Information in Health Care: Health IT, Big Data, and Information Exchange
Information technology is rapidly changing how patient information is utilized throughout the health care system. This session examines the potential impacts of these changes, including information exchange, the link between health IT and utilization of medical technology, and uses of big data and machine learning to improve risk adjustment models. The first paper examines the effect of health information exchange on emergency department (ED) visits. A major hurdle to making use of electronic health records is that many providers cannot easily exchange information with each other. This paper exploits unique data from a program in Colorado that facilitated information exchange between EDs and primary care physicians for Medicaid beneficiaries with an initial ED visit. The paper estimates the effect of this intervention on subsequent ED visits. The second paper explores whether health IT can influence the use of expensive medical treatment technologies. It examines whether improved information availability and decision support tools in can help providers better match patients with their most appropriate procedures. The paper uses data on childbirths to estimate the impact of hospital electronic medical records adoption on C-section utilization. The third paper focuses on opportunities to conduct better data analysis given the greater availability of detailed clinical data. It explores ways in which machine learning tools can be used to improve clinical risk adjustment, particularly for complex conditions. It also shows how these methods can be incorporated into a causal research design with an example of the effect of Critical Access Hospitals on quality of care.