Predictive analytics can improve clinical and operational decision making by identifying meaningful patterns in complex electronic health record (EHR) and claims data. The LHS Program brings scientists together with care delivery partners to design and implement predictive models to inform proactive outreach and improve care, with a special focus on members at high risk of health complications.
During the 2018/19 flu season, and LHS Program team piloted a predictive risk model to identify members who were at high risk for influenza-related complications, followed by outreach to encourage vaccination for these members. This brief, low-cost, targeted intervention was associated with increased vaccine uptake among high-risk patients. Based on these findings, Kaiser Permanente Washington has adopted this intervention systemwide. (See story by KING 5 News, Dec. 2, 2019.)
When COVID-19 hit, we leveraged our analytics capabilities to support Kaiser Permanente Washington's rapid response to the pandemic. Utilizing existing infrastructure from our flu vaccination intervention, we were able to rapidly deploy outreach efforts to Medicare members to make sure their care needs were being addressed. We then quickly expanded this to include outreach efforts to non-Medicare members with complex medical and social needs and other chronic conditions. Most recently, we launched work to identify member characteristics that might help us understand patterns of delaying care due to fear of contracting COVID-19 during the height of the pandemic.
Through the LHS Program, other predictive models have been developed to:
Our ongoing work in 2021 will focus on evaluating predictive models for sepsis, mortality, suicide risk, and opioid use disorder/opioid risk and advise care delivery partners on their implementation.
How our Learning Health System Program is using statistical methods and machine learning to respond to COVID-19. Healthy Findings, August 2020.