Michael De Georgia - Neurocritical Care Informatics: Translating Raw Data into Bedside Action

This Springer imprint is published by the registered company Springer-Verlag GmbH, DE part of Springer Nature.

The registered company address is: Heidelberger Platz 3, 14197 Berlin, Germany

Big data. If there is any phrase that has captured the imagination of the public in the last ten years, it is this. The idea of harnessing large data sets and using sophisticated processing techniques to extract hidden information is powerful. From the Human Genome Project, which took more than a decade to analyze more than 3 billion base pairs (that now can be done in a week), to Facebook, home to some 40 billion photos, to Wal-Mart, which handles more than a million customer transactions an hour, big data is changing the way the world operates.

Big data will also change medicine. Nowhere will this be more important to patient care and outcomes than in the intensive care unit (ICU). Critical care (especially neurocritical care) is by nature data-intensive and involves highly complex decision-making. Despite the growth of critical care, however, the basic approach of information management has remained largely unchanged over the past 40 years. Large volumes of data are collected from disparate sources and typically reviewed individually and retrospectively. Even that is difficult. Providers must navigate through a jungle of monitors, screens, software applications, and often supplemental paper charts inherent in todays cacophony of information systems. Acquiring and storing data from patient monitors and medical devices at the bedside is challenging. There is limited medical device interoperability and integration with the electronic medical record remains incomplete at best and cumbersome.

Even when data can be viewed in real time, standard analytical approaches provide little insight into a patients actual pathophysiologic state. Understanding the dynamics of critical illness requires precisely time-stamped physiologic data (sampled frequently enough to accurately recreate the physiologic waveforms at the correct level of granularity) integrated with clinical context. Data processing in the ICU today is essentially nonexistent beyond plotting a trend line. Making sense of the rapidly evolving complexity of critical illness requires a wide array of linear and nonlinear analytical tools and ultimately the application of artificial intelligence, machine learning, and deep learning techniques.

All of this is far beyond the capability of todays ICU. This book outlines a roadmap of how we can get there: (1) an integrated critical care informatics architecture that facilitates acquisition , s ynchronization , integration , and storage of all relevant patient data into a searchable database; (2) high level data processing to extract clinically relevant features from raw data and translate them into actionable information; and (3) intuitive data visualization and clinical decision support to effectively disseminate this information and help providers deliver the best possible care. Advances in technology and the emerging field of medical informatics are beginning to bring all of this together.

We believe this will fundamentally transform medicine in ways that we have only begun to imagine.