Every autumn, CB Insights publishes its Digital Health 50 - a data-driven ranking of the world's most promising private digital health companies, selected from a pool of more than 12,000. It's one of the few lists in our industry worth taking seriously, precisely because it isn't an opinion. It's built on predictive signals: funding momentum, partnerships, hiring, investor quality.
We recently went through the last three cohorts (2023, 2024, 2025), company by company. 150 slots, 140 unique winners (a handful earn repeat appearances; Hippocratic AI is the only company to make it three years in a row). We were looking for patterns to inform our own investing.
The count
Of 140 companies, 18 are European. That's 13%. The UK leads with five Germany and France have three each, Spain and Belgium two, and then a scattering of single appearances - Austria, Sweden, Hungary.
In three years and 150 slots, exactly one company from all of Central and Eastern Europe has made the list: Turbine, a drug-discovery simulation company from Budapest.
A plausible explanation would be that the analysts in New York just don't look at Europe closely enough. But that explanation doesn't survive contact with the methodology. The Mosaic score doesn't care about passports - it measures traction, capital, partnerships, growth. So why does HealthTech money prefer the US?
The obvious (and incomplete) answer
The obvious answer is size. The United States spent $5.3 trillion on healthcare in 2024 - 18% of its GDP, $15k per person, growing at 7% a year. The EU spends roughly 10-11% of its GDP, and even its most expensive systems (Germany and France, at 12%) sit far below the American level. One country, one language, one regulatory regime, and a healthcare market bigger than the entire German economy. Of course the capital, the startups, and therefore the list-makers' signals concentrate there.
But size alone doesn't explain the composition of the list - and the composition is where it gets interesting.
Look at what the winning companies actually do. A surprising share of them (roughly half, by our count) don't diagnose anything, treat anyone, or discover any molecule. They exist to make sure someone gets paid. Revenue cycle management, prior authorization, medical coding, claims review, eligibility verification, price transparency. Entire venture-backed categories - some of the hottest companies in HealthTech - dedicated to administering the flow of money between the people who deliver care and the people who pay for it.
To understand why, we have to understand one structural fact: in healthcare, the person receiving the service, the person delivering it, and the person paying for it are three different actors. Europe simplified this triangle - the payer is the state or quasi-state insurance funds, prices are set administratively, and getting paid is a reporting task. America turned the triangle into a battlefield: thousands of hospitals and physician groups negotiating secret prices with hundreds of insurers, each of which operates thousands of plan variants, each with its own rules for what it will cover and how hard it will fight before paying.
Every front in that war is a business model. Walk one patient visit through the system and you can watch venture-scale companies appear at each step.
Six businesses hiding inside one doctor's visit
Eligibility. Before an American clinic treats you, it must verify that you're insured, by whom, under which plan variant, and how much of your deductible remains. Get it wrong and the claim is dead before it's born. Sohar Health (CG Insights winner in 2025) automates this with AI. In Prague, this step is a card lookup that takes two seconds, because everyone is insured on essentially identical terms.
Prior authorization. For expensive drugs and procedures, the insurer demands permission before treatment - clinical justifications assembled, faxed, argued over the phone. Insurers, rationally, make this slow: every abandoned request is money saved. Infinitus, one of the 2025 winners, deploys voice AI agents whose core job is literally to call insurance companies and sit through the phone tree. The company has raised over $100M so far because in the world's most advanced healthcare market, humans must phone other humans to beg permission to treat a patient.
Coding. The visit must be translated into billing codes, and the same visit can legitimately be coded at different intensity levels with very different price tags. So providers employ certified coders to code as high as defensible, insurers employ auditors to push it back down, and Fathom (a 2024 winner) sells autonomous AI coding to replace entire departments of these people. Polish doctors code too, but into a fixed tariff where gaming yields almost nothing. No arms race, no market.
