THIS WEEK IN HEALTHCARE
What happened in healthcare this week—and what we think about it.
- General Catalyst announces intent to buy a health system. On Sunday, venture capital (VC) firm General Catalyst unveiled the Health Assurance Transformation Corporation (HATCo), a new subsidiary company which aims to acquire a health system to serve as a blueprint for the VC firm’s vision of healthcare transformation. Sharing this news on the first day of the HLTH 2023 conference in Las Vegas, General Catalyst declined to comment on which health systems are targets, or how much it is willing to spend, but CEO Hemant Taneja suggested that investment returns would be evaluated on a longer timeline than the typical 10-year venture capital horizon. Marc Harrison, the former CEO of Intermountain Health who joined General Catalyst in 2022, has been tapped to lead HATCo. The new company will build on General Catalyst’s previously announced partnerships with health systems, including Intermountain, HCA Healthcare, and Universal Health Services, with the goal of connecting healthcare startups with health systems in order to test and scale their technologies.
The Gist: While private equity firms have backed health systems before, a VC firm expressing interest in health system ownership is a surprising development. Even on a longer timeframe than most venture plays get, it’s difficult to imagine a health system ever delivering the outsized returns VC investors usually demand. It’s possible HATCo’s true value will come from scaling and selling the services of tech startups in General Catalyst’s portfolio after vetting them at their health system “proving ground”. HATCo’s more ambitious aim to align payers and providers in a pivot to value-based care is a familiar one, but the new venture will find itself up against skepticism from insurers and other entrenched stakeholders, which has been difficult for even the most motivated health systems to overcome.
- Two retail pharmacy giants reveal new virtual care offerings at HLTH. On Monday, Walgreens announced that it will begin offering direct-to-consumer virtual care in nine high-population states where nearly half of its pharmacy customers live. It will offer chat visits for $33 and video visits ranging from $36 to $75, relying on an undisclosed third-party partner to provide the service. The service will be cash-pay only when it begins later this month, but Walgreens plans to incorporate insurance coverage in the future. This week, Walgreens also appointed Tim Wentworth, former CEO of Cigna’s Evernorth health services organization and pharmacy benefit manager Express Scripts, as its new CEO. Wentworth will take the helm as the retailer tries to shore up its finances while continuing to expand into healthcare services.
And on Tuesday, Walmart announced the nationwide expansion of its virtual primary care option for employees. The 1M+ US associates and their dependents covered by the company’s health plan will have access to virtual primary care visits with zero copay, along with some specialist visits and home-based lab testing services. Walmart first piloted virtual primary care in 2020, working with Included Health, and cited employee health improvement and reduced hospital admissions as drivers of its decision to expand the program.
The Gist: Though consumer use of virtual care use has declined since its early pandemic peak, nearly all large retail and pharmacy chains now offer virtual on-demand services, bringing consumers a wide range of convenient and affordable care options for low-acuity conditions. Walmart’s strategy to leverage telehealth to expand primary care access for its employees makes a lot of sense as, prior to 2020, nearly half of its covered population was not receiving primary care. Walmart stands to lower its overall healthcare spend by giving employees increased access to routine care and earlier intervention, with an opportunity to expand the offering to all consumers down the road.
- California takes a step toward establishing universal health coverage for residents. California Governor Gavin Newsom signed a bill directing the state’s Health and Human Services Agency to work with the federal government to create a waiver allowing Medicare and Medicaid funding to be reallocated toward a universal health insurance system for its residents. The established timeline sets California on track to submit its final waiver for federal approval in 2026. The law does not specify whether universal coverage would be via a single-payer system, which is what Newsom favored in 2018. The California Nurses Association opposed the bill on the grounds that it does not commit to a single-payer outcome, while the California Association of Health Plans protested against its threat to end private coverage in the state.
The Gist: This is California’s 10th attempt at universal care, with all previous attempts having ended in failure because, despite both popular and political support in the state, there has not been consensus on how to pay for it. This most recent bill only passed because it was separated from a funding bill, since shelved, addressing the over $300B in tax revenue needed to pay for it. This process-first approach may be seen as a calculated appeasement of the Democratic Party’s left wing, as Governor Newsom clearly holds aspirations for higher office—but so far, healthcare has not ranked among the top issues for the current roster of candidates targeting the White House in 2024.
Plus—what we’ve been reading.
- How predictive AI models degrade over time. This informative article documents a concerning phenomenon in the fast-expanding world of AI-powered diagnostics: when hospitals deploy multiple AI models that track the same metrics, the success of one or more of them can create a feedback loop that undermines their collective predictive power. This makes it difficult to tell which AI-powered clinical models are effective in healthcare settings where multiple programs are implemented, for example in an intensive care unit. Even more concerning, retraining an AI model on patient data affected by other interventions can make that model perform even worse. This suggests that even if health systems carefully monitor their predictive AI models and catch when their performance degrades, it will be difficult to correct problems that arise.
The Gist: As health systems embrace AI models that promise to improve clinical outcomes, it’s important to grapple with the real-world limitations these breakthroughs still face. Unlike AI-powered productivity enhancers such as ambient notetaking, predictive AI clinical models are trained and refined based on live patient data that’s constantly changing, due both to external factors and the successful interventions of various models. And unlike other, more-centralized industries deploying AI, each health system and provider organization will bear the burden of ensuring their AI models continue to be effective.Careful and continuous monitoring of predictive AI programs will be essential, along with effective oversight and coordination of model deployment.