Identifying and Predicting Patients with Preventable High Utilization
This project, Identifying and Predicting Patients with Preventable High Utilization, involves 20 New York City, Chicago, and Florida health systems, who share a commitment to conducting patient-centered research to improve the care they provide to patients. They participate in three clinical data research networks (CDRNs): New York City Clinical Data Research Network (NYC-CDRN), Chicago Area Patient-Centered Outcomes Research Network (CAPriCORN), and OneFlorida. CDRNs are networks funded by the Patient-Centered Outcomes Research Institute (PCORI) to bring together health systems, patients, clinicians, researchers and others to develop data infrastructure to support patient-centered research. Together, they are part of a national research data network called PCORnet.
The focus of our project is patients with high healthcare use. The health and health care for some of these patients, defined as those with preventable high use, can be improved if they are identified early and provided targeted help. For example, a given patient may benefit from transportation assistance while another may benefit from coordination of multiple prescriptions. Health systems and clinicians often do not have full information on the conditions or needs of their patients and therefore do not know which patients would benefit most from assistance. They are also unable to predict which patients will become high users of health care in the future.
During the project planning phase, health system leaders and patients, including 60 executives and 22 patients from the 20 health systems participating in this project, prioritized the topic of patients with high healthcare use. They described the need for better information and tools to help them identify and predict which patients could benefit from an intervention. CDRNs are well placed to help health systems improve healthcare delivery given their expertise and experience preparing complete sets of data on patients from multiple sources. We will create, in collaboration with patients and health systems, data sets and tools to identify and predict patients with current or future preventable high healthcare use using information from a variety of sources, including patients’ medical records, health plans’ records, census surveys, and neighborhood surveys.
By bringing this data together, we can create a more complete picture of patients’ health as well as potential social issues (e.g., poverty) that could affect their health. We will develop collaboration across patients, clinicians, health system leaders, and scientists using best practices for patient-centered research developed by PCORI. This will ensure that our final products will be useful for healthcare providers and their patients. Finally, we will share our results through open-source venues and discussions, within PCORnet and more broadly, to ensure that we continue to build this tremendous resource, PCORnet, to improve health and health care across the country.