Claims review. The insurer receives a claim - hospital bills can run to hundreds of line items - and searches for reasons to deny or reduce it. Alaffia Health sells AI to payers for exactly this. Which brings us to:
Revenue cycle management. The provider's counter-industry: software to fight denials, resubmit claims, and recover revenue (Adonis, Janus Health, and half a dozen more across the three lists). The phrase "revenue cycle" barely translates into most European languages. There is no cycle. You report, you get paid.
Price transparency. Because every provider-insurer pair negotiates rates in secret, nobody in America knows what anything costs. The federal government had to pass regulation forcing hospitals to publish machine-readable prices - and Turquoise Health built a venture-backed business just cleaning up that data. In Europe, the price list is the regulation. A price-transparency startup here would be an oxymoron.
Add it up and the numbers stop being funny: the US spends an estimated $500 billion or more per year on billing and insurance-related administration alone. Roughly the entire healthcare budgets of Germany, spent not on care but on moving money between actors who don't trust each other.
This is the uncomfortable insight hiding in three years of Digital Health 50 lists: in American healthcare, the dysfunction is not a flaw in the market - the dysfunction is the market.
And investors fund it for entirely rational reasons. The customer's pain is measurable in dollars. The buyer has a P&L and competitors breathing down its neck. The ROI case writes itself - "we recover 4% of your denied claims" closes deals in months, not years. It is, in our view, one of the most reliably monetizable inefficiencies in the global economy.
What Europe's 18 companies have in common
Now let’s look back at the European winners with this lens and a pattern jumps out. With almost no exceptions, they are B2B deep tech - computational pathology from Berlin and Hamburg, radiology AI from Munich, ECG algorithms from Madrid, ambient documentation from Paris, drug-discovery simulation from Budapest. Clinically validated, regulatory-grade technology that mostly earns its revenue by exporting to the US market. Europe doesn't place care-delivery companies on this list, because scaling patient-facing care across 27 different reimbursement systems is a contradiction in terms.
The two exceptions prove the rule. Anima grew inside a single national system (the NHS) that it was purpose-built for. And Neko Health, the Swedish full-body-scan company co-founded by Daniel Ek, simply refused to play the reimbursement game at all: consumers pay cash, and over 350,000 people are globally on its waiting list.
Here's the detail we keep coming back to: Neko and Function Health - the American cash-pay longevity membership - carry the two highest Mosaic scores in the entire 2025 cohort. The market's most data-driven list is telling us, in effect, that the most promising business models in digital health right now are the ones that don't wait for a payer's permission.
What we take from this as a Central European investor
Three conclusions, stated with the usual caveat that this is a (simplified) view from someone deploying capital in Central Europe, not a policy paper.
First: We cannot copy the American playbook here, and that's fine. Half of the US health tech gold rush is friction arbitrage, and Europe doesn't have the friction. Anyone pitching a "European Tennr" is solving a problem their market never created.
Second: The European models that work are visible in the data: export clinically validated technology to where the money is; embed so deeply in one national system that you become part of its plumbing; or route around the payer entirely and sell directly to people. That third lane is more open in our region than most investors assume - out-of-pocket healthcare spending is already normal consumer behavior across CEE, where private diagnostics, dental, and ophthalmology have trained a generation of patients to pay for speed and quality.
Third: Europe's slowness is also its moat. American health tech monetizes chaos; European health tech has to create actual clinical or operational value, because there's no administrative inefficiency to arbitrage. That's a harder game to enter and a slower one to win. But revenue that comes from delivering real care inside a stable, price-regulated system doesn't churn, doesn't get competed away on price, and doesn't disappear when the arms race finds a new weapon.
The Digital Health 50 will likely stay American-dominated for years, and the lists are worth reading anyway - not as a scoreboard, but as an X-ray of where healthcare's money actually flows. Let’s just remember what we're looking at: a market where an enormous share of the innovation exists to manage the fighting. Europe never built this specific battlefield.
The healthcare opportunities here lie elsewhere - let’s talk about it in the next article.
