What We Learned About Scaling Clinical Data for HEDIS

Over the past year, we have supported health plans in consuming new data sources and optimizing the utilization of available clinical data. Our comprehensive approach to integrating new clinical data sources prioritizes rigorous data validation, identity management and primary source verification, tailored to meet the specific objectives of the health plan.

During these client implementations, we encountered several challenges that could potentially hinder health plans in accurately reporting for HEDIS measures.

Challenges in Utilizing Clinical Data

Our process involves validating each entry in Fast Healthcare Interoperability Resources (FHIR), ensuring it includes essential components such as date of service, structured coding and a status indicating the service was performed. We encountered several challenges with the health plan clinical data in this area:

• Structural complexity of FHIR standard: The FHIR standard, designed to facilitate the electronic exchange of healthcare information, introduced complexities in structured database queries. Thestandard allows for a variety of data types to be associated with each resource type, which provides a great deal of flexibility, but challenges more established data constructs. For instance, an observation resource could reference another code, as in the presentation of structured responses to a survey as a codeable concept, or a numerical value like a systolic blood pressure or a string value such as an HbA1c value of "7 %" or a string such as "normal." Our team developed parsing strategies to interpret the variety of datatypes appropriately.
• Standardization of Structured Data: Many records lacked standard coding compatible with HEDIS value sets. These records, often sourced from third-party providers, were consistently structured but presented challenges in translation to standard codes. Our thorough approach enabled us to bridge these gaps effectively.
• Interpretation of Blood pressure readings: Blood pressure readings, presented as separate entries for diastolic and systolic measurements using the same generic LOINC code, required refinement to enhance data clarity and alignment with NCQA's (National Committee for Quality Assurance) HEDIS value sets. Collaborating closely with the health plan and their electronic medical record (EMR) vendor, we facilitated the adoption of specific LOINC codes for diastolic and systolic readings.
• Inconsistent immunization records: A significant portion of immunization records indicated a status of "not-done," failing standard business rules for HEDIS inclusion. However, further analysis revealed additional attributes, such as lot numbers, suggesting the immunizations were administered. Working collaboratively with the health plan and their EMR vendor, we refined transformation logic to accurately reflect immunization status.

Navigating HL7 CCD Format

For supplemental data standardized on the HL7 Continuity of Care Document (CCD) format, we encountered similar challenges.

• Missing Member Identifiers: Despite the CCD standard allowing for insurance identifiers to support member identity, many files lacked this crucial information, leading to unattributed data. Through innovative matching techniques based on member demographics, we successfully attributed member IDs to the majority of records, with a high success rate.
• Inconsistent use of standard value sets: Challenges analogous to those encountered with FHIR entries were observed in processing CCD data. Collaborating with vendors providing CCD files, we successfully addressed issues such as the association of blood pressure readings with generic LOINC codes, ensuring compliance with reporting standards.

Ensuring Data Integrity through Primary Source Verification

Before fully operationalizing the process, each EMR data stream that was not provided through NCQA's data aggregator validation (DAV) program was subject to primary source verification (PSV). We then delivered comprehensive supplemental data impact reports, tracing each numerator and exclusion back to its source. This ensured transparency and accuracy for data sourcing. As a final step, auditor approval was obtained. This ongoing process enables seamless delivery of care gap reports at regular intervals.

Positive Impact on HEDIS measures

Analysis of data from the health plan's FHIR server revealed positive impacts across 33 HEDIS measures. Notably, we observed substantial improvements in Blood Pressure Control for Patients with Diabetes and Weight Assessment and Counseling for Children - BMI Percentile, with impacts of up to 17%.

If you would like more information on supplemental data read our blog Navigating Supplemental Data for HEDIS Measures, where we explore this topic in more detail.

Conclusion

Our extensive experience with HEDIS measurement proved invaluable in navigating and resolving these challenges. Effective collaboration and expertise in managing complex data ecosystems are paramount for optimizing healthcare quality measurement.

For a specific example, download our case study where we show how one health plan improved the quality of data and applicability to HEDIS measures, including steady improvement of Star Ratings in their Medicare HMO. Contact us today to engage our expertise for your HEDIS program.

Public link

Please use the above public link if you want to share this noodl on another website.