Weekly Health Tech Reads | 12/10/23

CVS Health investor day, MA enrollment data, virtual doula research outcomes, and more!


Sharing our perspective on the news, opinions, and data that made us think the most this week


Similar to UHG's Investor Day session last week, CVS's Investor Day on Wednesday is worth listening to just for the sheer breadth of topics they cover in the three and a half hour session. The changes to CVS's drug pricing model made the biggest headlines as CVS seeks to bring more transparency to the pharmacy industry (link to WSJ coverage). CVS thinks its reached a floor on reimbursement rate erosion in the pharmacy space, and this makes it the right time to shift to a transparent model to drive change moving forward. It's interesting to see the analyst questions on this topic - they understandably seem very focused on understanding business model impacts how CVS will roll out these changes.

Another key theme of the session was the benefit CVS is already starting to see from integrating the clinical, retail, pharmacy, and insurance assets that CVS now has. There seems to be a heavy emphasis on driving customers between these assets, which has the benefit of increasing customer lifetime value by 3 - 4x according to the session. The Signify and Oak Street acquisitions in particular spent a lot of time talking about the value of the integrated model here and the opportunity to cross-pollinate. One example given of this was that Oak Street is setting up tables inside CVS to convert CVS customers into Oak Street patients. It seems to be working reasonably well as Oak Street is converting 17% of CVS customers they talk with into visits at Oak Street. Signify is spending a lot of time talking to Aetna members in home, and as part of that is attempting to drive those members to Oak Street clinics - of the members Signify talks to, 19% want to be connected to Oak Street, and then 55% of those members actually schedule a visit, for a ~10% conversion rate in total.

Ultimately this activity should pay organizational dividends by moving Oak Street clinics to profitability faster. Oak Street mentioned a new metric in the session seemed designed to show exactly this by suggesting that if they add one incremental VBC patient per day in a clinic, it can speed up the J-Curve's profitability path for that clinic by two years. Conceptually the math seems to make sense, but Oak Street shared very little data on how that is happening in centers today, which is a bit unlike how they've historically communicated center-related metrics to investors. Oak Street did share some helpful data on the profitability of clinics generally - in 2024, they expect 120 out of the 129 centers they opened prior to 2021 to be profitable. The 39 centers that are over six years old will have $6 -$7 million in profitability, and the mature centers (defined as 2,000+ patients) will have $8+ million in profitability, although curiously I don't think they shared a number of centers at maturity. Lastly on this front, it was alluded to a number of times during the session that CVS is exploring moving Oak Street more into commercial, ACA, and Medicaid risk, with work already underway managing high acuity members. It'll be worth keeping an eye on the transportability of this model to those other insurance types as its a pretty big strategic shift for Oak Street from its time as a standalone business.

The care delivery asset is growing rapidly inside CVS - from $6 billion of revenue in 2023 to $10 billion in 2024. Value-based care also came up a number of times during the session as a key driver of growth. CVS shared that it recognizes it might be behind some of its competition today, but that it sees an opportunity to catch up and surpass anyone else in the space given the diversity of assets. CVS isn't that far behind UHG in terms of providers aligned to it - citing it will have 75,000 providers on its platform in 2024, with Oak Street and MinuteClinic employing 4,000 of those providers. CVS spent a good deal of time highlighting the success it has had building out its ACO model, with one million lives under management and ~60,000 providers aligned in 2024. This ACO business is largely a combination of CVS's ACO REACH business and Signify's Caravan Health business. An analyst asked about whether CVS will need to acquire any capabilities for VBC moving forward, and the answer was no, but it might choose to opportunistically pursue acquisitions like specialty assets.

