Update from the 2nd NCQA Digital Quality Summit, November 14-15th, Washington DC
Chun Li, PhD
Chief Data Scientist
Conference participants include providers, health plans, policy makers and technology vendors like Diameter Health. In addition to several keynote speeches delivered by industry leaders, the conference was organized around three tracks: point-of-care performance and quality, measurement-core elements, and dissemination and decision support. Within each track, there are workgroups focusing on a specific sub-area, and working through the challenges and proposed solutions during the 1 ½ day conference.
John D’Amore served as the co-lead for the data quality for quality measurement workgroup in the point-of-care performance and quality track. Their discussions started with the “who” (should be responsible for data quality), “what” (are each stakeholder’s role in quality control), and “how” (to measure the data quality outcomes). While they reached no consensus on the need for some quantitative rubrics, they focused their discussions mainly on the first two aspects. For the “who” aspect, they identified 8 stakeholders playing a role in ensuring data quality. This includes the patient, individual clinician, care delivery organizations, EHRs, HIEs, measure technology vendors, health plans, and policy makers/measure authors. For the “what” aspect, they believe that each stakeholder should aim for a set of different quality rubrics to ensure a safe exchange of data transmissions. For example, the data quality requirement for a measure technology vendor, such as Diameter Health, would focus on ensuring accurate data translation and compliance to standards and guidelines. For an HIE, on the other hand, it’s important to keep good data provenance for where the data is coming from.
Chun Li participated mostly in track 3 discussions – dissemination and decision support – with a focus on auditing requirements for the new HEDIS Electronic Clinical Data System (ECDS) Reporting. As highlighted in NCQA’s president Margaret E. O’Kane’s opening remarks, ECDS is an initiative by the NCQA to harmonize measures and make data more assessable and interoperable. Since this is the first year and ECDS data is required to be audited, there are a lot of questions around the process and potential impact. For eMeasure vendors like Diameter Health, the primary question is where our current eMeasure certification fits in the ECDS paradigm, especially regarding specific auditing requirements. We were pleased to learn that as a supplier of standard supplemental data, the data produced by our technology is an eligible standard supplemental data source for ECDS reporting with minimal auditing burden. This determination is congruent with the current benefit of standard supplemental data that bears minimum auditing requirement when used for HEDIS reporting.
In addition, Chun participated in another interesting workgroup that focused on utilizing artificial intelligence (AI) and natural language processing (NLP) for Quality measurement and improvement. The group has identified specific use cases where AI/NLP offers significant advantage. For example, for new measures, hybrid measures (those requiring information from both claims and clinical data), and complex measures, it can be challenging to understand the expected location and /or format of the required data elements, especially those from different data sources. AI, on the other hand, has the potential to learn and grasp the relevant information from various clinical domains, in various formats (e.g., clinician’s notes, answers to questionnaires) to inform the measure outcomes. The group’s conclusion was AI can play a very helpful role in healthcare and is not limited to quality measures.
In summary, the takeaways for Diameter Health clients and those interested in using clinical data are:
– Clinical data are increasingly being used to support payer-based quality measurement (e.g. HEDIS) and will be a requirement in the future
– Data from EHRs and other source systems vary substantially and require translation and normalization to make work for quality measurement (here’s Diameter Health’s recent paper on this topic: https://www.thieme-connect.com/products/ejournals/pdf/10.1055/s-0038-1656548.pdf )
– Data provenance and source tracking will be key parts of auditing of clinical information
Overall, it was an informative conference for Diameter Health to continue learning and sharing opinions on important topics and growth of the industry, as well as to understand the differing perspectives from providers, payers, and policy makers on quality measures.