Pharma Cases 2026: Doses, Patients and the Net Price

What arrives in a pharma case is never the list price

Three features separate this sector from anything a generic framework handles.

The payer is not the user. A physician prescribes, a patient takes, and an insurer or a government pays, so willingness to pay belongs to an institution with a formulary rather than to a consumer with a wallet.

The realized price is a fraction of the published one. Statutory rebates, commercial rebates, discount programs, wholesaler fees and patient assistance all sit between the list price and the cash that reaches the manufacturer. The industry calls the whole stack gross to net, and no manufacturer publishes its own rate for a named product.

And exclusivity ends on a date. When it does, price and volume fall together rather than in sequence, which is why the drop is called a cliff.

The rule that dates the cliff is public, and it is more precise than a patent expiry date. CMS states that a drug ceases to be selected for negotiation when a generic or biosimilar "is marketed pursuant to such approval or licensure", checked 29 September 2026 at cms.gov. Approval alone changes nothing. The FDA keeps the dated list of approved biosimilars and their reference products, which records entries such as "Zimrixby (rituximab-cdxx) / July 2026 / Rituxan (rituximab)", checked 29 September 2026 at fda.gov. So approval is visible and the commercial launch is the event.

SINHA, a pharma case recomputed from ZS's own figures

ZS publishes a pricing case with full answers, and it is the cleanest public arithmetic in the sector. Its client "expects the FDA will soon approve its drug, SINHA, for the treatment of chronic obstructive pulmonary disease (COPD)", checked 29 September 2026 at zs.com.

Take the published figures in order and check each one.

Prevalence first. The answer states that "five percent of the U.S. population (about 330 million) equals 16.5 million people affected by COPD". Check it: 5% on 330 million is 16.5 million. The firm then notes the client called it an estimate and suggests working with 15 or 20 million.

Dosing second, and this is the pivot of the whole case. "All treatments are administered every other week so patients will need 26 doses per year."

Cost third. The answer builds $50 a dose from three lines: "$6 million for 200,000 doses equals $30 per dose", "$375,000 per hour for 75,000 doses equals $5 per dose", and "$150,000 for 10,000 doses equals $15 per dose". All three divisions hold, and they sum to $50.

Capacity fourth. "75,000 doses/hour times 2,000 hours/year equals 150 million doses/year", and 150 million divided among 26 doses a patient gives 5.77 million patients, which the firm rounds to about 6 million and says so.

Price last. The recommendation is "to price SINHA at $300/dose", and the answer claims "80%+ profit margins" at that level. Check that too: the margin is $250 on a $300 price, which is 83.3 percent.

One factor of twenty six decides a pharma case answer

Everything above is per dose, and every question a payer asks is per patient. Convert, and the case changes shape.

A patient year costs 26 doses at $300, which is $7,800. The gross margin on a patient year is 26 doses at $250, which is $6,500. Those two figures are what a formulary committee, a health economist and a sales forecast all work in.

Now see the size of the error available here. Take a forecast of one million treated patients. At $7,800 a patient the gross revenue is $7.8 billion a year. A candidate who carries the $300 per dose figure into a per patient forecast reports $300 million. The gap is a factor of 26, and $300 million still reads like a plausible pharmaceutical number, which is why the slip survives a review.

The same discipline applies to capacity. ZS's 150 million doses a year sound like a large number until they are divided by 26, at which point the plant covers 5.77 million patients against an addressable 16.5 million. Doses measure the factory and patients measure the market, and mixing them will make a capacity constraint appear or vanish.

Gross to net, and the margin it removes

Assume a 34 percent gross to net deduction across the whole book. That number is an assumption we are declaring, because no manufacturer publishes its own and CMS publishes maximum fair prices for selected products, which is a different thing entirely. The stack behind such a rate has four kinds of entry, and a case answer should name them rather than quoting one percentage.

