Tuesday, July 15, 2014
Given the demand for better business data and the emergence of enterprise Electronic Health Records (EHRs), healthcare organizations have been embracing the potential of Big Data to spur cost savings, improve care outcomes, and drive greater business value. Similarly, there has been a surge of techniques involving Geographic Information Systems (GIS) that can prove valuable when their integration, spatial visualization, and geospatial analytic powers are properly harnessed. Duke Medicine has embraced these trends by implementing an enterprise EHR and supporting analytics team that serves the geospatial analytical needs of the organization. Additionally, Duke Medicine recognizes the large paradigm shift that has influenced our society to become heavily reliant upon on-demand information, requiring sophisticated, automated processes and robust hardware and software solutions that can quickly deliver information, rather than raw data. Consequently, Duke Medicine has incorporated geospatial information within its EHR, which adds thousands of new “big-geo-data” elements to a patient’s clinical record. Technical solutions are being developed in order to enable on-demand visualization and analytic capabilities that will help provide greater insight to patients’ risk, environmental history, and ability to be compliant with a care plan, given the sociocultural factors that may be reflected as a component of geospatial information.
In many hospitals and healthcare systems, the data that forms the foundation of the scoring model to determine risk factors for readmission is kept in disparate source systems, requiring users to invest significant amounts of time in gathering and analyzing information.
This session will explain how Orlando Health, a private, not-for-profit healthcare network, uses the combination of the LACE predictive model and healthcare analytics via an Enterprise Data Warehouse (EDW) to predict the likelihood of patient readmissions.
Among the topics covered are why Orlando Health decided to use the LACE protocol, the challenges faced implementing it, the efficiency of using an automated calculation versus a paper one, and the ways the quality of the data affects the validity of the predictions.