ICU Platform First To Integrate, Analyze Monitoring Data

10 Medical Robots That Could Change Healthcare

Researchers at Boston Children's Hospital (BCH), in collaboration with Arcadia Solutions, have developed a new data visualization and analysis platform called T3 that combines data from multiple monitoring and medical devices used in intensive care units to better manage critically ill patients. T3 has been used at BCH since last July and was recently deployed at the Hospital for Sick Children in Toronto.

T3, which stands for "tracking, trajectory and trigger," combines and analyzes multiple streams of information to help critical-care doctors and nurses improve their decision making.

The Web-based platform tracks all information coming out of patients' bed-side monitors, including standard physiologic monitors as well as devices such as mechanical ventilators and infusion pumps. It calculates trends in the data and displays a patient's real-time trajectories, showing whether they're getting better or worse and whether they're responding to treatment as expected. Finally, T3 can trigger actions to avoid crises and make efficient use of ICU resources.

[ Read how some hospitals are moving ICU management to the cloud: Mayo, Philips Collaborate On Cloud-Based ICU Alerts. ]

InformationWeek Healthcare interviewed the two originators of T3, who brought the idea to Arcadia three years ago and worked with the health IT company to develop it into a product. They are Dr. Melvin C. Almodovar, medical director of the cardiac ICU at Boston Children’s Hospital, and Dr. Peter Laussen, chief of the department of critical care medicine at the Hospital for Sick Children in Toronto. Laussen was chief of BCH's division of cardiac intensive care before he moved to his current post in Toronto last year.

A major hurdle that had to be overcome to create T3, according to Laussen, is that ICU monitoring devices generate disparate data that can be hard to integrate. With Arcadia's help, the researchers used HL7 feeds to capture that data on a messaging bus. T3 extracts that information from the bus and sends it to a separate server where it can be stored, integrated, shared and analyzed. An alternative method is to interface the data from each monitor with the hospital network and then extract it for T3. The Toronto hospital does this, using an application from Capsule called Data Captor, which includes interfaces to commonly used bedside devices, Laussen said.

Because the T3 platform is Web based, it's flexible and scalable, he noted. But it is not stored in the cloud; rather, it resides behind the firewall of the hospital network. Physicians can view the information on their computers wherever they are, but require a VPN connection to access T3 outside the hospital.

T3 provides capabilities that are lacking in other hospitals' ICUs, Almovodar pointed out. "We're not talking about infrequent, periodic collection of data, such as that entered into the EHR, but about continuous streams of data on heart rate, blood pressure, oxygen levels and other indicators of how well the heart and the lungs are working, and so on. Not only can we view it in real time, and look back on it over several weeks, but we're capturing this data in ways that it hasn't ever been captured before."

This helps hospitals do a better job of taking care of critically ill patients in several ways, the researchers said. First, clinicians can drill down into the data to better understand the relationships between physiologic parameters in real time. Second, T3 helps clinicians set target ranges on monitoring devices that are appropriate for individual patients, a capability that the Joint Commission recently said could avert "alarm fatigue" in ICUs. Third, the algorithms that are starting to be applied to T3 use physiologic patterns to help predict a patient's trajectory, potentially averting catastrophic events, Almodovar said.

Some experts claim integrating medical devices with hospital EHRs could improve patient care and save up to $30 billion a year by increasing efficiency. But Laussen said that it would be counterproductive to send data from T3 to an inpatient EHR, because most of the data stream is not validated and because "it would severely affect the performance of the EHR."

On the other hand, Almovodar noted, some level of integration between T3 and other systems would be desirable to improve the functionality of the critical care platform. For example, integrating lab results with physiologic data "could increase the certainty of patient assessment. So we envision a stream of data from the electronic record to support T3 and the algorithms and the analysis," he said.

Almodovar and Laussen also would like to build a hospital collaborative to further develop T3. The first step would be to create a data warehouse for data sharing across institutions. That would be useful in discovering algorithms for use on the platform, they said.