BRENNEN

Industries / Health / Healthcare / Pharmacy

Specialty pharmacy

Specialty pharmacies supply costly drugs for long-term and life-threatening diseases such as cancer, multiple sclerosis and hemophilia, drugs that are often injected or must be kept cold, and they also do the insurance paperwork, patient training and follow-up that come with them.

1985

Le Bonheur Health Systems, the parent of a not-for-profit children's hospital in Memphis, Tennessee, founded the pharmacy business that became Accredo, which by its 1998 fiscal year earned 46% of its revenue from drugs for Gaucher disease, 23% from multiple sclerosis drugs and 23% from hemophilia drugs.

sec.gov

39

Milestones achieved across 6 domains.

10

To watch for: announced, not yet confirmed outside.

12

Open deliverables. Nobody has shipped them.

In one minute

Where it started

In 1985 a children's hospital group in Memphis set up a pharmacy business for a few very costly drugs; as Accredo, it described its drugs in 1998 as injected, temperature-controlled and costing up to $200,000 a patient a year. Fewer than 30 such drugs were on sale in the mid-1990s and more than 200 by 2008, by one 2008 history. From 2005, when the pharmacy benefit manager Medco bought Accredo, the firms that run health plans' drug benefits also owned the largest specialty pharmacies.

Where it stands

US pharmacies dispensed an estimated $293.4 billion of specialty drugs in 2025, and the three largest specialty pharmacies, all owned by companies that also run drug benefits, took two-thirds of that revenue, by Drug Channels Institute's count. IQVIA expects specialty medicines to make up about 43% of the world's medicine spending in 2028. Outside the US the same work often sits with hospital pharmacies, as in Spain, or with homecare companies working under hospital contracts, as in England, where more than 500,000 patients use them.

What is open

Much of the work still runs on fax, phone and plan-by-plan rules: referrals arrive as faxed packets, each drug plan writes its own approval criteria, and the standard pharmacies use to report to drug makers can be read only by members of the body that wrote it. The open jobs below are public tests and tables built from made-up patients and public documents: scored referral packets, plan criteria side by side, safety-program duties coded for pharmacies, a map of which pharmacies may fill each drug, and copay program terms in one place. None of them dispenses a drug or advises a patient.

Domains

AchievedWatch forOpen

Advances to build with

1 research note

Open in specialty pharmacy

3 of 12 open deliverables

Open · Taking in referrals

Made-up referral packets, scored

A public set of 300 invented specialty pharmacy referral packets, each a fax-like PDF of 5 to 15 pages: prescription, drug maker's enrollment form, insurance card, clinic notes, lab results and consent page, with smudges, skew and handwriting. Each comes with an answer key of every field in the HL7 Specialty Medication Enrollment format, and a scoreboard of how well extraction tools fill it.

Built with what exists

Synthea, an open-source generator of made-up patients; Python with ReportLab and Faker to lay out pages; image noise scripts to imitate fax; the HL7 Specialty Medication Enrollment profiles as the field list; OCR such as Tesseract or docTR; vision-language models, including small open models fine-tuned on one graphics chip (see the Note "One GPU to fine-tune").

Credit · unclaimed

Open · Approval from the drug plan

Same drug, many plan rules

A public, machine-readable table of the prior authorization and step therapy criteria that every Medicare Part D plan sponsor posts for 50 high-spending specialty drugs, for plan year 2027. Each criterion is coded into fixed fields (diagnosis, prescriber specialty, drugs to try first, tests required, length of approval, renewal rules), linked to the page it came from, and marked where it goes beyond the FDA label. Repeat each October for the next plan year.

Built with what exists

CMS's Part D formulary files; each sponsor's posted criteria documents; a web agent such as Browser Use to find and download each sponsor's criteria file (see the Note "Deleting 80% to start over"); a low-cost AI model for repeated extraction, such as Meta's Muse Ember or Claude Haiku 5.5 (see the Note "The jobs agents repeat"), drafting coded rows for a pharmacist to check; FDA labels through openFDA; a CSV file in a public GitHub repository.

Credit · unclaimed

Open · Shipping cold drugs

Weather against shipper profiles

A public count of how often real weather on pharmacy delivery routes goes beyond the summer and winter test profiles that insulated shippers are qualified against. For 50 routes in the US and Europe (a pharmacy city to the places it ships), take hourly air temperatures for 2015 to 2025 from public weather records, build every possible 72-hour and 144-hour delivery window, and count by route and month the hours above the hottest or below the coldest point of the ISTA 7E heat and cold profiles. Publish the counts and the code, not the profiles, and state plainly that outdoor air is not the air inside a truck or on a porch.

Built with what exists

Hourly station data from NOAA's Integrated Surface Database and European national weather services; a purchased copy of ISTA 7E; Python with pandas; a public code repository.

Credit · unclaimed

Where it sits

Broader

Pharmacy