Predictive Sciences

The Miami-Humana Health Services Research Center is at the forefront of new applications for information and prediction sciences traditionally used to identify the expected future costs of populations or individual patients.  Using Humana’s rich claims data asset to study health outcomes over time enables us to explore new realms of prediction as we work to enhance and validate applications to predict, based on individual behaviors, which patients are most likely to seek care, join an intervention program, be non-adherent to medication regimens, or to assume the responsibility of self-management. 

This work gives rise to new knowledge discovery tools which become essential components of understanding and influencing health behaviors.  This type of modeling inherently recognizes that health and health behavior is multi-factorial and social.  The diversity of the greater Miami population offers unique opportunities to use these methods to explore significant emerging issues in health care, such as disparities in health care and cultural differences in health behaviors. 

Another area of interest is the detection of what stage of the disease process is most subject to remediation through self-management efforts or health coaching efforts. There is a great deal of work yet to be done to create predictive capacity that can guide intervention and thus improve health and reduce expenses.