Pandemic Saturday...time for S-1 club!

This week's entrant: @PriviaHealth

The prospectus draws some parallels to Oak & Agilon, but I read it more as a look under the hood of a traditional fee-for-service business, with a bit of (apparently successful) VBC layered on top. 1/n

Privia dates to 2007, but the real push started in 2014, with a $400M investment round led by a Goldman Sachs vehicle (Brighton Health) and Pamplona Capital. They've since expanded to 6 states and rolled up >2,500 "providers."*

*Defined in S-1 to incl MDs, DOs, NPs & PAs.

Structurally, Privia owns some medical groups outright, where that is allowed under state law, and operates as a management services organization (MSO) in others, collecting admin fees and (it seems) operating some ACO vehicles in which those affiliated groups participate.

That introduces some fuzziness in the P&L, but the core picture, across all owned & affiliated groups, seems to be:
- 2,550 providers
- Roughly 3 million total patients
- $1.3 billion in aggregate practice collections/revenue
- $27 million of EBITDA for Privia

The business is 85 – 90% fee-for-service, with a bit of recurring PMPM revenue ($27M in 2020) and shared savings.

Most of the latter appears to come from Privia’s participation in the Medicare Shared Savings and CPC+ programs, plus some bonuses from commercial ACO programs.

In general, this strikes me as likely a pretty typical picture for large-ish primary care groups, outside of the more focused value-based / total cost of care models operated by Oak, ChenMed or Cano Health.

The latter also means a tighter patient focus on patients 65+.

Privia’s revenue mix shakes out to $500 – 550K per provider, of which $200 – 250K prob ends up as provider compensation.

As an aside, this doesn't exactly suggest that traditional medical groups are “struggling”

An SMB where 40-50% of revenue is left for principals/owners?

The patient panel deserves a deeper dive. In short, Privia says their providers treated ~3M total patients in 2020, with ~682K “attributed lives” (subset of the 3M).

Those attributed lives are ~61% commercial (presumably commercial HMO plans), ~35% Medicare and 4% Medicaid.

In turn, the ~230K Medicare-eligible attributed lives are about 2/3 traditional Medicare enrollees & 1/3 Medicare Advantage.

Not too far off national averages, though the overall picture means that Privia has much less exposure to potential upside via MA or other TCOC models.

Triangulating a bit, I'd guess that Privia’s practices generate $300 – 400 annually per patient in FFS revenue (about half of One Medical’s revenue/mbr).

Then, Privia adds $100 – 150/patient for the quarter-ish of those patients who are attributed to a P4P contract of some sort.

The growth picture is mixed. Attributed lives & practice collections both grew by ~22% in 2019, but lives then dipped by 3% in 2020. This contraction is not well explained in the S-1, apart from a mention that “attrib lives in…commercial value-based programs decreased by 13.7%”

For context, in any case, Oak and Agilon () both saw revenue grow by >50% in 2020, while patients under management grew 22% and 45%, respectively.

Privia's GAAP revenue was up just 4%, while the attributed patient panel shrank a bit.

Pandemic disruptions may have dented revenue a bit (Privia’s FFS revenue from patient care was down by 4% YoY), but it may also just track the apparent loss of a block of commercial patients (or a commercial contract) at some point last year. Unclear.

Beyond that, I’d quibble with inflated TAM estimates (# of primary care physicians, a supposed *$1.9 trillion* mkt for “physician enablement,” etc), and the attempt to redefine EBITDA margins by first constructing a non-GAAP EBITDA measure and then dividing by gross profits.

That said, I do think "clinician enablement" is a reasonable way to think about this market, even if Privia is over-egging the pudding.

Most of these tech-enabled care delivery orgs launched in 2013/14; the IPO/SPAC filings shed light on different distribution & revenue models.

Finally, an interesting bit here for virtual care-istas.

Privia traces the virtual curve starting at 0.3% of pre-COVID visits, then jumping to 45% of visits in April 2020, before settling at 15-20% by end of year, where they expect it to stay.

Necessity...invention etc. /n