Round Up: A Tale of Two Health Systems, a Wave of Health-Tech News, and a Four Questions with Co-founder & CEO of Ambience

“A Tale of Two Systems.” That could be the title of my future book if I ever get around to writing another. I’ve been thinking a lot these days about how we seem to now have a burgeoning cash-pay healthcare system, coupled with an entirely distinct one that runs on insurance. The problem? The data isn’t connected between these two disparate systems. One example: I recently got a whole-body MRI scan, but my primary care doctor has no idea and hasn’t seen the results. Likewise, I’m speaking with countless people who are taking GLP-1s, HRT, and other therapies via direct-to-consumer websites, and this important and relevant information isn’t anywhere in their electronic medical record. 

On the one hand, those who can afford it are getting convenience and concierge care. But sometimes I wonder if we’re double or even triple paying. Most Americans are already paying a fortune to get health insurance through employers, both in terms of what’s coming out of the paycheck and the premium, while noticing that our take home pay doesn’t seem to be going up. And in addition to that, they’re paying out-of-pocket for care and struggling to meet their deductible.

Something has to give, which is why much of our content over the next few months will focus on the perfect storm that we’re currently in. We’re seeing general inflation, coupled with rising hospital costs and specialty drug costs. This is a moment that will force creative solutions because at some point the dam has to break. We are spending $2 trillion more on healthcare than our peer countries and yet somehow still underinvesting in primary care, behavioral health, and prevention. It might also be the breakthrough moment that digital health needs.

So, with all that said, any company in health-tech that is actually helping improve efficiency, reduce waste, and divert spending to the right places will thrive in this environment. A few of the companies listed in our news and deals round up are doing just that, so let’s dive in.

With Annalisa Merelli 

The news: Health insurance startup Sidecar Health is expanding to include 1,200 providers in Carrum Health specialty network, particularly for surgery, cancer, and addiction treatment. 

How it works: Sidecar Health shows the prices of its providers (now including Carrum Health’s network) to customers, allowing them to compare. If they pick a specialist priced below their benefit amount, they cash in half the savings. If the price is higher, they pay the difference.

Why it matters: This is a practical test of whether comparing prices and price transparency can work for healthcare. The big question is whether there’s opportunities beyond the areas where Centers of Excellence programs have been most effective, namely expensive surgeries where there’s a huge variation between providers and a tangible opportunity for cost savings. 

The news: One Medical is getting on with the menopause program. The company will offer a menopause-specific visit that can then connect the patient to medical resources to address any symptoms. 

What One Medical said: "This holistic approach allows us to provide more comprehensive and individualized care during this transition."

Why it matters: One Medical’s new offering is the latest confirmation that the momentum behind treatment for perimenopause and menopause is not slowing anytime soon. It’s also somewhat surprising it took this long, as One Medical has been around for a while and menopause is such a strong need for its female patients when it occurs.

The news: Aidoc received a breakthrough device designation from the FDA for its AI program that can triage multiple acute findings in CT scans within a single workflow, rather than a separate one for each finding. 

What Aidoc said: “We’re very excited by FDA’s leap in its recognition of the importance and value of multi-indication AI models, as reflected by this breakthrough designation, acknowledging our technology has the potential to help patients with a wide array of life-threatening and debilitating diseases,” said Aidoc co-founder and CEO Elad Walach. 

Why it matters: This is an intriguing update that shows a more friendly regulatory environment towards AI companies than we’ve seen in the past with other technology breakthroughs. Steps like these make AI more viable to payers and health systems.

The news: The first large randomized controlled trial of AI for breast cancer screening has been announced. It will take place in seven sites with a budget of $16 million, and will provide some insight into a technology already applied to millions of mammograms in the US.

What the experts said:  “We don’t really know if proposed solutions can actually improve the lives of patients or are just part of a growing AI fad that has permeated a number of industries,”  Sanjay Aneja, a radiation oncologist at Yale Cancer Center who studies mammography AI, told STAT reporter Katie Palmer. 

Why it matters: The trial will tackle many nuanced questions, including how AI can influence care and whether previous exposure to AI influences radiologists. There are already a number of companies in this space with venture capital behind them, so we’ll keep you posted once there are findings from the study.

The news: The president tweeted on Truth Social that cannabinoids could "revolutionize senior healthcare," and stocks of cannabis companies – as well as MSOS, the largest marijuana-focused EFT – skyrocketed.

