PHIA Explained: Population Health Impact Assessment with Markov Models

Answers: “PHIA population health impact assessment tool”

Under §910(c)(4), FDA weighs a product's effect on the population as a whole — including users and non-users, and both cessation and initiation. That is what a Population Health Impact Assessment (PHIA) quantifies.

How the model is built

• A 5/6-state Markov model tracks cohorts across smoking states over a 50–75 year horizon.
• Core outputs: PDA (Δ premature deaths averted), LYG (life-years gained), and an adoption ratio (example values: 19,251 PDA, 481k LYG, 8.0 adoption ratio).
• Dual tipping-point sensitivity tests both the switching benefit (smokers moving to the product) and the youth-initiation risk — the two forces FDA balances.
• Scenario A vs. B comparison plus Monte Carlo captures uncertainty.

Pros vs. cons. Pro: it produces the exact net-benefit framing (PDA/LYG) reviewers use, with transparent sensitivity to the youth-initiation counterweight — computed in minutes vs. a weeks-long health-economics consult. Con: population models are only as good as their behavioral inputs (switching and initiation rates), which carry genuine uncertainty; results should be presented as scenarios, not single truths.

Who should NOT buy: applicants whose submission strategy relies on commissioned epidemiological studies and who don't need rapid scenario screening.