Open Health Data Hub



All posts
|3 min read

The Prescribers Writing $1 Million in Drug Claims Each Year

top Medicare Part D prescribershigh-cost drug prescribersMedicare prescriber outliersprescription drug spending concentration

4.55% of Medicare Part D prescribers now control more than half the program's entire drug bill.

That's the arithmetic behind 2023's Part D data: 50,260 prescribers, each billing over $1 million in annual drug costs, collectively drove $107 billion of the program's $212.7 billion total spend. Their share: 50.31%. A decade ago, that same group (then just 9,276 prescribers) accounted for 18.03% of a much smaller total. The concentration has not crept upward. It has restructured the program.

From a Rounding Error to Half the Budget

In 2013, million-dollar prescribers were a statistical footnote: 1.15% of all prescribers, responsible for roughly one dollar in six. By 2023, they're 4.55% of prescribers and responsible for one dollar in two. The count grew from 9,276 to 50,260 over ten years, a 442% increase, while total prescribers grew only 37%.

Every year in between shows the same direction. The share of Part D spending from this group rose from 18.03% in 2013 to 24.50% in 2014, crossed 40% in 2020, and crossed 50% for the first time in 2023. No year reversed the trend.

For anyone modeling Medicare's long-term drug cost trajectory, this concentration is the central variable. Half the program's spending is now determined by the prescribing decisions of fewer than 51,000 clinicians out of 1.1 million.

The Drugs Driving the Concentration

Not all million-dollar prescribers are writing the same kinds of scripts. The top drugs by total cost within this group split into two distinct categories: high-volume maintenance drugs and low-volume specialty drugs with extreme per-claim costs.

Drug (Brand)Total CostClaimsCost per Claim
Apixaban (Eliquis)$5.54B6,682,811$828
Semaglutide (Ozempic)$3.37B2,393,561$1,409
Lenalidomide (Revlimid)$3.13B174,946$17,913
Ibrutinib (Imbruvica)$1.79B114,895$15,564

Apixaban leads by total cost at $5.54 billion across nearly 6.7 million claims, with a per-claim cost of $828. It's a widely prescribed anticoagulant, and its dominance reflects volume. Semaglutide ranks second at $3.37 billion, driven by 2.4 million claims at $1,409 each. These are drugs that reach millions of patients.

Lenalidomide tells a different story. At $17,913 per claim, it generated $3.13 billion from just 174,946 claims across 4,549 prescribers. Ibrutinib runs $15,564 per claim. These are oncology drugs, and their cost per claim is more than 20 times that of apixaban. A prescriber writing a modest number of lenalidomide scripts can cross the million-dollar threshold without being a high-volume prescriber at all.

Concentration Has a Compounding Effect

The practical consequence of this structure is that small changes in prescribing behavior among a narrow group of clinicians produce large swings in program costs. When a specialty drug's price rises, or when a new high-cost therapy enters the market, the impact flows disproportionately through this concentrated group.

Semaglutide's presence near the top of the list is a preview of what that dynamic looks like in real time. Its $3.37 billion in Part D costs within this group alone reflects a drug that barely registered in earlier years. As GLP-1 prescribing has grown, it has pulled more prescribers across the million-dollar threshold and added to the concentration simultaneously.

The question the data leaves open is whether the concentration itself is a problem or simply a structural feature of how specialty medicine works. Lenalidomide costs $17,913 per claim and is prescribed by only 4,549 high-spending prescribers. What share of those claims reflect appropriate oncology use versus potential overutilization or off-label prescribing is something the cost data alone cannot resolve, but it's the right question to ask when a single drug accounts for $3.13 billion in spending from fewer than 5,000 clinicians.

Explore the data yourself

Run your own queries against 240M+ rows of federal health data using natural language — powered by AI.

Start analyzing