𝐖𝐡𝐞𝐧 𝐚 𝐝𝐫𝐮𝐠 𝐬𝐡𝐨𝐰𝐬 𝐚 𝐡𝐚𝐳𝐚𝐫𝐝 𝐫𝐚𝐭𝐢𝐨 𝐨𝐟 𝟎.𝟖𝟎 — 𝐰𝐡𝐢𝐜𝐡 𝐬𝐨𝐮𝐧𝐝𝐬 𝐢𝐦𝐩𝐫𝐞𝐬𝐬𝐢𝐯𝐞 — 𝐚𝐧𝐝 𝐭𝐡𝐞 𝐭𝐫𝐢𝐚𝐥 𝐦𝐞𝐝𝐢𝐚𝐧 𝐬𝐮𝐫𝐯𝐢𝐯𝐚𝐥 𝐰𝐚𝐬 𝟏𝟓 𝐦𝐨𝐧𝐭𝐡𝐬 𝐢𝐧 𝐭𝐡𝐞 𝐜𝐨𝐧𝐭𝐫𝐨𝐥 𝐚𝐫𝐦, 𝐭𝐡𝐞 𝐦𝐚𝐭𝐡 𝐰𝐨𝐫𝐤𝐬 𝐨𝐮𝐭 𝐭𝐨 𝐫𝐨𝐮𝐠𝐡𝐥𝐲 𝟑 𝐚𝐝𝐝𝐢𝐭𝐢𝐨𝐧𝐚𝐥 𝐦𝐨𝐧𝐭𝐡𝐬 𝐚𝐭 𝐭𝐡𝐞 𝐦𝐞𝐝𝐢𝐚𝐧. 𝐍𝐨𝐭 𝐲𝐞𝐚𝐫𝐬. 𝐌𝐨𝐧𝐭𝐡𝐬. 𝐒𝐨𝐦𝐞𝐭𝐢𝐦𝐞𝐬 𝐰𝐞𝐞𝐤𝐬. The drug costs $28,000/month.
If a cancer drug showed “statistically significant survival benefit” in a clinical trial, what does that mean for the patient taking it? It might mean years. It might mean weeks. “Statistically significant” means the result is unlikely to be due to chance. It says nothing about size.
When a drug shows a hazard ratio of 0.80 — which sounds impressive — and the trial median survival was 15 months in the control arm, the math works out to roughly 3 additional months at the median. Not years. Months. Sometimes weeks.
That drug will cost $28,000 a month. It will go through a formulary process in which the manufacturer submits an economic model projecting its value over a 30-year horizon. The model will show, because of the way long-term survival projections work, something that looks considerably larger than 3 months.
A job posting this week describes the consultant hired to build that model: $150 an hour, 8 months, working across clinical, medical, and market access teams to “translate clinical data into inputs for economic models.” The collaboration with the market access team is in the job description. It is not incidental.
The patient navigating prior authorization, step therapy requirements, and cost-sharing for this drug does not know that the economic model justifying its price was built by a consultant hired by the manufacturer, using methods that contain documented directional bias toward favorable framing.
They were told it showed significant survival benefit. It did. Three months at the median. The rest is construction.