1. Statutory rebates. Owed to public programs under formulas set in law.
2. Commercial rebates. Negotiated with plans and their intermediaries for formulary position.
3. Channel fees and chargebacks. Paid to wholesalers and to sites of care that buy at a mandated discount.
4. Patient support. Copay assistance and free goods, which reduce net revenue without reducing volume.

Work it through. 34% on $300 is $102 of deductions, so the net price is $198 a dose. The margin becomes 198 less 50, which is $148, down from $250. That is a fall of 102 over 250, which is 40.8 percent of the margin.

Scale it to the forecast. On one million patients the gross revenue at list is $7.8 billion and the net revenue is 26 doses at $198, so $5,148 a patient, which is $5.148 billion. The list to net gap is $2.652 billion a year.

Read what that shows about the case rather than about the number. A pricing recommendation of $300 a dose is a recommendation about a number that nobody pays. The decision the client faces is which channels get which deduction, and a rebate given to one payer to win formulary position reduces the price for every patient in that channel rather than only the marginal one.

The structure of those deductions also moves without any negotiation. CMS revises the Part D benefit through annual instructions, most recently "for Calendar Year (CY) 2026", checked 29 September 2026 at cms.gov. A change in which party pays which slice of a prescription changes a manufacturer's deduction rate with the list price untouched.

A cliff dated by the first marketed competitor

Now put the end of the product's life in the model, because a five year plan that stops at peak sales has answered half the question.

Assume that at loss of exclusivity the net price per dose falls 60 percent and the product keeps 22 percent of its volume within eighteen months. Both are assumptions and both should be said aloud.

Start from the $5.148 billion of net revenue. 40% on $5.148 billion is $2.059 billion after the price fall. Then 22% on $2.059 billion is $453 million after the volume loss. That is 8.8 percent of what the product earned the year before.

State the conclusion in the unit the client decides in. The product returns $453 million a year after the cliff against $5.148 billion before it, so 91 percent of the net revenue disappears. Any launch investment has to earn back inside the exclusive period, and the exclusive period ends when a competitor ships rather than when a competitor is approved.

That distinction is worth one more sentence because it is where the sourcing helps. The FDA list shows approvals against reference products in 2026, and CMS ties its own consequence to a product being marketed. A candidate who dates the cliff from the approval date will be early, sometimes by years, and a candidate who dates it from patent expiry alone will miss the biosimilar pathway entirely.

Where a pharma answer goes wrong

Our own grader puts four weights on a piece of case arithmetic. They are our weights, read from the rubric code in this repository, and ZS publishes three practice cases with answers and no scoring of any kind. Where each of the four weights comes from is explained on how an answer is assessed.

Drilling the pharma case dose to patient chain

Rework SINHA with one input changed at a time, and say the unit out loud before every product.

Dosing frequency first, because it is the factor that does the most damage. Suppose a monthly formulation replaces the fortnightly one, so a patient year is 13 doses rather than 26. Hold the patient year price at $7,800 and the price per dose becomes $600. Hold the price per dose at $300 and the patient year becomes $3,900. Those are two completely different products, and the only thing that changed was the schedule.

Deduction rate second. Take gross to net from 34 to 45 percent. The net price becomes 0.55 times 300, which is $165 a dose, and the margin becomes $115. On one million patients the net revenue becomes 26 doses at $165, so $4,290 a patient, which is $4.29 billion. Eleven points of deduction cost $858 million a year, which is worth knowing before a rebate is offered.

Then check the prevalence assumption the whole forecast rests on. ZS's own answer suggests using 15 or 20 million rather than 16.5 million, so the addressable population carries a spread of a third. Every revenue figure above moves with it, and a candidate who presents one point estimate has hidden the widest uncertainty in the case.

Our breakeven calculator handles the fixed cost recovery question when a launch has development spend to earn back, and the math drill times conversions like doses to patient years and grades Units on its own line. Both are free and need no account.

For the payer side of the same system, see healthcare cases. For diagnostics and devices, where a fee schedule sets the price instead of a rebate stack, see life sciences cases. For pricing a product with no third party payer in the chain, see pricing strategy cases.