Beyond Interoperability: Why Healthcare Needs Shared Understanding in the AI Era

In the age of artificial intelligence (AI), healthcare has made significant strides in interoperability-systems can now communicate with each other more seamlessly than ever before. However, this progress masks a deeper issue: the lack of shared understanding about what the data actually means. This gap is particularly evident in medical coding, where clinical interpretation and financial outcomes often diverge, raising concerns about accuracy and fairness.

A recent report by the BlueCross BlueShield Association (BCBS) highlights the stakes. The study identified $663 million in additional inpatient spending due to increased coding intensity, driven in part by AI technologies. This has sparked a debate: are these tools introducing inaccuracies, or are they finally capturing patient complexity that was previously underrepresented? The answer lies not in the technology itself but in how we interpret and contextualize the data.

Bridging the Context Gap

The disconnect between clinical documentation and revenue cycle management is a prime example of this context gap. Health systems configure their electronic health records (EHRs) differently, including variations in documentation templates, order sets, problem lists, coding workflows, and data mappings. Providers also document care differently, leading to operational behaviors that introduce variation across the ecosystem.

Clinical documentation is narrative and contextual, capturing the nuances of a patient's condition and treatment. In contrast, revenue cycle data is abstracted and optimized for payer-specific reimbursement. By the time clinical data is transformed for billing, the same patient story can be represented very differently across systems. This discrepancy fuels tension between payers and providers over accuracy, "coding intensity," and suspicions of over-coding for higher reimbursements.

For instance, consider a patient with diabetes and complications. One system might code this to meet minimum payer requirements, while another could capture the full complexity of the condition, leading to different financial outcomes. This disparity not only affects billing but also impacts care coordination and patient safety. Without a shared understanding of what the data means, healthcare organizations struggle to provide consistent, high-quality care.

What Comes Next

To address this challenge, healthcare must move beyond mere interoperability and focus on creating a shared understanding of data. This involves standardizing clinical documentation practices, aligning coding workflows across different systems, and ensuring that AI tools are transparent and context-aware.

One potential solution is the development of standardized clinical documentation templates that capture both narrative details and structured data. These templates can help ensure consistency in how patient information is recorded and interpreted. AI technologies can be designed to provide explainable insights, allowing clinicians to understand the reasoning behind coding decisions and adjust them as needed.

Collaboration between payers, providers, and technology vendors is crucial. By working together, these stakeholders can develop common standards and best practices that promote accurate and fair reimbursement while maintaining high-quality care. This collaborative approach will also help build trust in AI-driven systems, making it more likely that they will be adopted and used effectively.

Ultimately, the goal is to create a healthcare system where data not only moves freely between systems but is also interpreted consistently and accurately. By achieving this shared understanding, we can ensure that AI technologies enhance rather than complicate care, ultimately improving outcomes for patients and reducing administrative burdens for providers.