The industry’s feelings: “We're encouraged by the U.S. administration's recognition of cannabis and its potential role in supporting health and wellness for seniors," a Ontario-based company, Canopy Growth, told Reuters.

Why it matters: Trump has said his administration plans to reclassify marijuana, which would likely further push up cannabis stocks.

$4.5 million for Moneta Health: The company, which offers AI-powered cognitive rehabilitation therapy, announced its latest funding round, as well as its expansion into Montana through a partnership with Benefis Health System.

$400 million for Capital Rx Judi Health: The investment, which included a $252 million Series F round, brought a rebranding for the health technology and benefit administrator company. Judi, the company explained in a press release, is short for adjudication. 

Prosper AI gets $5 million in seed round: The company is building a specialized voice AI platform to take care of costly and critical front and back-office work. The round was led by Emergence Capital. 

Oura Health Oy is raising $875 million in Series E: The maker of the Oura ring is reaching a total valuation of close to $11 billion, or double its Series D valuation. The round is still ongoing and could exceed $900 million. 

Datavant acquires DigitalOwl: The health data network Datavant will acquire the startup DigitalOwl, which offers AI-based medical document analysis. The deal is estimated to be around $200 million in a mostly cash deal.

Premier is going private: The service provider, which works with health care systems to cut costs by coordinating purchases of equipment and supplies, agreed to go private in a $2.6 billion deal. 

€14 million ($16.4 million) for RDS: French medtech RDS closed its Series A funding led by the SPI (Sociétés de Projets Industriels), a fund managed by Bpifrance on behalf of the French government. The funds will go toward MultiSense, a connected patch for remote patient monitoring.

It seems that sometimes we're telling ourselves in healthcare that AI can solve everything. What can it realistically solve - and what can't it? 

NB: The reality is we’re at a place and time where we have no choice but to leverage it. Most health systems are partnering on AI, not building it themselves. So for this next phase, it’s all about the right partners on the software side -- whether it’s solving the clinician burnout problem, the rev cycle problem, the quality problem, and more. The solution in the past was to throw labor at these problems. AI has enormous potential but the challenge is the frontier models off the shelf aren’t good at health-specific tasks. There’s no evidence they’ll get good without the right company distilling the expertise of the world’s best experts into these models. And then these partners need to own the outcome, so the viability of their organization only exists if the problems actually get solved. To underline that, it means software vendors that are selling AI solutions need to go at risk.

Should physicians and other providers be doing work with AI companies (the labeling, the reviewing, and so forth) at the potential expense of themselves - especially if this AI is coming for their jobs? 

NB: There’s so much to unpack here with this question. I think most people misunderstand how to build high-quality models and they might be collecting bad data to create bad models. So there's that. But from a clinician POV, it’s a tricky one. It really is. The reality is whether we like it or not these models will get smarter and they’re going to get smarter in all domains. Foundation model makers want to make them really good at some domains like software engineering. There’s still work to be done, however, to make foundation models good at healthcare-specific things like prior authorization and coding. This speaks to a bigger question of what the role of people should be in the industry as the models get better and better. Even at our own company, the job of an engineer has changed so much. So should software engineers be making models that code better than them? And should we be asking about how we reimagine our role, in collaboration with them? On the clinician front, it's not about taking jobs. It's that we don’t have enough providers to go around to take care of the population, especially as we age. 

That's an interesting point about software engineering. As a CEO, how are you hiring differently because of AI? 

NB: I think the reality is hiring a smaller number of truly world class people, and pushing the bounds of how we give people leverage in terms of AI. Engineering is the most mature in terms of leveraging code capabilities, triaging root cause for system downtime, looking for security vulnerabilities. What previously required a staff engineer surrounded by 5 others can now be done with 1 engineer. We are finding that we don't need 250 engineers, when the same work can be done with 50. We are seeing that play out over other parts of the org now also. So one of our big projects: How do we make sure that every conversation we have with a partner or prospect gets encoded into a shared brain — we have a dedicated team building a company brain so every person can see it. How do we take the lived experience & context of our 10th employee and give it to our 100th? Mike [Ng, cofounder & President of Ambience] is spending time right now on building that brain to give everyone leverage. So bottom line: We are hiring far fewer people, and our bar is much higher.

So why raise venture capital? Couldn't you get by with far less given you are becoming more operationally efficient? 

NB: We believe companies like ours should put revenue at risk with our health system customers. We get paid if we succeed at what we say we will. If you have $100m of revenue at risk, for example, you can’t do that with $70 million in the bank. That’s a major reason that prompted us to raise our $243m series C.