Weekly Health Tech Reads 6/21/2026

Thoreau makes its platform move and joins the AI-enabled healthcare admin platform race, acquiring Ensemble for ~$12 billion
On Tuesday, it became official that Matt Holt’s Thoreau made a strategic investment in Ensemble at a ~$12 billion valuation, with Ensemble’s current owners, Warburg Pincus, Berkshire Partners, and Bon Secours Mercy Health remaining minority shareholders. Thoreau didn’t take long to recover from its failed quest to acquire his former portfolio at New Mountain Capital and found Ensemble as a new platform asset. Ensemble has been on the block for some time now, as it was reportedly going through a dual-track IPO/M&A process late last year at a ~$13 billion valuation, which seems to hint at the public market reaction to a potential IPO here. Warburg and Berkshire bought Ensemble from Golden Gate Partners in 2022 at a ~$5 billion valuation.

It will be fascinating to watch what comes next for Thoreau. Holt’s playbook at New Mountain, along with the original vision for Thoreau, certainly indicates that we should expect to see a number of other AI-centric assets bolted on to Ensemble. Thoreau itself is described in the press release as an investment platform, which is perhaps a bit different than the operating company I think many envisioned. We were chatting in Slack this week about how Ensemble’s JPMorgan presentation earlier this year featured a refreshingly human narrative, highlighting some of the challenges AI faces in automating RCM while underscoring the value of a boots-on-the-ground approach. Given Thoreau's AI-centric narrative, that perhaps provides some breadcrumbs indicating the path of travel here.

A key blow to anti-tiering / anti-steering clauses potentially opens the door further for alternative plan designs in the commercial market
The DOJ announced this week that it has settled its antitrust lawsuit against OhioHealth. As part of the settlement, OhioHealth has agreed that its anti-tiering and anti-steering clauses are unenforceable, although it appears that all-or-nothing clauses have emerged unscathed as I read the settlement. The DOJ originally brought this lawsuit to help drive competition in the commercial insurance market, as it opens the door to more creative plan designs. On Friday, the White House released a report from the Council of Economic Advisors suggesting that banning anti-tiering/steering and all-or-nothing contracting practices could reduce hospital prices by 18% and employer premiums by 6.5%.

The DOJ has filed a similar lawsuit against New York Presbyterian, and it will be worth keeping an eye on if that has a similar outcome. If this is the beginning of the end of anti-tiering/steering clauses, it seems like a huge opportunity in the commercial insurance market, and a massive tailwind for the alternative plan design market, which is already benefiting from a huge tailwind around cost increases. Very interesting to think about the negotiating strategy surely going on behind all of this and how health systems like OhioHealth evaluate what they “give” on here. I’d probably give on the anti-tiering / steering clauses first over all-or-nothing and bet that, given the brand position in the market, it won’t ultimately impact commercial volume all that much.

The employer innovation market continues to heat up

On a related note, we saw two interesting signals this week about activity in the employer innovation market as rising medical costs create an opportunity to sell new offerings into employers seeking to better address cost increases. First, HCSC launched its copay-only plan design, Easify. Various Blues plans in the HCSC umbrella are rolling out this offering, as copay-only plans seem to continue building on the wave Bind/Surest started a few years ago. Second, Marathon Health and Lantern partnered to combine Marathon’s primary care footprint with Lantern’s surgical network. The logic of combining an advanced primary care model with a specialty care platform seems pretty straightforward, and I’d imagine we'll see more partnerships like this as platforms seek to help employers manage healthcare costs holistically while also ensuring access to high-quality care.

Chaos remains in the Medicare Advantage Stars program
CMS seemingly pre-empted a series of Stars-related lawsuits by sending a letter to the industry that it will recalculate Stars scores after Clover’s recent legal win. Plans with a higher score can resubmit bids by June 29th. In a post-Chevron environment where the Stars program faces existential questions, this decision seems to buy CMS time to figure out what to do next.

The stock market reaction to the news was muted, which provides some sense of the overall impact of this change — it’s not nearly the swing we saw around advance/final notice earlier this year. Still, it seems like another positive nudge in an industry that had a solid Q1 earnings season, providing a general sense of optimism about overall 2026 performance after several challenging years. I have to imagine that CMS leadership is keeping an eye on all of this as we move through 2026 in the ongoing negotiation between the federal government and the industry over healthcare costs. In roughly six months, we’ll be talking about the 2028 Advance Notice, and I’d think these sorts of wins will not be lost on CMS when evaluating its stewardship of taxpayer dollars.

Indiana’s effort to cap hospital prices is one of the more interesting regulatory efforts to curb hospital prices in the country at the moment, with hospitals that exceed the price cap potentially losing their tax-exempt status if they’re not below the threshold by 2029. This is sparking debate over which numbers should go into the threshold — if practitioner services are included, hospitals would appear to be below the overall cap. It would seem logical that all the services a health system offers should be included in the threshold. But also, if the underlying issue is that inpatient costs are too high, including practitioner costs in the calculation allows systems to lower their weighted average below the threshold, which seems to defeat the purpose of a price cap entirely.

