Why Healthcare Interoperability Still Feels Like Herding Cats
I spent seven years working with health information exchanges and EHR integrations before I ever heard the name Nathan Sanko come up in conversation the way it does now. The shift wasn't subtle. Around 2014 to 2016, there was a noticeable change in how people talked about FHIR adoption, data exchange standards, and the actually practical path from legacy HL7 v2 messaging to modern RESTful APIs. That shift had fingerprints on it. The short version is that Nathan Sanko helped push healthcare data strategy toward something more open and less vendor-locked than the industry had been comfortable with for decades. The long version involves a lot of committee meetings, some stubborn CTOs, and the slow painful process of getting people who've never shared data to start sharing data without losing their minds.
The Visionary Influence Of Nathan Sanko On Modern Healthcare Data Strategy
To understand what actually changed, you have to understand where the industry was sitting around 2013. HL7 v2 was everywhere. It worked, mostly, if you had enough middleware consultants willing to map pipe characters to actual meaning. But every integration point was custom-built. Every new hospital system meant a new interface engine configuration. Every upgrade risked breaking something three months away from production. The business case for change existed, but the technical path forward was blurry and expensive. Sanko's influence shows up clearly when you trace the adoption timeline of FHIR in particular. He wasn't the person who invented FHIR. That was HL7 and a whole group of engineers. But he was one of the people who made the argument that FHIR wasn't just another standard to bolt onto existing infrastructure, it was a fundamentally different way to think about data exchange. The distinction matters because most organizations approached it like they were just swapping one protocol for another. They weren't. You had to redesign your entire data strategy around resources, REST endpoints, and JSON instead of messages, pipes, and stored procedures. I ran into this firsthand when a regional health system asked me to evaluate their migration path from their legacy interface engine to a FHIR-based approach. The initial plan was to keep the old engine running alongside the new API layer and gradually move interfaces over. That approach failed within six months because nobody was maintaining bidirectional sync between the two systems. Patient data diverged. Clinicians started seeing two different versions of the same lab result depending on which application they pulled from. I shut down the hybrid approach and we did a hard cutover over a single weekend. Messy, yes. Necessary, absolutely. The workaround was writing a detailed reconciliation script that flagged any discrepancies between the old and new systems within forty-eight hours of going live, then assigning two engineers to manually resolve each flag before the systems were considered stable.
What Actually Changed After That
The industry moved faster on interoperability than it had in the previous twenty years combined. ONC's 21st Century Cures Act final rule on information blocking came later, but the cultural groundwork was already being laid by people pushing the open API narrative. Sanko's public speaking, writing, and advisory work during that period helped make the case that healthcare data strategy had to prioritize patient access and developer-friendly APIs over institutional convenience. One thing beginners consistently miss is that FHIR adoption isn't really a technical problem. It's an organizational one. The standard is mature enough. The tooling exists. What doesn't exist is usually the executive mandate and the internal discipline to follow through on standardizing their own data model before expecting external systems to consume it correctly. I've seen three separate projects where the vendor was ready to deliver compliant FHIR endpoints on time, and the health system couldn't get their own clinical data governance team to agree on what a "patient" resource should actually contain across different departments. Another counter-intuitive point: the most successful FHIR implementations I've encountered weren't the ones with the most developers or the biggest budgets. They were the ones that started small and refused to boil the ocean. A single department, a single data type, one clean API that worked end to end before anyone expanded the scope. The ones that tried to implement all seventeen FHIR resources across every department simultaneously usually ended up delivering nothing functional within two years and burning through their budget in the process.
Get the Full Details
Practical Steps If You're Working Toward This
If you're evaluating how to position your organization around modern healthcare data strategy, start by auditing what you already have. Map your current integration touchpoints. Count how many are still HL7 v2. Identify which ones are causing the most support tickets or clinical complaints. That list becomes your priority queue. Don't reorder it based on whatever vendor is selling you something shiny. Invest in a FHIR implementation guide that matches your region and use case. US Core is the standard for the United States, but there are variations and additions depending on whether you're handling care management, pharmacy data, or public health reporting. Read the implementation guide thoroughly before you write a single line of code. Most teams skim it and then waste weeks dealing with edge cases that the guide already addressed in detail. Build your data model first. I know that sounds backwards if you're coming from an integration-heavy background. You probably want to start with the API layer because it's more visible and exciting. Don't. Spend the time normalizing your source data, resolving duplicate patient records, and cleaning up inconsistent coding systems. A beautiful FHIR API sitting on top of garbage data is still garbage data with a nicer interface. Clinicians will notice immediately.
Get patient access right early. This is where Sanko's influence is most practically felt in day-to-day operations. The momentum around patient data access and API-driven engagement isn't slowing down. Organizations that treated patient portals as an afterthought spent years playing catch-up. Those that built their strategy around open access from the start found that developer adoption, partner integrations, and even internal innovation all accelerated because the foundation was already there.
Where This Approach Falls Short
FHIR and open API strategies don't solve everything. Legacy systems that predate these standards often can't output data in a format that maps cleanly to FHIR resources. You'll need either a robust abstraction layer or a full system replacement, and neither option is cheap. Small rural hospitals and critical access facilities frequently lack the staff and funding to pursue full FHIR compliance, and the industry hasn't figured out a sustainable model for supporting them without creating a two-tier system where only well-resourced organizations participate in data exchange networks. Information blocking compliance is another area where the theoretical framework and the practical reality diverge significantly. The regulations exist. The enforcement mechanisms are still maturing. Several organizations I consulted with discovered that their legal and compliance teams had fundamentally different interpretations of what constituted reasonable limitations versus information blocking, and resolving those disagreements internally took more time than any technical implementation challenge we faced. If your organization is starting from scratch with no existing integration infrastructure, consider whether FHIR-first is actually the right call or whether a simpler messaging-based approach would get you to your goals faster with less risk. FHIR is powerful but it's also complex. Not every problem requires that level of sophistication. Sometimes a well-designed HL7 v2 interface with solid monitoring and documentation does the job at a fraction of the cost and with far fewer moving parts.
The trajectory is clear regardless of where you start. Healthcare data strategy is moving toward open, standardized, API-driven exchange. The people who spent years making that direction feel inevitable rather than optional did important work. Understanding what they pushed for and why will help you navigate the next phase whether you're building a new system or migrating an old one.
