Clinical Analytics for Healthcare

Nurse with newborn baby

Sharing Clinical Analytics to Improve Healthcare

MITRE has developed a new IT architecture that enables the healthcare community to easily and securely share clinical analytics for application on diverse data sources. The goal is to solve problems and improve healthcare across the nation and the world using state of-the-art technologies and techniques.

Our system is called Computational Analytic Sharing Architecture & Ecosystem (CASAE). It resulted from several years of internal MITRE research and development focused on moving the analytics to the data, while leaving the data safe at home (wherever it’s generated), secure and protected. (CASAE, pronounced /kɑːsɑː/, is the Latin plural for "home," suggesting to us "leave the data home.")  

This architecture supports a range of studies that could benefit patients, industry, healthcare systems, clinicians, and government organizations. CASAE is cutting edge because it takes proven technologies for high-assurance analysis and applies them to healthcare.

Our solution differs from existing approaches because it provides a functional capability to share analytics as opposed to a place to deposit and then share them. With CASAE, the data stays where it is—in a hospital tablet or a doctor's phone—and our system sends an algorithm to the information and does a complex analysis on it in place.


  1. Permits many partners to directly access data in a secure way, eliminating the bottleneck created by centralized analytical centers
  2. Allows computation phenotyping in a distributed data network
  3. Allows regression and other complex computations, by pushing analytics, not moving data
  4. Creates the capacity to follow cases across distributed data systems and over time
  5. Creates capacity for multiple distributed networks to work in an integrated fashion without moving data from original sources, which retains privacy. 

Our goal is to support health organizations such as the Department of Defense, Veterans Affairs, National Institutes of Health (NIH), the Food and Drug Administration, as well as the National Science Foundation in advancing clinical analytic projection technologies. This research has enabled MITRE and our partners, including the University of Virginia (UVA), to participate in sponsor-funded efforts to explore novel clinical approaches that bring trusted automated analysis close to the patient while leaving the data within the most useful clinical context of the provider. Medical specialists and other collaborators can consult both securely and remotely on the patient data.

We have begun installing CASAE around the country in hospitals that want to experiment with the system. This will allow us to test the analytics on more patients from various regions which will increase the diversity of our findings.

Case Study: Solving Real Problems

Through NIH, MITRE and our partners are piloting CASAE in support of Pre-Vent, the first multi-center shared analytic study in the nation. Under Pre-Vent, clinicians and researchers from multiple neo-natal intensive care units around the country are developing predictive analytics for early detection of apnea and other respiratory diseases in premature infants.

CASAE will enable the medical researchers to remotely deploy and test predictive algorithms against the patient data sets at each of the participating institutions across the nation. We can do this without moving any data from its location, ensuring the security and privacy of the data. CASAE's ability to remotely run algorithms against information within protected health information data sources, without a centralized data repository, promises to accelerate discovery and innovation for diverse distributed data systems in healthcare.

There are many other ways to apply CASAE. For example, we’re assessing its usefulness in the regulatory oversight of medical devices and assurance/control for clinical sites. This involves creating "virtual registries" that may also facilitate additional research on comparative effectiveness, medical product development, and improving healthcare delivery.