you see the words "cost-effective" on Step 3, mentally substitute "ICER below the stated WTP threshold" — never "cheap," never "saves money," never "clinically superior alone." This single discipline solves the majority of biostatistics value questions.

Step 3 management: Pair every cost-effective coverage decision with implementation foresight — formulary tier, prior auth, V-BID copay design, and post-discharge affordability checks — because an unfilled prescription has an effective ICER of infinity.

Compute and compare: ICER = (Cost_new − Cost_old) / (Effect_new − Effect_old); cost-effective if ICER < WTP. Default US WTP for Step 3 = $100,000/QALY unless specified otherwise.

Dominance first, math second: Strategies that are cheaper AND more effective dominate (adopt without calculation); strategies that are costlier AND less effective are dominated (reject). Only the "more costly + more effective" quadrant requires ICER computation against WTP.

Sequential ICERs, not pairwise vs cheapest: With 3+ strategies, eliminate strongly and extendedly dominated options, then compare each remaining option to the next-most-effective non-dominated option along the efficiency frontier — moving up only while sequential ICER < WTP.

Context beyond the ratio: Cost-effective ≠ affordable (budget impact analysis), ≠ equitable (evLYG, distributional CEA), and ≠ automatically covered (ACA Section 1182 prohibits CMS QALY-based denial). Rule of rescue, rare disease, severe unmet need, and patient financial toxicity legitimately modify pure threshold-based decisions in ethically-framed Step 3 vignettes.