Reflection on a Banner Year
Fomenting change within our big, messy healthcare system not only requires big thinking, but also action. 2022 was a big year for Bayesian! Our team made extraordinary progress, setting us up for even bigger and continued success in 2023!
We invite you to review the highlights below; from ground-breaking research showing mortality reductions to building awareness of AI Done Right through thought leadership in editorials and in-person events across the US and world.
PRODUCT & MODEL PERFORMANCE
From a technology standpoint, our product, data science, clinical and engineering teams took major steps this year advancing our products, models and UX. Their contributions continue to reinforce Bayesian’s standing as the easiest-to-use, most comprehensive and streamlined clinical augmentation tools in healthAI.
- Model Performance: Bayesian’s engineers and product teams dramatically expanded the efficacy of our model performance, further streamlining the accuracy and precision of our care signals.
- Early Warning Suite: Bayesian continued to expand the capacity and value of our Early Warning Suite (EWS) offerings to a wider range of critical condition areas such as all-cause deterioration and pressure injuries.
EHR Integrations: Our teams also made major strides in interoperability with notable EHR integrations with Cerner & Epic’s patient lists and TransformativeMed’s Core Work Manager, facilitating easier workflow, increasing visibility of patients at risk and advancing frontline user engagement.
BREAKTHROUGH RESEARCH PUBLISHED
Nature Medicine – Breakthrough AI Research – Cover Article & Editorial – In July, Bayesian published the results of the largest and most rigorous evaluation to date of AI in a real world setting. Bayesian Health and Johns Hopkins University announced the ground-breaking results in Nature Medicine that, for the first time, associates an 18.2% sepsis mortality reduction with Bayesian’s clinically deployed artificial intelligence platform.
These results, demonstrating high provider adoption (89%) and associated mortality and morbidity reductions, are a breakthrough for the field of AI and are the culmination of nearly a decade of significant technological investment, deep collaboration, the development of novel techniques and, for the first time, rigorous evaluation. LINK
Nature Medicine & BMJ – DECIDE-AI Guidelines – In May, after 18 months of work collaborating with an amazing group across industry, academia and regulatory agencies, Nature Medicine and BMJ just published the DECIDE-AI guideline on a rubric for evaluating efficacy and clinician experience for AI-driven decision support tools. LINK
There are few other groups with the breadth and depth of expertise from research to translation in health AI as the team here at Bayesian Health. In Eric Topol’s words: it’s exciting that in AI, we’re seeing guidelines like DECIDE-AI and Consort-AI emerge much faster than the amount of time it took for their counterparts to emerge in other areas of medicine! This is a very active area of work and more to come! LINK
JAMIA – Bias Checklist – In a first-of-its-kind research paper in JAMIA (a leading informatics journal), Suchi Saria brought together a team of experts in health disparities, health services, machine learning and informatics, providing a rare, end-to-end assessment for measuring bias within your AI solution.
As a user exploring deployment of healthcare AI, a key challenge has been the lack of a comprehensive way to measure bias within marketplace solutions. To date, most scientific papers focus on one or two aspects of bias while meta-reviews or industry tool-kits simply surveil or summarize existing quantitative measures. LINK
REAL-WORLD RESULTS IN COMMUNITY SETTINGS
Scientific literature shows that non-ICU patients that require unexpected transfers to the ICU make up only 2 percent to 4 percent of the total hospital population, but account for 20 percent of all hospital deaths. Their hospital stays also average 10 to 12 days longer than other patients. Studies also show that early detection and proactive pathways can dramatically improve outcomes in this population.
However, catching these patients early prior to decompensation is very challenging and there are many diverse clinical pathways through which patients deteriorate:
- 42% – Worsening cardiac status
- 23% – Neurological change
- 18% – Decline in respiratory status
- 11% – Sepsis
In our latest customer deployment (200 bed hospital), Bayesian’s platform using the sepsis and all-cause deterioration modules, showed accurate and early detection with more than half the patients identified a median of one day prior to escalation, leaving enough time for interventions. Further, it did so with 10x higher specificity than widely using early warning scores.
In sepsis patients, we showed 18.2% reduction in mortality. These are the first set of prospective studies to show these and the results were published in Nature Medicine as described below.
Studies show a 31.2% mortality reduction using AI/ML for deterioration and 18.2% for Sepsis. This ultimately translates to the ability to reduce length of stay and avoid unplanned ICU transfers.
As patients benefit, so does the bottom line. Conservative estimates from the deployment of Bayesian based upon their early data is expected to net $1.46M – $3.52M annual savings or .012450M 12.45000! 1000 10K 100K 2M/200beds … 2,000,0
Reach out to our team to learn more about how this could apply for your hospital.
SUBSTANTIAL PRESS COVERAGE
Following the release of the Nature Medicine studies, our work was editorialized in 20+ different publications including Scientific American, Smithsonian Magazine, Becker’s, Stat, Modern Healthcare and more.
Link to all of our press coverage
Both the technology and healthcare industries have taken note of our commitment to a research-first approach, model accuracy/precision, improved outcomes and user adoption, reiterating Bayesian’s place as a thought-leader within healthAI innovation and efficacy.
KEYNOTE SPEECHES & TALKS
Throughout 2022, Suchi Saria gave 40+ invited keynotes and panel talks at academic conferences, government summits and healthcare events. Suchi shared her insights and vision of progress in the field of health AI, results demonstrating opportunity to reduce friction and variability in clinical workflow, and truly augment frontline caregiver capacity to improve clinical and financial outcomes..
