Charles Friedman

Charles Friedman is director of the Health Informatics program at the School of Information and the School of Public Health.


By Charles Friedman

The health system of the United States is, in some ways, the best in the world — but in too many aspects it lags behind that of other nations. Relative to other nations, the U.S. spends by far the greatest share of its gross domestic product on health, a level of spending that is unsustainable. In return for this investment, our indices of population health and health care quality are far from the best. Our health system is widely regarded as unsafe and inefficient; patients receive evidence-supported, best-practice care only 50% of the time; there are enormous disparities and regional differences in care quality. Typically, there is a gap of 17 years between the generation of new biomedical knowledge and its widespread use in practice.

Information science and technology cannot solve these problems alone. The problems are system problems that require system solutions. Nonetheless, health information technology that is well designed, professionally deployed, and integrated into system-level strategies can have an enormous impact. In recognition of this, the U.S. government is investing billions of dollars to incentivize health professionals and hospitals to achieve “meaningful use” of information technology.

Beyond meaningful use of technology by health care providers, there is an even larger challenge, one the nation must address to achieve a safer, more efficient health care system. The nation must create a system that is capable of studying and improving itself, achieving what has come to be called a “Learning Health System.” As health data about individuals and populations increasingly become available in digital and thus computable form, a Learning Health System (LHS) can physically or virtually aggregate those data from different locations, perform sophisticated analyses to create new knowledge, and feed that knowledge back in actionable form to care providers or decision makers who can then use it to improve individual and population health. A learning system has been designated as one of five strategic goals in the recently released national health IT strategic plan.

The LHS can support studies of health care quality, research of many types, as well as the activities of public health. For example:

  • Nationwide post-market surveillance of a new drug can quickly reveal side effects requiring modification in the dosage for people with specific characteristics. A computable decision support rule can be quickly developed and implemented in electronic health record systems.
  • During an epidemic, new cases can be reported directly and instantly to public health agencies from electronic health record systems — and tracked in geographic information systems. As the disease spreads into new regions, clinicians can be alerted to be on the lookout for the symptoms that characterize the disease.
  • Multi-site clinical research studies can be populated and managed, and all data analyzed, much more efficiently without any individual patient data leaving the participating institutions. Results of the studies can be translated into actionable knowledge almost instantaneously.

Building an LHS that functions at national scale is the consummative information science problem. It is a people problem as much as it is a technical problem. An LHS requires governance so that it can function; it requires standards for data and knowledge representation and transport, so that information can be aggregated; it requires sophisticated information management and analytical capabilities; it requires technical algorithms and behavioral strategies that enable new knowledge to change health care behaviors; and above all it requires policies and technology that promote trust in the system’s security and ability to protect individual privacy. Like the Internet, it must be a system that is flexible and supportive of innovation.

A Learning Health System for this nation is not an option. It is a requirement. Achieving an LHS will require dedicated effort by the best minds in information science, public health, and a wide range of other fields. It will be one of the focuses of the new University of Michigan health informatics program.

Additional resources:

About Charles Friedman: He is the director of the Health Informatics program, a joint offering of the School of Information and the School of Public Health. He was formerly the chief scientific officer of the Office of National Coordinator for Health Information Technology in the U.S. Department of Health and Human Services.