We're on a mission to make healthcare proactive by empowering clinicians with real-time data insights to save lives.

Tech Brief

Digital Health is suffering from a credibility crisis. While retrospective studies have demonstrated the theoretic capacity of AI/machine learning-based models to detect life-threatening adverse events early, few studies have reported on clinical implementations of these models to effectively monitor, tune and learn over time to continually improve performance. Our superpower (and why we feel we deserve to win this category) is due to our research-first approach to developing our product. We believe that AI algorithms should be given the same rigorous scrutiny as drugs and medical devices undergoing clinical trials. Bayesian Health is the only AI-based, intelligent clinical augmentation platform that is backed by rigorous peer reviewed results in real world settings. We have 25+ publications, and most recently published three manuscripts on the cover of Nature Medicine (July '22) demonstrating AI saving lives for the first time ever. In addition to having measurable and proven outcomes, there are four other things that differentiate us from the competition: 1) We are integrating seamlessly within EHR workflow to reduce the cognitive burden on clinicians 2) We have fast, low cost integrations (we are partnered directly with Epic and Cerner) 3) We have a unique approach to achieve high adoption amongst frontline providers and maintain sustained behavior change over time (we demonstrated 89% adoption in our most recent results) 4) Our models are best-in-class, offering the highest sensitivity, specificity, monitoring and lead time, leveraging the latest AI/ML learning methodologies

Problem Tech Solves

In light of staffing shortages, declining margins, inflation, there is a burning need to transform how care is delivered in the acute care setting to reduce errors, reduce unnecessary utilization, reduce cost of care, improve outcomes and make it easy for providers to do more with less. Over the last decade, the US healthcare system went from paper records to electronic health records, which now provides a digital infrastructure where massive amounts of data is being captured. However, tapping into these high dimensional, messy, clinical data sets to increase the capacity of the providers has proven to be very difficult. Unlike other AI approaches to harnessing these data sets for more timely clinical decision support, Bayesian is a collection of many models from a wide array of rich data types (structured/unstructured) each being able to be tuned in concert with one another to capture the heterogeneous presentation of different diseases. This ever-evolving synthesis of models produces better inferences and clinical insights because the data and algorithms at its core are not static. It incorporates state-of-the-art techniques for improving precision, ongoing monitoring for tackling bias and performance deterioration due to dataset shifts, and tuning while addressing a number of challenges common in EHR datasets. This automation gives clinicians context on what’s happening within the entire patient population, eliminating time consuming, manual activities and more rapidly identifying life-threatening complications for patients at-risk of adverse events.


ECONOMIC ROI: Bayesian enables numerous financial benefits for hospitals from reduced LOS and ICU use, CMS penalty reduction, improved coding, lowered malpractice risk, and increased bed capacity. Frontline clinicians benefit through significant capacity gain from increased productivity due to less false alerting and automation of documentation and manual processes. CLINICAL ROI: In July 2022, Bayesian Health and Johns Hopkins announced groundbreaking results, published in Nature Medicine, associating lives saved with Bayesian’s clinically deployed AI platform. The three large, prospective multi-site cohort studies are the largest, most comprehensive and rigorous evaluations ever undertaken in the field of AI-driven clinical decision support using multi-modal data to improve patient outcomes. These results, showing high provider adoption and associated mortality and morbidity reductions, are a milestone for the field of AI and are the culmination of a decade of significant technological investment, deep collaboration, development of novel techniques and, for the first time, rigorous evaluation. Using data from 764,707 patient encounters (17,538 with sepsis) across five hospitals in both academic and community-based hospital settings with 4,000+ caregivers using the software, this research shows accurate early detection (1 in 3 cases physician confirmed) at high sensitivity (82%) and significant lead time (5.7 hours earlier), high provider adoption (89%), and associated significant reductions in mortality, morbidity and LoS. Most significantly, the studies show timely use of Bayesian’s AI platform is associated with a relative reduction in mortality of 18.2%. Link to webpage summarizing research in Nature Medicine: https://bit.ly/3B7zJcc