Pieter De Leenheer

Pieter De Leenheer's core argument is that American healthcare's problem is not a lack of data but a failure to combine and trust the data that already exists. He describes an industry where "clinical data is today only used by pharma for A/B testing in trials and claims data only for insurance risk underwriting" ¹, two rich datasets kept in separate silos. His central insight is that "once you start combining data these two sources together you get in one plus one is more than two effects" ¹ — new use cases like care management and population health emerge only when the walls between clinical and claims data come down. This is the throughline of his thinking: the technology to unlock value already exists, the obstacle is structural and cultural separation, not invention.

He frames the stakes in blunt economic terms. "Today the healthcare system costs us about 5 trillion dollar a year, which is nearly 20% of the GDP of the American economy, and if we don't do anything about it, it will just lead to the bankruptcy of the system overall" ¹. For De Leenheer, shifting from fee-for-service to value-based care is not a policy preference but a survival requirement, and better data is the mechanism that makes value-based reimbursement possible at all. He extends this to disease prevention, pointing out that the four major chronic disease categories — cancers, metabolic disorders, cardiovascular and neurodegenerative diseases — share overlapping risk factors, and that correlating risk, behavior and inflammation data could eventually make prevention itself data-driven ¹.

A recurring theme is his frustration with the industry's technological backwardness. "The healthcare industry basically still kind of lives in the 1980s when it comes to technology" ¹, he says, and he is pointed about who profits from that stagnation: "Companies like Iron Mountain make billions of dollars in filling airplanes with patient records, 9 billion a year" ¹. This is not incidental color for him but evidence of a system built around paper and proprietary lock-in rather than interoperability. He is explicit that the number one challenge is not technical capability but that healthcare CIOs need to make the same cloud migration that financial-sector CIOs already completed decades ago ¹. He tempers any urgency for quick fixes with realism: "it takes also a long time to revert an industry that has been going the wrong direction for almost 50 years" ¹.

On trust and governance, he draws a direct analogy to banking. "If you give your money to a bank you trust the bank with their fiduciary responsibility to invest it as such and give you full transparency into their investments, and I think we have to have the same mindset with data" ¹. He pushes this further by noting the asymmetry in how consumers already accept data tracking elsewhere — "I do know if I pull up my phone here what did I order at Amazon for the last 10 years, and Google even knows where I was 5 years ago" ¹ — suggesting healthcare's caution around data sharing is inconsistent with norms already accepted in commerce. Counterintuitively, he argues cloud platforms can make data use more observable and auditable than the status quo of records sitting on local doctors' laptops, reframing the cloud not as a risk to patient privacy but as a governance improvement.

He also sees regulation itself as foundational infrastructure rather than a constraint layered on top of technology, comparing it to AT&T's telephone cables reaching even a single remote house in Texas — without regulatory tailwind, he suggests, the connective tissue for data-sharing simply doesn't get built ¹. That said, he is critical of specific regulatory choices, flagging Europe's AI Act clause banning applications that "influence behavior" as troublingly vague and potentially excluding legitimate healthcare opportunities ¹.

Looking ahead, De Leenheer is candid that near-term gains from better data will likely accrue first to insurance companies through quality-based government reimbursements, with patient control over their own data only rebalancing later ¹. This is a notably unsentimental prediction — he doesn't promise patients immediate empowerment, but rather a two-stage process. On AI specifically, he draws a sharp line: healthcare AI should focus on narrow, human-in-the-loop applications like care management recommendations, not generative AI, which he considers still prone to hallucination and far from ready for clinical use ¹. The takeaway across his remarks is consistent: unlock and combine existing data, treat it with fiduciary seriousness, build on regulatory infrastructure, and apply AI narrowly and cautiously rather than chasing generative hype.