When Partners HealthCare, now Mass General Brigham, budgeted for a new electronic health record rollout, the plan called for $600 million. The system ultimately spent $1.2 billion. The gap wasn't a rounding error or a scope change caught late. It was the ordinary cost of moving a health system's data out of one platform and into another, a cost that's rarely visible until an organization is already committed to paying it.

That story isn't unusual so much as it's illustrative. It's worth asking why switching costs in health IT run this high, and what it actually takes to avoid getting stuck the same way.

Lock-in is a business model, not an accident

Vendor lock-in happens when the cost of leaving a platform, financial, operational, or technical, becomes so high that an organization stays even after the relationship stops serving it. In healthcare IT, this isn't a side effect of complexity. Peer-reviewed analysis has argued that the dominant position of the largest EHR vendors isn't primarily a story about superior technology, but about structural forces: federal incentive programs, historically weak interoperability requirements, and network effects that make switching progressively harder the longer an organization stays put. One vendor alone now provides the EHR for roughly 42% of U.S. acute care hospitals and captured nearly 70% of new hospital contracts in a recent year, a level of concentration that makes lock-in a market-wide condition rather than a one-off risk.

The stakes go beyond IT budgets. In December 2025, the Texas Attorney General filed an antitrust suit against a major EHR vendor, alleging that the company leveraged control over hundreds of millions of patient records to impose steep penalty fees on hospitals that tried to use competing solutions. Whatever the outcome, the case is a sign of how seriously regulators are starting to treat data portability as a competition issue, not just a technical one.

The regulatory floor is rising, slowly

Federal policy has been pushing in the other direction for several years. The 21st Century Cures Act prohibits health IT developers from blocking data export, and the Office of Inspector General can levy civil penalties of up to $1 million per violation. Starting this year, the CMS Interoperability and Patient Access Final Rule requires Medicare Advantage plans, Medicaid managed care plans, and qualified health plan issuers to make claims, encounter, and clinical data available through standardized FHIR APIs.

These rules matter, but they set a floor, not a guarantee of easy portability. A system can technically comply with an export requirement while still making that export slow, incomplete, or expensive enough to function as a practical deterrent. One industry survey found that more than 60% of providers say they'd switch EHR systems if they could, while only about 15% are actively planning to in the next year, a gap driven less by satisfaction with their current system than by fear of the disruption and cost involved in leaving it.

Why this matters beyond the EHR

Most of the public conversation about vendor lock-in centers on electronic health records specifically, but the same dynamic applies anywhere clinical or operational data lives in a system built to make leaving expensive. Care management platforms are a clear example. If a health system's care coordination data, care plans, task histories, patient engagement records, only exists in a format native to one vendor's system, switching away from that platform later means the same kind of costly, disruptive migration project as an EHR switch, just on a smaller scale.

The way to avoid inheriting that problem isn't to pick the vendor that promises the easiest export today. Export promises are only as good as the format underneath them. HiiveCare is built on Medplum, a FHIR-native platform, which means every piece of data in the system, care plans, task histories, patient records, already exists as a portable FHIR resource rather than something that has to be converted into one on the way out. That's a structural difference, not a feature flag: it's not about whether an export button exists, but about whether the data was ever locked into a proprietary shape to begin with.

Choosing a platform that's portable by design doesn't just make a future switch easier. It changes the negotiating position a health system holds for as long as it stays, since the option to leave stays real rather than theoretical.

Sources: CaboLabs — Avoiding Vendor Lock-in in Healthcare · PLOS Digital Health — A problem of Epic proportion · EHR Source — Switching EHR Systems · Nirmitee — EHR Migration with FHIR Data Portability