EHR Integration Blocks Startup Adoption
Brendan Keeler, Senior PM at Zus Health, on building infrastructure for digital health
The hard part in healthcare is rarely getting a doctor to like a product, it is getting a hospital to approve it, connect it to the EHR, and fit it into clinical workflow. In practice that means security review, contracting, and data mapping often kill demand before software ever reaches patients. This is why digital health startups increasingly sell to newer virtual care companies first, where the buyer and the user are much closer together.
-
Legacy hospital integration still runs through slow, messy pipes. Older HL7 interfaces were built for hospital to hospital data exchange, not modern developer workflows, while FHIR is improving the experience but remains unevenly implemented across EHRs. The result is that shipping one product to many health systems still means long setup work.
-
An integration layer became its own product category because startups kept losing deals at deployment. Companies like Redox sell a single connection point that translates among EHRs, clinical networks, and digital health apps, reducing the need to rebuild the same interface for every customer.
-
The buyer is often not the clinician. In large health systems, software choice is filtered through IT, security, and executive approval, so the winning product is often the one that clears procurement fastest and plugs cleanly into Epic or another incumbent system, not the one doctors simply prefer.
This pushes the market toward infrastructure first. More value will accrue to companies that compress implementation from months to days, whether through better APIs, reusable integration networks, or tools that automate security and approval work. In digital health, distribution increasingly depends on removing operational friction, not just building a better feature.