Two Prescribers in the Same Practice, Converting Patients at Very Different Rates

Conversion reporting normally works at the level of the site or the dispensing pharmacy. Both views average together every prescriber writing inside a practice, so a writer who struggles to get patients through prior authorization is covered by colleagues who do it well. We benchmarked at the writer level, comparing prescribers who share an account, a payer mix, and similar patient profiles.

In 42% of accounts with more than one writer, colleagues at the same site differ by 20 percentage points or more in how many of their patients ever reach a paid fill. These are not thin accounts. Fifty-one of them differ by 40 to 60 points and cover 1,632 patients, and another 24 differ by 60 to 90 points across 1,101 patients. The losses do not happen at the counter. They happen after a rejection, at the point where someone has to work the claim back through to approval. On a specialty product of this kind that work usually succeeds: among patients already established on therapy, the vast majority of rejections resolve to a paid fill in a median of five days.

The same patients convert under someone else

We took 812 patients whose first attempt was rejected and who never reached a paid fill under that prescriber. Within twelve months, 207 of them tried again under a different prescriber, and 133 of those reached a paid fill. Two out of three patients who never converted for one writer converted under another.

The second prescribers under whom these patients converted were, as a group, higher converters across their own patient pool. Time also passes between the two attempts, so some share of these conversions could reflect a plan or a benefit year changing rather than the second writer. Most of these patients remain in regular medical care after the failed attempt.

Where we could associate the organization for both prescribers, half the patients who moved to a second prescriber stayed inside the same practice. Those patients converted at 70%.

Why: conversion depends on follow-through, and low converters stop early.

When a patient's first claim is rejected, the two writers diverge within the first month. High converters submit again within 30 days for 80% of those patients. Low converters do it for 43%. Widening the window to 90 days barely moves either number, so a claim that is not followed up quickly is generally not followed up at all. This is measured on whether a second attempt happens, not on whether the patient was eventually paid, so it does not depend on observing the final outcome.

Getting a rejected patient to a paid fill takes repeated attempts. Among patients who were rejected and then recovered under a high converter, it often took more than 10 separate claim attempts.

Low converters show less follow-through after rejection. Among their patients who were rejected and never reached a paid fill, 69% have one observed claim attempt and no subsequent attempt. Because specialty pharmacies typically rerun a claim once coverage is resolved, the absence of a second observed claim is consistent with an unresolved coverage process.

Restricted to writer pairs where both prescribers fill the majority of their patients at the same pharmacy, the gap persists, so this is not an artifact of where the two writers send their scripts.

This is not explained by harder rejections or by less prepared patients. Coverage-related rejections make up a similar share of both groups' rejections. Their patients are also comparable at baseline. We checked what exists in the record before the first attempt, including a recent echocardiogram and documented prior therapy on a beta blocker or calcium channel blocker, and found no meaningful difference between the two groups. The two groups see similar patient profiles and resolve a different number of them.

Across these pairs, we counted each prescriber's patients who have HCM or present like it. The high converter sees about 60 such patients. The low-converting colleague sees roughly twice as many, but the share with a coded HCM diagnosis is about four times as high in the high converter's panel. Same building, same payers, but for one of them this is a regular part of the week and for the other it is an occasional patient inside a much larger panel. Few practices build a durable authorization workflow for a drug they write twice a year.

The pattern follows them well beyond HCM. Across the rest of their prescribing, on therapies unrelated to this disease, patients who hit a rejection under a high converter reach a paid fill 88% of the time against 76% under a low converter, and low converters are about twice as likely to have a patient with one claim and nothing after it.

A real-life account example

Two prescribers in the same, large cardiology practice, 54 patients between them. The first has 29 patients and gets 96.6% of them to a paid fill, with rejections resolving in a median of four days and every patient who starts on a regular fill cycle going on to a second one. The second has 25 patients and converts 56%. Four in ten of their patients who hit a rejection never work back to a paid fill. Nineteen patients appear under both prescribers. In eighteen of them the low converter wrote first and the high converter picked the patient up afterward. All nineteen were eventually paid.

They are matched on the factors that would normally explain this. Both books are commercial-weighted, 62% and 68%. A similar share of each book runs through the same pharmacy benefit manager. Both route to the same dominant dispensing pharmacy, which rules out the possibility that the gap is an artifact of where their patients fill. Both submit prior authorizations. They share a meaningful number of patients on the medical side but rarely bill the same patient on the same day, so these are colleagues with overlapping panels rather than a physician and their nurse practitioner (further comfirmed by specialty codes).

Their panels differ in one way that matters, and it points at the same conclusion. The low converter's panel is roughly two-thirds new starts, and new starts convert worse everywhere. But within their own panel, their continuing patients almost always recover from a rejection and their new starts recover only 40% of the time. No patient on a second fill or regular fill cycle has stopped filling, so their patients are not abandoning therapy. They are failing to get through the first approval.

What this is worth and who owns it

Across three dozen accounts with a 20-point internal gap, roughly 190 patients sit without an observed paid claim under the low-converting writer, about 106 commercial and 71 Medicare. Free goods and bridge dispensing produce no paid claim, so some of these are patients on assistance rather than patients lost. We checked for the claim activity a bridge would generate and found it in only a small minority. Often the intervention is making sure someone in the office closes out the coverage question, rather than anything involving a specific payer, and in half of these accounts the person who already does that works in the same practice.

What else the quarter covered

The same scan tested whether a practice's competitive behavior on new patients could be predicted from what the account looks like. It cannot. More than a dozen structural and operational variables were tested against matched accounts, including panel size, academic versus community setting, payer mix, health system affiliation, dispensing channel, prescriber role, and how heavily a practice images and works up its patients. None separated the accounts that shifted from the ones that did not. Two comparable flagship programs in different health systems sit at opposite ends of the range, and the one with the larger program is the one that shifted least.