Effective blood product safety monitoring faces many obstacles, including difficulty detecting unreported adverse reactions to blood transfusion and difficulty recognizing actionable patterns in the sheer volume of potentially relevant clinical data due to lack of standardized reporting. For example, signs and symptoms of transfusion reactions may be buried in the free text of nurses’ notes and vital signs flowsheets, making automated review problematic. Access to administrative data and missing patient information are additional challenges.
The U.S. Center for Biologics Evaluation and Research (CBER) is addressing these obstacles through its Biologics Effectiveness and Safety (BEST) system. Launched in October 2017, this program seeks to expand and enhance access to new and better data sources, methods, tools, expertise, and infrastructure to conduct surveillance and epidemiologic studies. Standardized data models such as the HL7 Fast Healthcare Interoperability Resources (FHIR) and secure data exchange protocols such as "SMART on FHIR" can facilitate the development of innovative approaches to utilizing electronic health records (EHR) effectively in order to establish semi-automated adverse event identification, validation, and reporting. To solve the challenge of embedded EHR text, natural language processing and other machine learning methods are potentially powerful approaches. The future in this area is very promising, but success will require close collaboration between numerous disciplines, including clinical stakeholders, “big data” scientists, and informatics experts.
In this session, we will define the basic concepts related to “big data” analysis, including data management, interoperability, exchange of protected health information, and machine learning. We will then discuss how these concepts are currently being employed in the application of blood safety through the BEST system. Finally, we will address successes and challenges of this approach, as well as the potential benefits of public-private partnerships in blood safety and beyond.
Discuss concepts and tools used in large-scale data analytics and how these are applied to data extraction and monitoring for blood safety.
Describe CBER’s Biologics Effectiveness and Safety (BEST) program.
Discuss prototype development of the BEST program within a hospital system, including benefits and obstacles.