Why does connected data still leave care teams searching for answers?
A patient arrives at a hospital. Somewhere in the healthcare network are their medical history, recent test results, and medication records.
The systems are connected. The data can be exchanged. Yet the care team may still struggle to piece together the healthcare information needed to make a decision.
This is the paradox of modern healthcare interoperability. We have made progress in connecting systems, but making the information useful remains a challenge.
And regulations alone won’t solve it.
That’s the opportunity beyond healthcare interoperability compliance. And it’s where AI could make a real difference.
What is healthcare interoperability compliance?
Healthcare interoperability compliance is meeting the regulatory requirements for securely exchanging and accessing health data across payers, providers, and patients. In the US, CMS-0057-F requires impacted payers to support standardized Fast Healthcare Interoperability Resources (FHIR) APIs, including electronic prior authorization, by January 2027.
The biggest challenges in healthcare interoperability
Connecting systems is only part of interoperability. Making the information usable is often harder.
- Fragmented data and legacy systems: Patient records sit across hospitals, labs, and EHRs that rarely come together, even when each has an API. This makes data integration across electronic health record systems difficult.
- Clinical meaning and terms: Payers and providers often use different code sets. So the same clinical data may be read differently across systems. This can happen even with standards like ICD-10 and SNOMED CT.
- Data in the wrong place: A procedure filed as an observation won’t be found by an automated prior authorization workflow.
- Multi-vendor complexity: Electronic prior authorization can involve several utilization management vendors, EHR platforms, and implementation-guide versions that don’t always align.
- Workflow integration: Clinicians often still switch between portals, even when the underlying systems are connected. Healthcare professionals may need to search multiple applications for medication history, lab results, and treatment plans.
- Data governance and consent: As more organizations share data, controlling access, confirming patient identity, protecting privacy, and managing consent become harder. Gaps in data management put individual patients’ privacy at risk.
- Medical device integration: Imaging relies on the DICOM standard, but monitors and ventilators often use proprietary formats that other systems can’t readily use.
- Adoption and change management: Teams used to document-based, manual processes often struggle to adopt end-to-end electronic exchange.
Solving everyday healthcare problems with interoperability
The value of interoperability is clearest in the problems it helps solve.
- Streamlining prior authorization: Electronic prior authorization pulls the required clinical information straight from the EHR, reducing manual document gathering.
- Strengthening care coordination: Care teams can be alerted when a patient is admitted to or discharged from the emergency department, so follow-up starts sooner. Access to medical records across the healthcare system can support more coordinated care delivery.
- Supporting value-based care: Timely clinical data helps teams act on patient needs earlier, instead of waiting for claims that arrive weeks or months later.
- Improving digital quality measurement: FHIR-based data can turn periodic quality reports into near real-time insight into care gaps. This supports data analysis, quality improvement, and evidence-based decisions.
AI-ready healthcare starts with interoperability
AI is often seen as a shortcut around messy healthcare data. In practice, it works best when the data is already interoperable.
- Standardized data makes AI efficient: HL7 FHIR codes carry their coding system, so AI can read data from many sources without costly reinterpretation.
- Standardized device data enables real-time care: Once monitor and ventilator data is converted to HL7, AI can support real-time monitoring in the ICU and operating theatre. Access to real-time data can help support timely clinical decisions.
- Trusted data makes AI reliable: AI needs relevant, well-governed, and monitored data, since ungoverned data creates privacy and regulatory risk. Strong healthcare data quality practices help ensure that AI works with high-quality data.
At Impelsys, this work powers our connected care solutions. We standardize device and imaging data. AI applications can then use it.
Without quality data, there are no quality outcomes from AI.
Beyond compliance: Where healthcare leaders should start
Going beyond compliance means building intelligence on top of interoperable data.
Compliance gets data flowing between providers, payers, and patients. For most healthcare organizations, that’s where the effort ends. The APIs work and the data moves, but very little of it is ever put to use.
A better approach is to treat that data as a lasting asset. Instead of collecting records for a one-time review and filing them away, store them in a FHIR-based repository. Track where each record came from, and use it across programs. The same data that satisfies a mandate can then close care gaps, power quality measures, and answer a CMS audit with confidence.
That’s where the real return on interoperability lies. It’s also what makes data ready for AI.
Getting there doesn’t start with technology. It starts with a problem worth solving.
- Start with a pain point: Find where better information sharing would make a clear difference. For example, look at prior authorization delays or data gaps in a value-based care program.
- Pilot a few high-value use cases: Get them working reliably end to end, across every vendor and EHR involved, before scaling further.
- Fix data at the source: Ensure clinical information is coded correctly and stored where automated workflows can find it. Accurate data collection helps preserve clinical meaning.
- Invest in adoption: Train clinical and administrative teams. Redesign workflows so electronic exchange reduces effort, not adds steps. This helps health information technologies fit into existing clinical workflows.
- Assess data before AI: Export FHIR data from a certified EHR. Measure it against frameworks like PIQI. Do this before starting an AI pilot.
- Build for reuse: Design each project so the next requirement or use case can build on it.
Conclusion
The January 2027 CMS interoperability deadline will get more healthcare data moving than ever before. But moving data was never the real goal.
The goal is getting the right information to the people who need it, when they need it. That means data that is accurate, built into everyday workflows, and trustworthy enough for AI to use.
Organizations that start with real problems, fix data at the source, and plan for the long term will gain more.
They will get more from interoperability than a simple compliance checkmark. They’ll be ready for what AI can do next.
Authored by Sherlin Jannet Arputharaj.