Speaking of that, we also heard a bit of conversation about the value of local market density in terms of care delivery asset. Chicago was given as the example where CVS Health is building this density - slide 104 (link) had a nice heat map of Chicago. In Chicago, CVS has 28 Oak Street clinics, 120 MinuteClinics, an ACO with 1,500 provider partners, and Signify doing 32,000 home visits. With that density it is serving 290,000 patients in Chicago today. Chicago's population is roughly 2.7 million people according to the census, so by my math CVS Health's most densely penetrated care delivery market in the country has ~11% market share of patients today. It's interesting to think about how much opportunity there is in front of CVS to grow that number and what it might max out at in a given market.

All in all it was an interesting session, very focused on highlighting how if CVS can use its massive footprint to drive members into more offerings, it benefits the organization tremendously. It's a logical thesis, and a unique opportunity for CVS given the breadth of assets it has across retail, pharmacy, insurance, and care delivery.


Sharing a visual or two from the week that made us think

In a recent post on X, Mario Schlosser highlights some interesting data related to Medicare Advantage enrollment, unpacking how population growth might have been helping MA enrollments in recent years, a trend that will flatten in coming years.

Mario Schlosser's Reaction:

Mario Schlosser is the Co-Founder, President of Technology & Chief Technology Officer of Oscar Health and previously served as CEO of Oscar. Prior to co-founding Oscar, Mario also co-founded the largest social gaming company in Latin America, where he led the company's analytics and game design practices. He shares his reactions to the MA enrollment data below:

A strong case can be made that in incumbent-dominated markets (which healthcare certainly is), you need some type of structural or regulatory disruption in order for startups to break in. In health insurer land, Centene rode up the growth of managed Medicaid, Humana the growth of MA, Oscar the growth of the ACA (that story is still being written!), and so on. But it is also interesting to try to see when the winds might shift. Take Medicare Advantage, a massive growth driver of anyone doing anything in healthcare over the past 10 years. First, I like to joke that the best illustration of how little we understood about healthcare when we started Oscar was that we entered the ACA in 2013, and not Medicare Advantage. Because at that time, MA was just about to embark on its growth trajectory, partly because the US was about to enter a phase where the number of those turning 65 was going to double every year. But I do distinctly remember that it wasn’t as obvious a coming juggernaut as it looks like in hindsight. Partly because the market had actually been broken until a few years prior: exactly like the ACA, the early MA market shrunk for more than half a decade, until 2007 or so [see chart above]. So this was a still relatively small market, and merely on the mend from earlier instability. (That is also the story of the ACA.) My other guess is what also wasn’t clear at the time was just how profitable risk-adjusting against the government would be. (Unlike in the ACA, where it is zero-sum.)

Importantly, we might be entering another phase. The number of people turning 65 is now going to be flat or down for the next two decades, so that second-derivative tailwind is gone. The market competitiveness exploded around 2019, when the number of MA plans per beneficiary doubled. (Great lesson: US healthcare needs 10 years after a market turns obviously very profitable to finally figure it out broadly.) The government is stepping down on risk adjustment. It’s still a growing and profitable market, with plenty of opportunities. But for startups, I think the opportunity is to seek out micro-disruptions within MA: for example, who should be making money in it but isn’t yet? (Health systems, which are also leaving MA networks.) And look for the next MA market, because it will take incumbents 10 years to figure that one out.

A study on the impact of a virtual doula model

Population trends and Medicare Advantage growth. Forest plot of doula utilization with odds of cesarean birth (N=8,989). Model 1 adjusts for age, race, Social Vulnerability Index score, presence of medical conditions, body mass index, history of anxiety or depression, current pregnancy-related anxiety, and parity. Model 2 adjusts for all covariates present in model 1, in addition to overall Maven utilization (in minutes). Black vertical line indicates the null value of the odds ratio (1.00). All horizontal lines not crossing the null value indicate statistical significance.

A recent publication in Obstetrics & Gynecology looked at the association between the use of a virtual doula via Maven Clinic's digital platform and birth outcomes. The study looked ~9,000 users on ~9,000 15.5% of whom attended one doula appointment, and 8% attended two or more. Similar to more traditional models, the study showed a positive impact on birthing outcomes using a virtual doula (this is a nice summary of the impacts from Maven's CMO, Neel Shah).

