Better interoperability… Yet poor clinical data quality could cause a “mis-FHIR”

This blog was co-authored by Mike Noshay, Chief Strategy and Marketing Officer of Verinovum, Michael Barr, MD, President of MEDIS, and Rick Moore, PhD, Chief Executive Officer of MTC Group, LLC.

The U.S. healthcare system’s information technology infrastructure is a house of cards when it comes to interoperability and clinical data usability.

Whether intentionally designed or not, the US healthcare system built this house with multiple, mostly siloed technologies such as electronic health records, payment clearinghouses, registries, information management systems, health information exchanges, and so on. Efficiently sharing patient and member information among clinicians, practices, payers, and patients is an ongoing challenge and this house of cards often collapses due to complex interoperability and data quality issues.

Many pioneering organizations have been focused on solving these challenges, some for decades. Institutions like HL7 International, which developed HL7 v2, HL7 v3, and HL7 v4 (FHIR) standards, and the Regenstrief Institute, with its LOINC codes, have created the framework for the transmission and standardization of healthcare data. Despite these standards, healthcare system stakeholders continue to struggle with retrieving accurate information, sharing information efficiently, and creating usable, high-quality data.

In 2011, the Centers for Medicare and Medicaid Services (CMS) provided incentives for the adoption and meaningful use of certified electronic health records technology through the Medicare Promoting Interoperability Program (formerly the Medicare and Medicaid EHR Incentive Program). The meaningful use program, incentivized by the American Recovery and Reinvestment Act, has been met with lukewarm reception over time. The costs for installing, maintaining, and using these systems have been notable. Despite the adoption of certified electronic health records technology by 94% of general acute care hospitals and 78% of office-based physicians, these systems, while helpful for capturing patient information, do not provide the necessary standardization needed for codifying or exchanging that information between systems.

Equally, if not more importantly, the systems do not address the quality of the data ingested by the system. The information found within an EHR is only usable if it is present, coded correctly, accurate, and accessible in the clinical record. In Verinovum’s experience only 40% of the EHR data we receive is usable for many important operational and clinical use cases.

While CMS was establishing the EHR Incentive Program, Graham Grieve, along with many collaborators, was writing the First Resources for Health. Over the span of ten years, this system evolved into the Fast Healthcare Interoperability Resources standards, often shortened to FHIR® (pronounced “fire”). Developed by HL7 International, FHIR provides standardization protocols for medical message information segments, called resources and the means to exchange data via application programing interfaces (APIs).

FHIR adoption is a current priority of the Biden-Harris administration, but it’s not a panacea. FHIR enables API-driven retrieval and transmission of data formatted to established standards, but it does not address problems with data quality further upstream — a “mis-FHIR” that could cause the health data interoperability cards to collapse and therefore limit the effectiveness of FHIR to support a core need of value-based payment models – better differentiation of quality performance through better (digital) quality measures.

Participation in programs like the Medicare and Medicaid Promoting Interoperability Program is predicated on reporting EHR-derived electronic clinical quality measures. Health plans submit Healthcare Effectiveness Data and Information Set, or HEDIS®1, measures to the National Committee on Quality Assurance to demonstrate plan performance. Many of these measures are calculated using EHR data. And for Medicare Advantage plans, several of these measures are used for performance payments known as the CMS Star Ratings program.

That data, however, whether used for quality determinations or transmitted to another institution or patients, is likely to be inaccurate and incomplete. This leads to the question, should the data be fixed at the source in every EHR in every practice, health system, and hospital? Or is there a way to effectively and securely curate, normalize, and standardize data outside these systems to create data usable for multiple purposes?

We believe that Verinovum’s Data Curation as a Service (DCaaSSM) methodology is more expedient and efficient. There are too many health systems and electronic health records for a scalable fix at every source, which is why we recommend a two-phase approach to improve data quality.

During Phase One, Verinovum evaluates the quality of the data from a provider via the following questions:

  • Are messages complete?
  • Are message fields accurate?
  • Are there local codes to be translated?
  • Can the data be standardized?
  • Are patients identifiable?

Advancement to Phase Two involves the application of hundreds business-defined rules to enhance data usability by as much as 40%, while also ensuring conformance to:

  • Standard data science best in class practices
  • United States Core Data for Interoperability
  • Client-specific use cases

This two-phase approach provides value from the increase in data quality to as much as 90% for use cases such as HEDIS® or CMS Star reporting. Data will be more accurate and complete no matter the format — HL7 v2, HL7 v3, or FHIR.

We need enhanced clinical data quality and usability to support value-based payment models through better data and better (digital) quality measures. Once the data is cleaned, standardized, and normalized, there will be less likelihood of a “mis-FHIR” and more opportunities to drive improvements in care based on reliable, actionable, more robust assessments of quality.

For more than 30 years, Dr. Michael Barr has delivered programs, policies, and services aimed at equitable, accessible, evidence-based, and person-centered care. As President and Founder of MEDIS, Dr. Barr delivers personalized, client-driven healthcare consulting services with objective-oriented strategies and support for organizations and individuals.

As NCQA’s former Executive Vice President of Quality Measurement and Research, Dr. Barr oversaw performance measurement development, research and analysis, management of contracts and grants, and NCQA’s consulting activities. He also led the digital measures strategy for HEDIS® and contributed to public policy, educational programs, and other initiatives.

Rick Moore, PhD is an award-winning, board-certified, healthcare strategist with more than 25 years of experience in the federal government and non-profit sectors. He is Founder and CEO of MTC Group, LLC, where his focus is on uniting teams to drive efficient data collection and management through the design, implementation, and operation of informatics infrastructures.

As NCQA’s former Chief Information Officer, he was responsible for the vision and strategic direction of the Information Services, Information Technology, and Information Products. In addition, he is an expert in privacy/security-based compliance with federal mandates, including HIPAA and HITRUST, and led the only program in the country to receive Office of the National Coordinator for Health Information Technology approval as a quality measurement test system.

1HEDIS® is a registered trademark of the National Committee for Quality Assurance (NCQA).