The HEOR Director is told “take data from cancer drug trials and feed it into economic models that determine what the drug is worth.” 𝐇𝐞𝐫𝐞’𝐬 𝐰𝐡𝐚𝐭 𝐭𝐡𝐚𝐭 𝐚𝐜𝐭𝐮𝐚𝐥𝐥𝐲 𝐢𝐧𝐯𝐨𝐥𝐯𝐞𝐬. 𝐓𝐡𝐞 𝐝𝐫𝐮𝐠 𝐚𝐭 𝐭𝐡𝐞 𝐞𝐧𝐝 𝐨𝐟 𝐭𝐡𝐢𝐬 𝐩𝐫𝐨𝐜𝐞𝐬𝐬 𝐜𝐨𝐬𝐭𝐬 $𝟐𝟖,𝟎𝟎𝟎 𝐚 𝐦𝐨𝐧𝐭𝐡. 𝐓𝐡𝐞 𝐭𝐫𝐢𝐚𝐥 𝐬𝐡𝐨𝐰𝐞𝐝 𝟑 𝐦𝐨𝐧𝐭𝐡𝐬 𝐨𝐟 𝐦𝐞𝐝𝐢𝐚𝐧 𝐬𝐮𝐫𝐯𝐢𝐯𝐚𝐥 𝐛𝐞𝐧𝐞𝐟𝐢𝐭. 𝐓𝐡𝐞 𝐦𝐨𝐝𝐞𝐥 𝐬𝐡𝐨𝐰𝐬 𝐬𝐨𝐦𝐞𝐭𝐡𝐢𝐧𝐠 𝐭𝐡𝐚𝐭 𝐥𝐨𝐨𝐤𝐬 𝐜𝐨𝐧𝐬𝐢𝐝𝐞𝐫𝐚𝐛𝐥𝐲 𝐦𝐨𝐫𝐞 𝐢𝐦𝐩𝐫𝐞𝐬𝐬𝐢𝐯𝐞. 𝐓𝐡𝐚𝐭 𝐢𝐬 𝐰𝐡𝐚𝐭 𝐭𝐡𝐢𝐬 𝐣𝐨𝐛 𝐢𝐬 𝐟𝐨𝐫.
Cancer drug trials run for 3–5 years. Economic models that justify drug prices need to project 5–10 years into the future beyond the clinical follow-up end date, because that’s what payers require. The gap between what the trial observed and what the model projects is filled by statistical extrapolation — a mathematical technique that extends the survival curve beyond the data.
Different extrapolation methods, fitted to the same trial data, can produce radically different long-term projections. A drug that extended median survival by 3 months in a trial might look, in one model, like it extends 10-year survival by 6 percentage points — or, in a different equally valid model, by 1 percentage point. The QALY gain, and therefore the justifiable price, differs by a factor of six depending on which model gets selected.
The person being hired makes that selection. They also select how transition probabilities move patients between health states in the model, which quality-of-life weights to apply, and how to model the costs the drug allegedly saves the health system.
Each choice is defensible. The aggregate is not neutral.
– 𝐕𝐨𝐥𝐭𝐚𝐢𝐫𝐞 So here’s my question: Should a national policy treat perfectly healthy, “at risk” people for Alzheimer’s before they show a single symptom?
I’ve spent the last several months researching and writing exactly that — a new national prophylaxis framework for Alzheimer’s prevention. Not early detection. Prevention before onset. [Spoiler alert: in June I will send the 1500-word policy proposal to Health Affairs]
The uncomfortable reality (why this is not academic)
Most healthcare policy waits for a diagnosis. My research flips the model: identify genetic, biomarker, or lifestyle-based risk in healthy individuals, then intervene with drugs, protocols, or monitoring.
The upside is obvious — delaying or stopping Alzheimer’s entirely. The downside is less discussed: labeling healthy people as “pre-patients,” potential over-medicalization, and a massive shift in who pays for what.
Whether you love or hate the idea, it’s coming. And that changes your industry.
The career hook (why you should care even if you hate policy)
Here’s where your job search enters the room.
A national Alzheimer’s prophylaxis policy would create entirely new roles:
• Genetic risk counselors for employers
• “Pre-diagnosis” care coordinators in insurance
• Compliance and ethics officers for at-risk data privacy
• New training specialties for geriatric nurses, data scientists, and benefits managers
If you work in health tech, HR, benefits brokerage, pharma sales, or public policy — this is a near-future skill set you can start building now. Ignoring it means competing against people who saw it coming.
What I’m actually doing with this research
I’m not just theorizing. My current writing outlines a state-level pilot framework that answers:
• Who consents for a healthy person?
• What happens if prophylaxis fails — or has side effects?
• How do employers handle “at risk” designations without discrimination?
I’ll be sharing key sections over the next few weeks. First up: the liability question that keeps corporate counsel up at night.
Voltaire was right: the right question is more revealing than any tidy answer.
My question to you — whether you’re in healthcare, tech, or just planning a 30-year career:
Are you waiting for Alzheimer’s prevention to become mainstream before you learn how it affects your job market? The framework is designed to be implementable at the primary care level — no specialist required. I practice what I preach — I’ve been following this framework myself for two years. The built-in design means that even if I was never at elevated risk, I’ve already realized measurable health and cost benefits. That’s the win-win.
Intrigued? The first installment drops next week.