A Little Girl in a Hospital Waiting Room, with a Folder Full of Medical Records. Can Interop 2.0 Help Her?

Act I · The Paper Era

Once upon a time, the patient was the courier.

In 2005, a little girl sat in a waiting room next to her mother, a manila folder in her lap. The corners were soft from being carried to every appointment. Inside: immunization records, a handwritten referral, a radiology report on carbon paper, an allergy list her mother kept updating in pen. She was seven. She had been to four doctors in three cities. The folder was the only thing holding her medical life together.

For most of American healthcare in 2026 — that folder is still Tuesday. Not because providers are careless. Because the infrastructure behind them was built for a world where records lived in one building, faxes counted as "innovation," and the patient was the most reliable transport protocol in the system.

That folder represented something the industry took a long time to measure. A 2019 JAMA analysis by Shrank et al. put US healthcare waste at $760 billion to $935 billion annually — roughly a quarter of all healthcare spending. A big share of that number is administrative complexity and failures of care coordination. Duplicate imaging because the prior scan couldn't be located. Readmissions because the discharge summary never reached the primary care doctor. A Johns Hopkins study, also often cited, estimated that preventable medical error contributes to roughly 250,000 deaths per year in the United States. Not every death traces to a missing record. Enough do.

The founding question of interoperability, then, is not technical. It is moral. Why does the patient carry what the system should know? Every standard, every API, every regulation that follows is an attempt to answer that question — badly at first, then better, then finally at scale.

Act II · Interop 1.0

What $35 billion actually bought.

Interop 1.0 did not start with patient care. It started with billing. EDI transactions — 834 for enrollment, 837 for claims, 835 for remittance — were standardized under ANSI X12 in the 1970s and 1980s because insurers and employers needed to move money. The patient record tagged along as a side effect. Healthcare's first serious interoperability was a financial byproduct, and that origin shaped everything that came after.

Then came HL7v2, the workhorse. A pipe-delimited messaging standard that still runs the majority of hospital interfaces on the planet in 2026. ADT feeds moving patients through admissions. ORM and ORU messages moving orders and results. MLLP sockets opening and closing at 2 AM while an interface engineer somewhere debugged a Z-segment a vendor had invented to fit local workflow. HL7v2 worked. It scaled. It kept hospitals running. It also produced a generation of integrators whose entire craft was mapping one hospital's dialect to another's.

The policy wave came next. The HITECH Act of 2009 committed roughly $35 billion in incentive payments to push American providers onto certified EHRs. It worked — in one dimension. Per ONC data, physician EHR adoption moved from 9% in 2008 to 96% by 2021. The United States got digitized. What it did not get was interoperable. Meaningful Use paid for software, not for data flow. Providers bought the systems, attested to the measures, and then discovered that sending a summary across town was still harder than printing one out.

The global picture rhymed. In the UK, NHS Spine and the Summary Care Record gave Britain a centralized backbone earlier than most — imperfect, politically fraught, but functional at national scale. In the EU, epSOS (2008–2014) was the first serious attempt at cross-border patient summaries across member states, laying groundwork that the European Health Data Space would later inherit. India, in the pre-ABDM era, ran on a patchwork of state health systems, paper-dominant outside the private hospital networks. Every country learned the same lesson in a different dialect: you can digitize without interoperating, and most of the world did.

Interop 1.0 built the pipes. It created the category of "integration engineer" as a real profession. It gave us CCDAs, Direct messaging, and a generation of HIEs — most of which struggled for participation and sustainable funding. The honest ledger reads like this: 1.0 was necessary. It was also not enough. It proved data could move. It did not prove data would move to where it was needed, when it was needed, in a form the receiving system could actually use.

That gap — between "digitized" and "interoperable" — is what Interop 2.0 was built to close.

Act III · Interop 2.0

The global architecture shift.

Interop 2.0 is not a single standard or a single regulation. It is a convergence — four or five countries, working in parallel, arriving at the same architectural conclusion: health data should move through open APIs over the public internet, governed by shared standards and enforceable policy. What changed is not that anyone finally "solved" interoperability. What changed is that the entire stack — standards, policy, economics — started pointing the same direction for the first time in forty years.

The standards layer

FHIR R4, released in 2019, was the inflection point. Where HL7v2 assumed a dedicated socket and a custom mapping per partner, FHIR assumes a web. Resources are RESTful. Identifiers are URLs. Data elements are JSON. A modern developer can call a Patient endpoint without ever touching MLLP. Layered on top: SMART on FHIR for app authorization, USCDI for a common US data set, and the International Patient Summary (IPS) for a harmonized cross-border record.

