P214: Leveraging Real-World Data for a Culture of Learning in Health Systems: A Landscape Assessment
Poster Presenter
Matt D'Ambrosio
Policy Analyst
Duke-Margolis Institute For Health Policy United States
Objectives
This study explored how learning health system principles have been integrated into health system processes—specifically how systematic collection of real-world data translates into continuous learning that informs clinical and operational decision-making.
Method
This study was conducted over the course of late 2022 and early 2023. An online survey was distributed to health system employees (including providers, researchers, administrators, and technical staff), and a subsample of respondents were interviewed via video call.
Results
This study collected data on health system characteristics (e.g., mission, geographic location, clinical processes, etc.) and on how/whether real-world data is collected and used within the health system at various levels (e.g., leadership decision-making, provider decision-making, patient reported outcomes, etc.).
Across all types of health systems, electronic health record (EHR) and claims data utilization is prominent. In addition to EHRs and claims, survey respondents frequently indicated laboratory data, patient-reported outcome data, and registry data as widely collected and/or utilized in health system processes.
Several emergent themes and takeaways were identified from survey responses and interviews. A learning health system (LHS) involves the continual integration of available data and metrics to achieve predetermined goals or benchmarks. LHSs are driven by a variety of real-world data sources. Incentives are necessary to ensure different groups within a health system are aligned and fully engaged on LHS principle implementation. Routine feedback and open communication with a receptive leadership group is critical for LHS success regardless of the context or type of LHS principles that are pursued. Barriers to implementation include lack of interoperability between internal and external data platforms, insufficient and unrepresentative patient sample sizes within individual health systems, low engagement with particular stakeholder groups (e.g., providers or patients), and lack of expertise with data curation and informatics.
Conclusion
Efforts remain underway to support the notion and practice of leveraging RWE to achieve a culture of learning within and across health systems. Though many health systems are making efforts to become learning systems, our landscape assessment of health system practices shows that specific organizational, interpersonal, and structural elements are critical to achieve this goal.
Key elements include a robust data infrastructure built on internal and external interoperability, a culture that prioritizes research and learning, and incentives that encourage sustainability. Federal agencies, policymakers, and other RWE policy stakeholders must strive to meet health systems where they are in their LHS journey to support these efforts.
Learning health principles that are established must include a cultural reset within health care systems. Learning health system cultures should embody a clear mission and vision championed by all health system stakeholders and driven by leadership to deliver evidence-informed care, which achieves sustainable practices toward patient wellness and health system operations. This culture should include, at minimum, addressing burnout among health care providers who use evidence- generating and -based solutions.
As more health data is created and leveraged for a wider range of decision making, it is important that patients know how their data is used. As learning health concepts proliferate, health systems have an obligation to share enough information for patients to make informed decisions, even in circumstances where informed consent might not be required.
