Health Economics Metrics

Topic 43 of 76 · HTA frameworks and prevention economics

Health Technology Assessment (HTA)

HTA is the formal, institutionalized process by which health systems decide whether a technology — drug, device, or software — is worth paying for. It combines clinical-effectiveness evidence with economic evaluation under a published, mandatory methodology.

Why it matters

If you sell into a national health service, an HTA body may literally decide your market access. Knowing the local process is knowing your real regulator-of-value:

  • NICE (England): statutory appraisals under a defined reference case — QALYs from EQ-5D, NHS+PSS perspective, 3.5% discounting, PSA required — judged against £20k–£30k/QALY with severity modifiers; highly specialised technologies up to £100k+ with weighting.
  • ICER (US, non-governmental): evidence reports with a health-benefit price benchmark — the price at which a product would be cost-effective at $100k–$150k per QALY/evLYG — used as negotiating leverage; plus budget-impact "affordability alerts."
  • Canada (CADTH → CDA-AMC): reimbursement reviews at ≈CAD$50k/QALY; historically requested price cuts in ~95% of submissions.

The math

HTA's power is not a formula but a mandated method: every submission computes the same ICER under the same reference-case rules, so results are comparable across products and years. The reference case specifies outcome measure, utility instrument, perspective, comparator selection, discount rate, time horizon, and uncertainty analysis — removing every degree of freedom a sponsor could game.

Worked example

A digital therapeutic submits to NICE-style evaluation:

Model: ΔC = +£450/patient, ΔE = +0.03 QALYs → ICER = £15,000/QALY ✓ under £20k
Reference-case checks:
  utilities from EQ-5D-5L with UK value set              ✓
  comparator = current care pathway (not "no treatment") ✓
  PSA: 71% probability cost-effective at £20k            ✓ (reported)
  severity modifier: shortfall below ×1.2 bounds         — none claimed
Recommendation: routine commissioning, with real-world data collection.

The sponsor's own preferred analysis showed £9,000/QALY; the reference case pushed it to £15,000 by forcing the honest comparator. That gap is why reference cases exist.

Software engineering connection

The transferable artifact is the internal reference case: one mandated method for all tooling/platform business cases — declared comparator, standard unit costs (see national tariff and unit costs for the pattern), fixed discount rate, required sensitivity analysis, standard template. An "AMCP-dossier for tools" submitted to a platform council makes proposals comparable and gaming visible, exactly as HTA does for medicine. Start smaller than NICE did: a two-page template plus a published price book beats no standard at all.

Pitfalls

  • Treating HTA as a formality after regulatory clearance — CE/UKCA/FDA clearance says a product is safe; HTA decides if it's worth buying. Different hurdle, different evidence.
  • Building the economic model after the trial — evidence generation should be designed backwards from the reference case's requirements.
  • Ignoring jurisdiction differences: an ICER fundable in the US at $120k/QALY fails NICE at £30k; plan evidence and pricing per market.

Sources