This is a complete phase-one report for a five-site dermatology group. It is eleven pages of findings, evidence and sizing — and the practice it describes has never existed for a single day.
Northwood Dermatology Partners is invented. So are its five locations, its eleven clinicians, its payers, and every number below. We built it specifically to be shown.
The reason is the same promise you would be buying. We cannot show you a real client's report — it is not ours to show — and when the report is yours, it will not be ours to show either. No client of ours appears in a case study, a pitch deck, a conference talk or a website, named or thinly disguised. That is not a policy we can suspend for a good prospect, which means a specimen is the only honest way to answer “what do I actually get?”
So: the company is fiction. The structure, the vocabulary, the evidence standard, the way a finding is sized, and above all what we refuse to conclude — those are exactly what arrives.
Is there enough here to justify a full review? Not “is anything wrong” — something is always wrong. The question is whether what is wrong is concentrated enough, and large enough, to be worth four more weeks of anyone's attention.
Short answer for Northwood: yes, but narrowly. Four of the six findings below sit in one place — the gap between what a drug costs the practice and what the practice is paid for it. That concentration is the argument for phase two. Had it been six unrelated findings of this size, our recommendation would have been to fix them and not commission us again.
Seven sources. Two of them — the distributor invoices and the point-of-sale export — had never been read against the clinical data, because nothing joins them automatically. Four of the six findings are in that join.
Sized to an order of magnitude, annualized, and stated as a range. A diagnostic that quotes a recovery to the dollar is quoting a number it cannot have.
Northwood's five sites joined at different times, and four of them kept the payer contracts they arrived with. For the group's twelve highest-volume CPT codes, the allowed amount from the same payer differs by site — in one case by 31% for an identical procedure performed by clinicians who now share a compensation pool.
This is not a negotiation failure. Nobody ever had all five contracts open at once.
Northwood purchases, holds and administers biologic and injectable product worth $6.1M a year, and is reimbursed against a published payment formula. The spread between acquisition and allowable is positive on most product and materially negative on four items, which between them account for roughly 9% of drug volume. The practice carries the inventory risk in between on its own balance sheet.
Nobody is doing anything wrong here. The purchasing decision and the reimbursement consequence are simply made by different people, months apart, in systems that do not speak.
Where a single-dose vial is larger than the dose administered, the remainder is reimbursable when billed with the JW modifier and documented. Across 24 months the modifier appears on 11% of eligible administrations at one site and on essentially none at the other four.
The constraint is documentation, not billing. Phase two would test whether the documentation to support the modifier exists before recommending it be claimed — billing it without support is a worse outcome than not billing it.
Northwood reads its own dermatopathology at two sites. Claims are submitted globally. For two payers representing 23% of that volume, the published rules for the place of service recorded require the technical and professional components to be split, and the global submissions are being reduced on adjudication rather than denied — which is why this has never appeared in the denial rate anybody watches.
Cosmeceutical, injectable and optical stock is recorded when it sells and when it is ordered, and never counted in between. Shrinkage, expiry, staff purchase and a genuinely slow month are therefore indistinguishable from one another in the accounts. Gross margin on this line is reported at 61% and is not calculable from the available data.
We are not saying anything is being taken. The most common cause of a gap this shape is expiry on slow-moving stock. It is stated here because an uncounted $2.4M is a control weakness whatever its cause, and because you cannot rule anything in or out until it is counted.
Provider compensation is calculated from a productivity report that does not separate ancillary work. Two clinicians ranked in the bottom third of the group carry the highest proportion of pathology and dispensary-attached encounters, and contribute more margin than three clinicians ranked above them.
This one recovers nothing. It is here because a compensation formula that mismeasures contribution eventually costs a group the wrong clinician, and that is more expensive than everything else on this page. It is also the finding most likely to be resisted, which is worth knowing before it is raised.
| Finding | Low | High | Confidence |
|---|---|---|---|
| 01 Payer contracting spread and underpayment | $310,000 | $520,000 | Firm |
| 02 Buy-and-bill acquisition against allowable | $240,000 | $390,000 | Firm |
| 03 Discarded drug not claimed | $85,000 | $140,000 | Probable |
| 04 Pathology component split | $190,000 | $300,000 | Firm |
| 05 Uncounted retail and dispensary stock | $120,000 | $260,000 | Indicative |
| 06 Compensation mismeasurement | — | — | Firm |
| Indicative annual range | $945,000 | $1,610,000 |
Against $41.2M of collections, that is between 2.3% and 3.9%. If a report tells you it has found 20% of your revenue, read it very carefully.
A diagnostic that overstates its own authority is worse than no diagnostic. So, explicitly:
The same report, three audiences. A document that only works for one of them stalls in the room where it is presented.
Practice administrator or CEO, with the managing physician on clinical scope. They read the question, the summary table and the refusals — roughly two pages.
Physician owners whose compensation formula finding 06 decomposes. Worth deciding how that is raised before the meeting, not during it.
The revenue cycle lead, who can usually name the problem sites before we open anything. Findings 01 and 04 will not surprise them; the evidence for them is what they have lacked.
Same structure, your registers, about two weeks. If the answer is that nothing is concentrated anywhere interesting, the report says so and we do not propose a review — a worse outcome for us and a perfectly good one for you, and the reason this is worth commissioning rather than taking our word for it.
And it stays yours. Your report is never shown to anyone, in any form, named or disguised — which is why the one above had to be invented.