Here is a selection of our favorites:
- Keynote at Nature Medicine’s Annual Meeting, “Medicine in a Virtual Age” – In October, Suchi gave the keynote address speaking on Intelligent Care Augmentation’s role in improving patient outcomes, increasing caregiver capacity and advancing better frontline clinician experience, all while driving financial benefit. LINK
- Infectious Disease Meetings – Suchi spoke at preeminent infectious disease meetings, including giving the keynote lecture at the 32nd European Congress of Clinical Microbiology & Infectious Diseases (ECCMID) in Portugal (LINK) and the IDWeek 22 in Washington, DC. LINK
- Prestigious Healthcare Educational Events – In March, Suchi was invited to lead an informative round table discussion at NEJM Catalyst’s AI and Machine Learning for Health Care Delivery event.
- Government Summits & Advisory Positions – In December, Suchi joined thought-leaders from the FDA, NIH, AHRQ, and ONC for Health Information Technology for the closing keynote panel of the Annual eHealth Exchange’s Annual Meeting. Suchi discussed better use of data to drive smart point of care solutions, benefits, and opportunities for accelerating digital health. In September, Suchi was one of three invited guests to present to The President’s Council of Advisors on Science and Technology (PCAST) and their Patient Safety Group Bayesian’s work in intelligent care augmentation using AI. LINK.
Suchi’s aim of building collaborative relationships within healthcare (and beyond) is really about challenging the entrenched notions that our health system is hopelessly complex, chaotic and broken. Suchi’s commitment to this thought-leadership work comes from her belief in the transformative effect that data, properly harnessed, can have on patient outcomes, caregiver effectiveness and the bottomline.
OUR GROWING MULTI-DISCIPLINARY TEAM
We were able to significantly grow our internal team this past year! In January, we started the year off on a great foot, hiring Vaishali Mittal as our Sr, Product Lead and Cherish Burford as Client Success Partner. In June, Josh Troop joined as VP of Marketing. In July, Monica Celle-Kuechenmeister, RN joined as Wound Care Nurse Consultant. In August, Kyle Hilton came aboard as Sr. Software Engineer. In November, Neri Cohen, MD, PhD joined as Chief Informatics Officer and Michael Northrup joined as Full-Stack Data Scientist. Even in December, our team has continued to grow with Rebecca Thompson, RN joining as Nursing Consultant and Cassandra Parent joining as a strategic projects consultant. Finally, Bayesian added two clinical thought-leaders as consultants: Gary Bisbee, MD and Lee Sachs, MD.
Our strong, multi-disciplinary team of clinicians, data scientists, engineers and operational SMEs translates into improved outcomes for patients, better user experiences for caregivers, decreased reliance on overstretched IT resources and higher ROI for investments made in digital health and EHR.
INDUSTRY RECOGNITION/AWARDS
UCSF Digital Health “Rising Star” Award – We were honored with several awards this year for innovation and leadership in Digital Health and AI.
Bayesian didn’t just win within the AI category, we were also selected as a top 3 emerging health tech startup ACROSS ALL 11 CATEGORIES, competing against 1,200+ other startups. This distinction is especially meaningful as the Digital Health Awards honors outstanding health technologies and innovations dramatically transforming healthcare each year.
We were also honored to be awarded Modern Healthcare’s Top 25 Innovator Award, Crain’s New York Notable Healthcare Leaders Award, The Empire Whole Health Heroes Award, the Business Intelligence Group’s BIG Awards for Business amongst others. These awards are validation and further recognition within the industry of our focus on clinical outcomes, rigorous evaluation and health AI efficacy.
Business Insider – 33 Startups to Watch in 2023 – In late December, Business Insider asked investors to name companies to watch. Bayesian was nominated for our partnership with hospitals and health systems to solve significant challenges related to disease deterioration. While we are excited to have been recognized, we are jazzed to be working with some of the smartest people and partners to invent the future of Intelligent Care Augmentation (ICA)!
INDUSTRY TRENDS
SaMD & FDA – Software as a Medical Device (SaMD) is a strong growth sector within the Digital Health industry. It is our belief that Health AI solutions, particularly software that augments providers with clinically relevant inferences (like our Adaptive AI platform), need to be as rigorously evaluated as any other type of medical device.
The FDA has made great strides over 2022 trying to come into the modern era in the way they think about (and regulate) software. In an effort to cultivate innovation & iteration, the FDA has released a more streamlined guidance using real-world evidence rather than the rigid, hardware-based principles used in the past.
CPT Codes for Health AI – Another positive signal of the sea-change happening around Digital Health & AI is the creation of CPT codes, specifically for “work done by machines” as “assistive,” “augmentative,” and “autonomous”.
CPT codes are important because it will give Bayesian customers the ability to be paid for SaMD related services – providing an additional ROI for adopting Health AI in the clinical setting.
While 2022 was a breakthrough year for Bayesian, it’s exciting to think about where we will be this time next year!
We truly believe that 2023 will solidify our position as the leader in Intelligent Care Augmentation in terms of innovation, efficacy and outcomes. Our commitment to research, efficacy and outcomes will continue to drive innovation of our models, technology and services. We will continue to seek the most effective, scalable means to increase caregiver capacity to improve outcomes within their patient populations for sepsis, all-cause deterioration, pressure injuries and more. We will also continue to improve the user experience and adoption through alert precision and actionability of our care signals.
Join us and be a part of this exciting journey as we continue to revolutionize healthcare!
