Reducing Harm with Predictive Modeling and Impact Analysis

The effective use of data is critical in the field of health care. Southcentral Foundation has been refining and improving its data systems for many years, including establishing our Data Services department, creating a central repository where data without PHI is accessible to all employees, and creating web-based tools to make data easier to use. One new area where SCF has begun using data is for predictive analysis and impact evaluation, to be able to act proactively before a customer reaches a crisis point. Predictive modeling helps SCF calculate risk for customer-owners, and impact analysis allows SCF to evaluate interventions to determine their effectiveness.

So far, there is one main area where SCF has been using these methods. Since SCF has determined that providing home health services can reduce ER usage and hospital readmission, being able to predict the impact of these services can help target where home health may be best applied. Customer-owners can have potential health issues addressed and receive support for overall wellness through home-based services before an ER or hospital visit becomes necessary.

Another area where SCF may use predictive modeling in the future is for suicide prevention. Although SCF is not currently using predictive modeling for this purpose, SCF undertook two research projects related to algorithms that may be useful for suicide prevention in the future⁽¹⁾. This modeling could allow SCF’s primary care teams to have conversations with customer-owners before they reach a crisis point.

By using data to predict when customers may need support, and evaluate the impact of interventions from the health care system, SCF is working to support customers in achieving wellness, and reduce the utilization of the most costly parts of the system.

Interested in finding out more about SCF’s innovative data systems? Sign up for the Nuka Conference now! The conference features sessions from SCF’s subject matter experts on integrated care teams, integrated behavioral health, data collection and usage, workforce development, improvement and innovation, trauma-informed care, and much more. If you have any questions about the conference, or SCF’s Nuka System of Care, feel free to contact the SCF Learning Institute.

• The research reported was supported by Alaska IDeA Network for Biomedical Excellence and the National Institute of General Medical Sciences of the National Institutes of Health under award numbers P20GM103395 and S06GM123545. This work was also made possible through partnership with the Mental Health Research Network, funded by the National Institute of Mental Health under award number U19MH121738. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.