Beware of Swapping Two Healthcare Data Silos for Three

By Steven Berkow, Head of Value-based Care Market Strategy, InterSystems

Data Integration Is Key to Achieving Needed Breakthroughs in Healthcare

Healthcare is overflowing with data. So why has it not seen the data-driven breakthroughs in operations and customer experience that have already transformed banking and travel? For health plans, the answer is often hiding in plain sight: clinical and administrative data still live in different systems and “speak” different digital languages.

This divide is becoming harder to tolerate. Better data integration is now essential to advancing value-based care or, more specifically, improving care quality, efficiency, patient experience and access all at the same time. Fortunately, multiple mandates are spurring this integration, from CMS requiring Fast Healthcare Interoperability Resources (FHIR^®) APIs, to NCQA embracing digital quality measures, to enforcement of information blocking rules. Health plans, however, must guard against swapping these two longstanding data silos for several new ones.

The Unintended Consequence of Advancing Data Integration Use Case by Use Case

Leading organizations are now making great progress in integrating data from disparate sources to streamline prior authorization, close care gaps and improve coding accuracy. Unfortunately, these initiatives risk a new set of data silos. They are often run by separate work groups, funded by separate budgets and implemented on separate timelines. The end result: significant time and money are being spent on narrowly scoped data integrations with limited returns. These integrations are optimized for a specific purpose and consequently underperform when asked to support other use cases.

Building One Source of Digital Truth for Multiple Applications

Best-practice organizations are instead investing in a centralized data foundation with the requisite flexibility, extensibility and scalability to support multiple use cases. This approach minimizes redundant technology, storage and—more importantly—conflicting information on an individual patient or member.

Achieving one source of digital truth for a complex healthcare enterprise requires a pragmatic understanding of your organization’s current reality. Key systems that use a wide range of formats and standards and store information at the document level are often too costly to replace. Additionally, you cannot assume other functions will embrace new workflows. A centralized data foundation must not only ingest information from disparate sources but also make relevant data readily available to diverse end-users in the desired formats to move from a new system asset to system-wide impact.

Must-haves for such an all-purpose resource include:

• Ingests data in a wide range of formats and via multiple transfer protocols and modalities. While FHIR is gaining traction throughout healthcare, other formats and standards continue to dominate legacy systems. Likewise, centralized systems must ingest data across transport modes (push and pull), cadences (real-time and batch) and volumes (single record and bulk). Ideally, your system will also evaluate ingested data for missing elements so feeds can be expanded to capture them.
• Parses, indexes, deduplicates, harmonizes and normalizes ingested data. These are all essential components of the data transformation process in healthcare. Without all of them, discrete data elements cannot be analyzed at the patient or member level and pushed into relevant systems supporting operations and care management.
• Stores data in relational tables and integrates with preferred end-user interfaces. You should not expect other functions and departments to change their workflows. To ensure application, your centralized system must make relevant data readily available to existing departmental tools, systems and analytics engines.
• Scales with no erosion in performance or security. Your solution must continue to ingest and output data fast enough to support frontline operations as the number and complexity of data elements expand. It must also accurately manage consent and access for an ever-increasing number of individuals.

Avoiding New Data Silos Is More Than a Technical Challenge

Ensuring your organization does not swap two longstanding data silos for several new ones is not just about architecture. It is also about governance. Again, the teams in many organizations who are responsible for prior authorization, quality, risk adjustment, analytics and digital member experience operate with separate budgets, leadership structures and success metrics. This can make it difficult to build shared infrastructure, even when everyone would benefit.

Health plans making the most progress are often the ones that bring these groups together early, identify common data needs and invest in capabilities that serve multiple goals. This cross-functional alignment is what turns a technical asset into enterprise value. Healthcare does not need more disconnected data projects. It needs a more disciplined approach to integration, one that creates a durable foundation for regulatory readiness, operational efficiency and better care.

For health plans, the challenge is no longer whether to integrate data. It is whether to do so in a way that prevents the next generation of silos before they take shape.

This blog is authored by InterSystems, and the views expressed are solely those of InterSystems.

HL7^® and FHIR^® are the registered trademarks of Health Level Seven International and their use does not constitute endorsement by HL7.