I recently started building products focused on healthcare affordability in the US. As I was ramping up on a new space, the biggest question that sparked my curiosity was: how did we get here? This question is the inspiration for this weekly series chronicling the decisions, accidents, and breakthroughs that built the US healthcare system.
To understand why Epic exists, you need to understand what happened to American healthcare after World War II.
The wage freeze of 1942 didn’t just create employer-sponsored insurance — it created a three-way relationship between patients, providers, and payers that hadn’t existed before. A doctor saw a patient. But now a third party — the insurer — decided what to pay, and under what conditions. To get paid, physicians had to document, code, and justify every interaction to a bureaucracy with its own rules.
Medicare and Medicaid in 1965 added the federal government as the largest single payer in the country. The documentation requirements didn’t just continue — they multiplied. A diagnosis needed a code. A procedure needed a code. The attending physician’s notes had to justify both. By the 1970s and 80s, as private insurers followed the same model, getting paid had become a full-time administrative task running parallel to the actual practice of medicine.
This created the conditions for a software behemoth.
In 1979, a computer science graduate student named Judy Faulkner started a company in a Madison, Wisconsin basement with $70,000 borrowed from friends and family. She called it Human Services Computing. Her co-founder wanted to raise venture capital and grow fast. Faulkner refused. He left in 1983.
What Faulkner built — slowly and deliberately over the following two decades — was deceptively simple in concept. A single database. She called it Chronicles. Every clinical record, every billing entry, every scheduling note, every lab result — all of it living in one place, all of it connected. No separate systems that needed to talk to each other. No data loss between the clinical side and the billing side. One source of truth for the entire patient journey.
The contrast with competitors was stark. Cerner, Epic’s main rival, had grown quickly through acquisitions — eventually an amalgamation of 24 different companies stitched together. Integration was always incomplete. Data lived in silos. Chronicles was the opposite of that.
By 2001, Epic had fewer than 90 customers and $50 million in revenue. Then the phone rang.
In 2003, Kaiser Permanente — the largest nonprofit health system in America, the one that appears earlier in this series — issued a request for proposals for a new electronic health record system. The contract was part of a $4 billion infrastructure project covering 30 hospitals, 400 clinics, and 11,000 physicians.
Kaiser’s initial instinct was to split the work. Use Epic for outpatient care — where Epic’s reputation was strongest — and Cerner for everything else. Judy Faulkner said no. Splitting clinical records between two systems meant data loss at the handoff between outpatient and inpatient care. Bad patient experience. Bad medicine. All or nothing.
Kaiser also asked for equity in the deal. Cerner offered 10%. Faulkner said no to that too. Epic didn’t negotiate. It didn’t discount. It didn’t give away equity. They knew they had better software and just prepared harder — Epic’s team flew in the night before the technical due diligence meeting and rebuilt their entire presentation to model Kaiser’s specific transaction flow.
Epic won. Revenue went from $50 million to $162 million almost overnight. The Kaiser deal established Epic as the gold standard for large health systems. The floodgates opened.
Then in 2009, the federal government accelerated everything.
The HITECH Act set aside $27 billion — eventually rising to $35 billion — to incentivize hospitals and providers to adopt electronic health records. In 2008, only 10% of American hospitals used EHRs. The incentives were enormous — up to $44,000 per physician, up to $6.37 million annually per hospital. After 2015, providers who hadn’t adopted EHRs faced Medicare reimbursement reductions.
Epic has faced criticism for its role in drafting the HITECH legislation — Judy Faulkner was involved in the process. Whether that shaped the outcome in Epic’s favor is debated. What’s less debated: Epic was already winning before HITECH. Kaiser had already chosen them. The legislation added tailwinds to a company already on the ascent. It also added complexity — the Meaningful Use requirements created new documentation mandates that compounded the billing burden EHRs were supposed to reduce.
Today Epic supports the records of over 250 million Americans and generates $6.7 billion in annual revenue. Faulkner, who still runs the company at 82, has never taken outside investment, never made an acquisition, and has committed to giving 99% of her assets to charity.
A unified patient record that follows someone across a health system catches drug interactions, flags allergies, and surfaces clinical history at the point of care. Hospital mortality statistics improve meaningfully after EHR implementations mature. MyChart — Epic’s patient portal launched in 1999 — gave patients access to their own records a decade before competitors thought to try it.
And yet physicians today spend roughly half their working hours on EHR documentation. The tool built to manage billing complexity created its own complexity. Doctors enter data not because it helps them treat patients but because it satisfies reimbursement requirements. EHRs became the single most cited cause of physician burnout in the country.
The software that runs American healthcare was built because billing became as complicated as medicine itself. It made medicine safer — and it made being a doctor significantly harder.
Paul Starr’s line applies here too. The dream of reason did not take power into account. Epic didn’t create the billing complexity. It was built to survive it. But in doing so, it became the behemoth that now defines how American healthcare runs — for all its good and all its challenges.
Consider what switching away from Epic actually means. When one major health system attempted to migrate to a different EHR, the project took over a decade and cost tens of billions of dollars — and still wasn’t complete. That’s not a software contract. That’s infrastructure.
Epic controls records for over 250 million Americans, sits at the center of clinical workflows in more than half of US hospital beds, and keeps 35% of its revenue invested in R&D. Every year it expands further — into scheduling, billing, staffing, AI documentation, population health. The question healthcare startups ask isn’t whether to compete with Epic. It’s whether Epic has noticed them yet.
In healthcare technology, you don’t fear running out of runway. You fear ending up on Epic’s roadmap.
A Learning a Day 