Because paying for primary care with insurance incurs administrative costs not encountered in direct pay models, a case can be made that direct primary care should cost a patient less than insured primary care. But most DPC advocates are themselves PCPs and they have little to gain from offering discount pricing and much to gain by offering a subscription service featuring deluxe, small patient panel service at a premium price. For direct primary care physicians as a group, the preferred way to square the value circle in patients’ eyes is with the proposition that premium care results in lowering costs for downstream care by more than enough to offset the premium price.
Health Care Cost Institute national claims data for 2017, grossed up to account for administrative costs and profits to the maximum allowed for by MLR rules, indicate that a middle-aged insured effectively pays under $48 a month for primary care. But subscription DPC services are typically priced at $75 a month.
A few DPC advocates apparently to have an agenda at least slightly different than that of the collective of all D-PCPs . One such advocate is Gayle Brekke, whose recent e-book, Paying for Primary Healthcare, was supported by the Free Market Medical Association and the Mises Institute. Brekke precedes the gravamen of her ebook with a humble brag of her credentials as an actuary and an academic. Then, before proceeding with her specific defense of subscription based small panel direct practice, Brekke lays out a broader argument purporting to show that using insurance, rather than direct patient to doctor payment, increases the cost of primary care by more than 50%. Per Brekke’s conclusion, then, a middle aged person could have her $48 per month primary care services provided in a direct arrangement for less than $32.
In prior blog posts — here, here, here, here and here — I addressed at (undoubtedly-too-much) length, many other issues raised by Brekke’s analysis as she originally presented it in blog posts on her own website. A separate post summarize Brekke’s argument and my chief responses.
By far the largest single factor in Brekke’s quantitive analysis, by itself accounting for nearly two-third of the 50% savings figure at which she arrived, was her estimate of a 25 to 35% range for the overhead of direct primary care practices. In this post, I reprise and refine a previous analysis of whether the sources from which Brekke has drawn that estimate offer meaningful support for the estimate she makes.
The mathematical core of Brekke’s argument
[T]he overhead for a typical traditional practice is roughly twice the overhead of a typical direct practice.5
5 We use a range of 25-35%, as little information about overhead in a DPC or DPC-like practice is available.
See Bujold and Forest:
Bujold, E., The Impending Death of the Patient-Centered Medical Home. JAMA Intern Med, 2017. 177(11):
p. 1559-1560.G, Brekke, Paying for Primary Healthcare
Forrest, B.R., Breaking even on 4 visits per day. Fam Pract Manag, 2007. 14(6): p. 19-24.
Bujold
Dr Bujold’s “DPC-like practice” actually reported 65% overhead in the very article to which Brekke refers. This is most clearly not evidence of 25-35% overhead for DPC practices. What it does evidence is that Ms Brekke, while offering herself as having significant academic credentials, handles source materials — even those of her own selection — with less than reasonable care.
(Bujold + Forrest) – Bujold = Forrest
Taking away the Bujold piece leaves Brekke with only Dr. Forrest’s article to support the idea that typical DPC practices have overhead of 25%-35%. Reliance on Forrest’s piece, however, further confirms Brekke’s carelessness. For, while the two largest items of overhead cost for typical medical practices are payroll and the cost of medical office space, Dr Forrest’s unique accounting methodology openly excludes the largest component of his payroll and actually treats office space as an income item rather than a cost item.
Forrest on payroll:
… I recently added a nurse practitioner, who sees most of our walk-in patients with minor acute care needs. She also helps manage patients with chronic conditions like hypertension, hyperlipidemia and diabetes. …
Our overhead has been consistent at 25 percent of total revenue. … Our overhead figure does not include the employment costs associated with our nurse practitioner. These are paid out of the revenue she generates by seeing patients that I would normally not have the time to see.
Brian R. Forrest, MD, Breaking Even on Four Visits Per Day
Dr Forrest’s rationale is nonsensical. “These are paid out of the revenue she generates by seeing patients that I would normally not have the time to see.” The effect of Forrest’s math is to sequester an important, skilled-labor-intensive, high overhead component of his enterprise, then report the remainder of the business as having low overhead. Beginning NP salaries when Forrest’s piece was published were around $50,000 and then moved went well-upward; Forrest reported that his NP earned an income above the average.
