The Next Buy-and-Bill Writer Is Already in the Building

New-writer reporting treats the prescriber as the unit. A prescriber writes a first script, gets counted, gets deciled, and either builds a panel or doesn't. The office they work in is background. We turned that around. We sorted new writers by how they came into the class, through a prescription filled at a pharmacy or through a dose given in the office, and then asked what predicted whether they kept writing: the prescriber, or the office.

About 62,000 prescribers wrote their first observed PCSK9 script through the pharmacy between March 2024 and August 2025. About 900 entered by administering the therapy in their office instead. Counting only the patients each writer started rather than took over from another prescriber, one in seventy of the pharmacy writers reached ten patients in their first year. One in six of the buy-and-bill writers did. And four out of five of those buy-and-bill writers started inside an office where a colleague was already administering. The ones who started it alone did about half as well.

Prescribers who changed offices kept administering only where the new office already did

We found 295 prescribers who administered the therapy in 2024 and then moved to a different office. When the new office was already administering, nine in ten of them kept doing it. When the new office had never administered, one in ten did.

These are the same prescribers. Every one of them had administered the year before. What decided whether they continued was not anything about them. It was whether the office they walked into already had the scheduling, the inventory, and the billing set up. A few of the 295 are prescribers whose work shifted between two offices they already used rather than a full move, and the result reads the same either way.

Why: a prescriber joining an office that already administers sees patients on day one. One starting alone has to build the office first.

Joining an administering office about doubles a new writer's first-year panel. Prescribers who joined an office already administering saw eleven patients in their first year. Prescribers who started it alone saw about five. Some of that is existing patients being handed over, because at a large office a new provider ends up giving doses to patients who were already on the therapy. So we also counted only the patients each prescriber started themselves. The joiner still sees about seven new patients in the first year. The lone starter sees about four. The gap gets smaller. It doesn't go away. And because medical-benefit claims arrive more slowly than pharmacy dispenses, the gap in starts may be larger than it looks here.

The setup works the day it exists. A prescriber who joined an office that had started administering the quarter before saw the same number of patients as one who joined an office two years in. How long the office had been buying and billing made no difference. The effect is binary.

Once an office starts, the provider who started it soon stops being the main source of patients. Across offices we tracked from their first dose, a year in, three-quarters of the office's new patients were coming through someone other than the first administering provider. Many offices never add a second administering provider at all. The ones that do outgrow the first one within the year. Prescribers who administer do so at nearly every office where it is already set up, and almost never start it where it isn't.

A third of the new administering providers in these offices are nurse practitioners and physician assistants, and a new-writer list built on physicians misses them. On patients they started themselves, APPs saw about eight in the first year against five for the physicians in the same offices. That count excludes follow-up doses, so it isn't one NP's name on every maintenance visit. And where an APP was the first to administer, with no setup to join, APPs and physicians look the same. The office is what makes the APP productive.

Offices bring new prescribers into the class. Hospitals add buy-and-bill to prescribers already in it.

About 900 offices and clinics administer the therapy, and about 730 hospital-affiliated sites do. Nearly every prescriber who came into the class through buy-and-bill did so at an office. The hospital sites, close to half of the places the drug is given, are not where new writers come from.

The difference is which claim came first. Hospital cardiologists who administer had usually written Repatha or Praluent before their first dose, 85% of them. They were in the class already and added administration as a second route. At offices, for most administering providers, a dose given in the office was their first PCSK9 claim of any kind.

Reimbursement doesn't account for it. Sites paid above 130% of ASP are mostly hospital outpatient, and there more providers administer but each sees about one new patient a year rather than two. Below that line, what a site was paid had no relationship to how deep its new providers went.

A real-life account example

One account shows the pattern end to end. A large multispecialty physician network in California was already administering when our data begins in August 2023, with fifteen providers doing so. By the first half of 2026 there were thirty-five, and they administered to 522 patients in that half-year, against 39 in the first. Most of the twenty who joined had been writing Repatha and added administration as a second route. A few had never touched the class. One physician started administering in July 2025, with 31 colleagues already doing so and no prior PCSK9 claim. Over the next twelve months that physician saw nine patients, eight of them new to the class, without ever writing a prescription for any therapy in it.

The pharmacy side of the network grew over the same period, from 535 Repatha patients in the first year we can see to 1,181 in the most recent, and its pharmacy-side new writers stall at the same rate as everyone else's. About three in ten of the network's PCSK9 patients get the drug in the office, a high share for this class.

What this is worth and who owns it

Anyone who has launched a buy-and-bill product knows the account matters more than the individual. What the data adds is how completely, and that the next prescriber to start can be picked out of claims a year ahead. Buy-and-bill is a small channel here, about 6% of new-writer patient volume, so this isn't a growth story. It is an account of where a small channel's next writers come from. The usual search looks for a champion who will bring administration into their practice. In this data the practice comes first and the champion follows. The prescribers who do start administration alone don't look like champions either: most are the only administering provider in their office, and they are less likely than joiners to have infused other therapies or to be cardiologists.

The next adopter can be seen a year ahead. Two things about a provider show up in claims: whether they already give infusions of other therapies, and whether they work in an office that already administers this one. Providers with both started administering the following year at one in 29. Everyone else started at roughly one in several thousand. That is what happens on its own, with no one targeting them. The account is already on the call plan. The name that matters next is not, because that provider has never written the drug and appears on no new-writer report. In this class the next administering provider is usually already inside a covered account, an infusion nurse or a physician down the hall, and there are fewer than two thousand of them.

For a pharmacy-benefit brand, administering offices are not a separate account list, and administration did not come at the pharmacy side's expense. Six in ten offices that administer also write pharmacy-benefit PCSK9. At the largest of them, Repatha patients went from about 120 in the six months before the office started administering to about 210 in the twelve months after; comparable offices that didn't start administering added about 40 over the same stretch. Praluent didn't move either way. Whether administration draws patients in or growing lipid practices adopt everything at once, the pharmacy side held at these accounts.

Two limits. Patients seen by joiners and patients seen by lone starters got the same number of doses in their first year, but the joiners' patients got the second dose on schedule less often, 27% against 41%. The big office sees more patients on a slower calendar. And most offices that administer never grow past one provider. About one in five add providers over two years, and they are where the joiners come from.

What else the quarter covered

The same scan asked what predicts anything for the 62,000 pharmacy-side new writers, half of whom never write a second script. Specialty, decile, and the number of prescribers a patient sees barely moved it. What did was whether the practice could get a rejected claim through. In small practices, new writers whose group was good at recovering rejections reached three patients at 24%, against 15% where the group wasn't. In large practices the gap almost vanished. Small offices live on process; large ones absorb it. A first patient whose claim never cleared left the writer at half the odds of reaching three, and it made no difference how many attempts the claim took, only whether it eventually went through.

Site economics predicted none of this. Negotiated rates, 340B status, and GPO affiliation had no relationship with how a site routed patients, how well they completed doses, or whether its providers adopted. Device presentation predicted patients switching from syringe to pen, not stopping.