Midjourney, the company generally known for its AI videos, is entering the healthcare market, announcing a 60-second full-body MRI scanning tool along with a 23,000 sq ft medspa / wellness clinic in San Francisco. Midjourney hopes that by 2031, it could deploy 50,000 of these MRI scanners across 5,000 locations around the world, casually throwing out it might need $20 billion to do so. The medspa location won’t open until the end of 2027. Midjourney has a licensing agreement with Butterfly Network to use Butterfly Network's ultrasound technology in its scanner. Midjourney suggests that with early imaging like this, the world will avoid 30% of all deaths and 50% of future healthcare costs, which feels like a great rallying cry for a company and also a very pleasant alternate universe to live in. The whole approach here is just so confusing to me — hopefully the technology works, but if so, why not just build the tech and sell it to other clinics? Why spend energy building out medspas too? Are people only going to use revolutionary scanning technology if it’s part of an AI-generated spa experience? I’m lost. Part of me feels like Butterfly is the more interesting technology to unpack behind this announcement.

OpenLoop Health announced a new offering that appears to be akin to a Shopify-for-D2C-telehealth, purportedly allowing influencers, gyms, dating apps, and anyone with a following to spin up a telehealth site within hours to start selling GLP-1s and the like. Whether you think this is good for access to care or enabling more MedVi scenarios, it certainly seems like an interesting time to be building in the wellness market.

In the PBM market, LucyRx and Abarca Health are merging, combining their platforms to create a modern PBM better positioned to compete against the Big Three.

Connie, an AI-enabled rollup of Medicare Advantage brokerages, raised $40 million. HealthQuest Capital led the round. Along with the funding, Connie announced its tenth brokerage acquisition. We’re seeing a lot of momentum in the brokerage market at the moment, specifically around the play of rolling up independent brokerages with AI, but more broadly around the importance of the broker as a strategic advisor in the MA market. As the MA market shifts, we think this theory holds water, and some companies will do quite well here by positioning themselves as trusted advisors to members/plans. One thing we’re watching as this market heats up: whether the competition for quality independent brokerages leads to frothy purchase prices and erodes financial outcomes here.

InStride Health, a virtual pediatric mental health model, raised $30 million. Echo Health Ventures and FMZ Ventures led the round. InStride reports it is expanding nationally, currently operating in 17 states and treating 5,000+ patients. It doubled revenue in 2025 and expects to double again in 2026. InStride has a cool background — its co-founders developed the clinical model at McLean Hospital and turned that into InStride.

OffPlan, a startup that is helping physicians open DPC practices and contract with small employers, raised $2.5 million in a seed round. It plans to launch in Virginia and Florida later this year before expanding nationally. Another indicator of the momentum in the DPC market right now, although this version of the play seems to have an extremely high degree of difficulty, as it attempts to build both what appears to be a DPC clinic-in-a-box model and an alternative plan design/network for employers simultaneously. We just saw Mending pivot to focus on the DPC network for employers’ play while exiting the ACA market, perhaps a nod at the benefits of focusing on one piece of the equation here.

NPR featured a story about access to GLP-1s via employers, highlighting two anecdotes: an employee who was unhappy with their experience with Vida Health, and a primary care provider who was unhappy with Virta Health deprescribing GLP-1s for one of their patients. The undertone of the piece is that these approaches help employers manage access to GLP-1s, implying they’re not on the patient's side. It’s a watch-out for this part of the industry from a PR perspective — as employers grapple with how to manage GLP-1 costs, it’s not hard to envision a scenario where these approaches face heavy criticism that looks a lot like traditional critiques of utilization management. The articles about frustrated employees being denied access to GLP-1s almost write themselves.

NPR also highlighted the conclusion of the Oregon emergency department staffing drama: PeaceHealth canceled its contract with ApolloMD and is moving forward with contracting with the local ED group that was at risk of being displaced. The Oregon Nurses Association has applauded the move as a “historic victory for all Oregonians,” and NPR highlighted the various corporate practice of medicine dynamics at play. Worth keeping this situation in mind as similar situations inevitably play out in other locales.

I am fascinated by this Substack post from an orthopaedic surgeon lamenting Higmark Health’s decision to move its prior auths from EviCore to an internal AI solution. The core of the reasoning is that AI is denying more claims because of ambiguity around surgical materials, whereas humans at eviCore would have interpreted the ambiguity differently and approved the claim. Note, this is the same eviCore that has been referred to as EvilCore, and Cigna has essentially deemed it worthless as it seeks to jettison the asset and move on. It is such a great example of what is actually happening on the ground as organizations shift to AI, why I remain pretty convinced that fully automating away payor/provider frictions is a pipe dream, and how humans will play a key role even in an AI centric world. If I were an investor, I’d be checking with Cigna’s corp dev team about the asking price for eviCore at the moment…

The WSJ highlighted the ongoing clash in Utah over the AI prescription-refill pilot, noting physicians' concerns about patient safety and medical liability. Meanwhile, the answers to these issues from Doctronic / Utah’s AI Office seem pretty straightforward. I continue to think that, of all the ways to roll out an AI-doctor pilot, Utah’s approach here is quite reasonable.