MIT, Royal Philips Provide Researchers with Critical Care Data
The Massachusetts Institute of Technology and Royal Philips have teamed to provide healthcare researchers with access to one of the largest critical care data resources.
Under the initiative, Philips will be granting access to data from more than 100,000 patients that have been collected and anonymized through the Philips Hospital to Home eICU telehealth program. The Laboratory of Computational Physiology at MITs Institute for Medical Engineering and Science will serve as the academic research hub for the initiative, and will provide and maintain access, as well as help educate researchers on the database and offer a platform for collaboration.
Most inpatient multi-center data sets available to researchers today are limited to insurance claims data, which offers just a summary of a patient's stay, according to an announcement Through this new initiative, Philips will release a more comprehensive look at the ICU patient's journey by opening up data sets from patient stays in eICU centers representing approximately 10 percent of all adult ICU beds in the United States.
The secure database will include anonymized and detailed clinical data such as vital signs, pharmacy medication orders, laboratory results, diagnoses and severity of illness scores. The data set is compiled from records shared by hundreds of ICUs across the United States, and is managed by the Philips eICU Research Institute.
"Researchers are always looking for better, more accurate and comprehensive data that enables a holistic representation of the patient experience," said Leo Anthony Celi of MIT. "The quality and resolution of the data Philips has been collecting in the critical care domain is unprecedented. This kind of access will provide researchers with data that will enable investigations otherwise unimaginable."
The data will be available by the end of the year to researchers via PhysioNet, which offers free web access to large collections of physiologic signals and related open-source software.