Cleaning Up Healthcare’s Dirty Little Secret

In our previous post, we mentioned the challenges payers face as they transition to FHIR.

The first of these challenges is simply how quickly the date of enforcement is arriving. By 2021’s first deadline, payers must go live with two different APIs—one providing members with access to their own clinical and claims data and the other delivering public-facing directory services for provider networks and drug formularies.

The second major challenge revolves around the quality of the clinical data that payers present to members. Health plans must use the FHIR standard to make these five types of data available: 1) adjudicated claims; 2) encounters with capitated providers; 3) provider remittances; 4) enrollee cost-sharing; and 5) clinical data, including laboratory results.

Some payers have previously implemented systems to process the first four types of data, which can be mapped to three parts of FHIR, specifically: explanation of benefits, patient, and coverage resources. The fifth type, clinical data, is generally the most valuable data for patients but also represents the most significant potential roadblock for payers. That’s because the success of FHIR as an exchange format depends on the data being high quality and appropriately structured.

Because clinical data is often incomplete, redundant, or inconsistently coded, its value for patient access via FHIR can be limited. Information that is poorly formatted in the source EMR will render incomplete as a FHIR resource. In many cases, clinical data is not structured or codified, making it difficult to understand. For example, critical measures such as blood pressure or blood test results indicating diabetes are recorded in the unstructured “notes” field of EMRs. Even when data appears in structured fields, it often lacks the detail required to achieve true semantic interoperability.

Not surprisingly, adoption of the new standards will have its leaders and laggards. Payers know they must have FHIR-enabled clinical and claims data so they can share it with their member population. But when it comes to ingesting that data, some institutions are only going to have basic capabilities. They may still be reliant on HL7 2.X and 3.X and not yet have FHIR capabilities within their EMR. Therefore, until there is maturation across the ecosystem towards new sophisticated technologies, payers must be prepared to meet data sources where they are.

Verinovum Cleans Your Data and Enables Interoperability

Due to the complexity of the healthcare continuum, the myriad of vendors and rules, and obsolete processes that may not take into account end uses, only about 14% of all clinical data is actionable. Verinovum empowers payers, providers, and other healthcare organizations by using a unique near real-time data curation and enrichment process that makes nearly 89% of clinical data actionable. Verinovum does this by focusing on the quality and completeness of data across the entire continuum of care, thereby unleashing the full potential of measurements, clinical outcomes, and innovation to deliver better patient outcomes and avoid unnecessary costs.

Each patient’s digital footprint is unique, and the information may be located in more than a half dozen different silos, each of which has different formats, modes of transmission, and rules for use. Most of them are required by CMS to use HL7, but due to local customizations, a lot of the data isn’t structured or semantically normalized in a way that makes it truly cohesive. Some of the data that’s incomplete or incorrect across silos of information is relatively minor, such as incorrect date/time stamps. But other discrepancies are far more complex, like having the wrong coding for a specific quality outcome. Verinovum solves this problem by working with payers to create a comprehensive view of the patient’s health, helping payers proactively intervene on a patient-by-patient basis.

In the past, providers and payers were looking at PDF printouts or directly at native EMR data on computer screens. Sometimes, they could solve discrepancies that might exist across multiple silos of information. Unfortunately, when you’re trying to piece data together for machine-readable applications and next-generation interventions to support advanced outcomes like artificial intelligence, machine learning, and case management interventions, the data has to be normalized. If you don’t have that in a machine-readable format that’s fully functioning, you end up with real discrepancies in treating those patients.

With the appropriate layers of data quality scoring and data quality improvement techniques, Verinovm can eliminate the burden of dirty information and make your data far more accessible to both the human eye as well as to the machines that are helping optimize your outcomes. Moreover, payers that go beyond just implementing API layers to support patient access can effectively enable Verinovum’s entire ecosystem of applications to serve the data needs across their business lines with FHIR enabled data. Having this level of robust information is a major strategic advantage, enabling internal stakeholders and empowering patient populations to better manage their own care.