None of this is theoretical in 2026. FHIR is embedded in every major certified EHR. SMART on FHIR apps plug into Epic, Cerner/Oracle, Athenahealth, and Meditech the way browser extensions plug into Chrome. The standards layer stopped being a debate and became a substrate.

The policy layer

The regulatory shift is the piece most technologists under-read. Between 2020 and 2026, every major economy put interoperability into law — with teeth. In the United States, the 21st Century Cures Act and the ONC Information Blocking Rule made it illegal for providers and vendors to unreasonably withhold electronic health information. TEFCA went live in late 2023; the first QHINs (Qualified Health Information Networks) began exchanging data in early 2024. CMS finalized the Interoperability and Prior Authorization Final Rule in 2024, mandating FHIR-based APIs for Medicare Advantage, Medicaid, and marketplace plans.

In Europe, the European Health Data Space (EHDS) regulation was adopted in 2024 and entered into force in 2025. EHDS mandates cross-border exchange of electronic health records, gives patients API-level access to their data, and creates a governed secondary-use regime for research and public health. It is the most ambitious single piece of health-data legislation in the world. The UK, outside the EU, is running a parallel track: the NHS Federated Data Platform contract signed in 2023, the NHS App maturing into a serious patient-facing front door, and a national "single patient record" ambition sitting on top.

India moved the fastest of any large economy. The Ayushman Bharat Digital Mission (ABDM), launched in 2021, has issued over 700 million health IDs. The Unified Health Interface (UHI) is designed as the "UPI for healthcare" — a protocol layer that lets any compliant app discover and transact with any compliant provider. India leapfrogged the 1.0 era the same way it leapfrogged landlines: by building on open protocols from the start.

Sitting above all of them is the WHO Global Digital Health Strategy 2020–2025, which positions digital health as foundational to Universal Health Coverage, and the WHO SMART Guidelines, which encode clinical guidance as computable FHIR-based artifacts.

The economic layer

The framing that makes the stack move is economic. The World Economic Forum's Global Health Data Charter argues that health data is a public good — its value compounds when shared. The World Bank treats digital health investment as a prerequisite for Universal Health Coverage in low- and middle-income countries, not an optional upgrade. The McKinsey Health Institute has estimated that closing the interoperability gap in the United States alone could unlock on the order of $250 billion in annual value — through reduced duplication, better outcomes, and care delivered at the right site.

The summary: 2.0 is not just a better standard. It is a shift from messages to resources, from interfaces to APIs, from point-to-point to networks, and from compliance projects to architectural commitments. The girl with the folder, if she were seven today, would still bring the folder. But her doctor would not need it.

Global Snapshot

Four economies, one direction.

The parallel policy moves across jurisdictions are worth seeing in one frame. Different political systems, different health financing models, different starting points — all arriving at FHIR-based APIs and networked exchange within roughly a five-year window.

None of these programs is finished. TEFCA is young. EHDS implementation runs through 2029. The NHS platform is politically contested. India's ABDM is scaling faster than any peer but still faces private-sector adoption gaps. Every one of them will disappoint someone. That is not the point. The point is that for the first time in the history of health IT, the policy direction and the architectural direction are the same direction.

Act IV · The Destination

When interop disappears.

Twenty years after the waiting room, the girl is grown. She has a daughter of her own. The question is no longer whether her daughter will carry a folder. It is whether she will need to walk into a clinic at all.

The destination of forty years of interoperability work has two parts, and the order matters. First: a health system that keeps her daughter out of the hospital. Not through austerity, but through continuity — predictive, preventive, ambient care, the WHO vision of Universal Health Coverage made real by infrastructure that knows her baseline and notices when it drifts. Second: when she does need a hospital, she carries nothing. No folder. No list of medications typed from memory. No phone call between clinics that may or may not happen. Her record is already there. Her clinicians are reading from the same page. The system did the work the patient used to do.

When that happens, interoperability stops being a topic. It becomes infrastructure, the way electricity is infrastructure. You only notice it when it fails. The engineers who spent decades building HL7v2 interfaces, mapping CCDAs, standing up HIEs — their work does not disappear. It becomes the foundation nobody has to think about anymore.

The honest closer is not triumphalism. It is orientation. Forty years ago, a girl carried a folder because the system had no way to carry it for her. Today, most of the technical and legal pieces exist to carry it for her daughter. Whether they actually do — whether the architecture becomes ambient the way electricity did — is the work of the next decade. Fax to EDI to HL7v2 to FHIR. 1.0 to 2.0. The journey continues.