A normal medical practice, DPC or otherwise, would consider the costs of each of its employees as part of overhead. And, no matter how Forrest computes his “overhead figure”, he surely lists the amount paid to his nurse practitioner on his practice’s tax return.
Bonus note. At times, small enterprises like DPC practices hire temporary workers to, for example, cover absences due to illness or to staff turnover; overhead for normal firms includes payment for the value thus received. But Forrest is different; rather than employ and pay “temps” Forrest relies regularly on labor donated by “many well-trained volunteers” – though the amount of these donated services has not been qualified or valued.
Forrest: my medical office space is even cheaper than free!
Dr Forrest’s practice occupies 2200 square feet and, to rephrase a popular meme, he lives there rent-free — inside his own head. Dr Forrest’s overhead computation shows this.
Dr Forrest reports owning 4400 square feet in an office complex and he rents out half of it to other doctors. After paying the expenses for the “rented half”, the rent he receives still nets him enough, to both pay the costs of “his own” space and to provide him an additional $3600 annually. Presumably, the rental amount paid by his tenants is near the average annual rental cost of 2200 square feet of medical office space near Forrest’s location at the time of his piece. That would likely have been well in excess of $40,000.
His use of “his own” space in this situation, however, is not cost-free. It is merely cash-free. To occupy “his own” space, he must forego the opportunity to rent “his own” space, at fair market value, to yet someone else. Assuming “his own” space and the “rented half” are more or less equivalent, he could be renting the entire 4400 square feet at double the rent he is being paid for the “rented half”.
I do not know for certain whether Dr Forrest understands “opportunity cost”. But I am certain that Dr Forrest’s accountant understands that a doctor who uses space he “owns” rather than “rent” is allowed to use the depreciation, mortgage interest, real estate taxes and other costs of “his own” space to offset the revenues from his practice.
Without the bizarre accounting, Forrest’s “overhead figure” more than doubles.
A typical DPC practice would arrive at overhead costs using normal accounting principles. That would mean including fair costs for his medical office space, and fair pay for his nurse practitioner.
Add a conservative estimate of $50,000 for the nurse practitioner to a conservative estimate of $40,000 for a 2200 square foot rental. Add this to Forrest’s $78,400, and back out that $3600 profit from Forrest’s real estate side hustle, and a realistic figure for Forrest’s overhead is $164,800. That’s actually more than double the Forrest-computed figure on which Brekke relied — to estimate 50% savings.
Additional bonus notes. Forrest’s budget includes no amortization or depreciation for medical equipment, albeit that cost item might be small because Forrest purchases used equipment. In a similar vein, Forrest recounts multiple other economies, such as foregoing janitorial service (he takes out his own trash); he sets the thermostat to lower the heat when the office is empty. Importantly, many of Forrest’s budget control practices would be available to any PCP, whether they accept insurance or not.
Forrest prides himself on being both an abstainer from insurance and a skinflint; his article gives us no specific information about how to apportion purported overhead savings between those two aspects of his unique persona.***
Critically read, Forrest’s article joins Bujold’s in doing substantially nothing to confirm Brekke’s claim that “the overhead for a typical traditional practice is roughly twice the overhead of a typical direct practice.”
Brekke wrote, “little information about overhead in a DPC or DPC-like practice is available.” Brekke might have looked more closely, before offering an estimate that is not meaningfully supported by the only material she can identify .
I do not deny that provider-side administrative costs are somewhat lower when a patient pays the provider directly for primary care. But the systemic consequences of that depend heavily on the exact magnitude of that effect. It is common for DPC advocates to suggest that provider-side administrative costs savings alone can support three-fold longer visit lengths patient and three-fold smaller patent panels. Brekke’s own effort to support similar claims is worse than guesswork.
*** Despite his longtime DPC advocacy, Dr Forrest’s work has yet to receive proper respect. For example, even despite his having publicized its astonishing findings, many do not recall that Forrest directed a team of NCSU graduate business students in canvassing primary care offices throughout the Raleigh area to observe directly the amount of time FFS-PCPs were actually spending with their patients. Even the students on Forrest’s NCSU team have no recollection that this field work ever happened.
DPC Referee 