Completing at least two appointments with a virtual doula was associated with a reduction in C-sections among all users, and was particularly meaningful for Black users. As I mentioned in Slack this week, doula models have always been one of the clearest examples for me of how healthcare gets so backwards. With doulas, you have a model that seems to pretty clearly improve birth outcomes for the patient, while reducing costs, and yet doulas are fighting to be paid living wages (i.e. see this article about doula payment rates for Medi-Cal).


A round-up of other newsworthy items we noticed during the week

MacKenzie Scott this week shared the list of non-profit organizations that she donated over $2 billion to in 2023. Among them include 210 healthcare organizations globally, 179 of which are in the US. She donated more than $518 million dollars to these 179 organizations in 2023. Among other things, Scott made a number of multi-million dollar donations to health centers across the US - i.e. she donated $6 million to Muskingum Valley Health Centers in Ohio. The single biggest healthcare donation Scott made was giving $25 million to Upstream USA, which works with health centers to expand access to contraceptives and reduce unplanned pregnancies. It's a good reminder that for as much as we talk about startup funding rounds in spaces like this, there are other avenues to making a meaningful difference on healthcare outcomes in this country.
Link / Slack

The WSJ wrote an interesting article suggesting the Medicare Advantage market might be cooling off, pointing at two key data points: UnitedHealth's 5% growth target for next year (it grew 11% last year) and Humana's willingness to discuss a merger of equals with Cigna. Seems to align from a thesis perspective with the tweet / data shared by Mario Schlosser above.
Link / Slack (h/t Michael Ceballos)

R1 RCM is set to acquire Acclara, a rev cycle management company from Providence (the health system), for $675 million in cash, plus warrants for 12.2 million shares of R1 stock priced at $10.52. The deal includes a 10-year agreement for Providence to use the rev cycle services of the newly formed entity. Acclara is expected to contribute $625 million in revenue and $185 million in Adjusted EBITDA to R1 by year 5 of the deal. The press release notes this was the first strategic cross-sell for R1 RCM with the Cloudmed customer base, which R1 RCM acquired last year (link). The press release also calls out that 95 of the top 100 US health systems are Cloudmed customers, suggesting a future growth opportunity for R1 RCM here.
Link / Slack (h/t Rohan D'Souza)

CB Insights released its list of the top 50 most promising digital health startups in 2023 based on its database. It interesting to look at the names of companies, but also the clustering of spaces that companies are in (and aren't in) - there's a heavy presence of diagnosis & imaging companies and care management companies.
Link / Slack (h/t Karra Hendrix)

Mosie Baby received FDA clearance for at-home intravaginal insemination, making it the first and only device to allow for this at-home procedure. The kit, priced at $129, serves as an alternative to traditional artificial insemination procedures, which often are more expensive and not covered by insurance.
Link / Slack (h/t Jaime Rooney)

Rural hospitals are increasingly losing out on new residency slots funding from Medicare, and as this Axios article discusses it appears to be worsening the provider shortage issues in rural areas.
Link / Slack


A collection of notable startup financing rounds across the industry

SCAN Health Plan, a nonprofit Medicare Advantage plan invested in Abridge, an AI clinical workflow tool aimed at automating medical documentation. SCAN has offered Abridge's products to select members since 2021.


A round-up of posts from the broader healthcare community this week that made us think

Marissa Moore over at OMERS Ventures drills down into the world of dental vertical software - unpacking dental industry trends, buckets of dental practice software, and more.

Six imperatives for AI-first companies by Morgan Cheatham and Steve Kraus

The Bessemer Ventures crew offers up a thesis on 6 imperatives for AI-first companies, from creating data advantages to recruiting AI scientist expertise to establishing moats and beyond.

An interesting read looking at what 2024 might hold in terms of M&A deals, suggesting we might see an increase in deal activity, particularly stock-for-stock deals, if the fed cuts interests rates next year as predicted.


or to participate.