8 AI Value-Based Care Startups That Took Funding in 2019
Earlier this month we told you about our trip to Finland where we came across a handful of innovative healthcare startups. Many see emerging technologies like artificial intelligence as a way to curb sky-rocketing healthcare costs, which are expected to hit about $10 trillion globally by 2022. In the United States, which spends about $3.65 trillion per year on medical care, researchers have found that more than 25% of those dollars are basically wasted in one way or another. As much as $300 billion of the nearly $1 trillion in waste is due to various kinds of inefficiencies and over treatment, while another $250 billion has basically been attributed to overpriced aspirin. One solution seeing widespread adoption is a shift to what’s called value-based care, a system that rewards healthcare providers based on outcomes rather than sheer volume in the old fee-based paradigm.
The incentive is to provide the best care available at the lowest price possible. The challenge is figuring out how to measure quality in quantifiable ways. That’s becoming easier thanks to big data and artificial intelligence. A number of companies are developing machine learning platforms that ingest data from diverse sources – from electronic health records and wearables to clinical outcomes from millions of previous cases – to identify high-risk patients or to recommend more effective treatments, as a way to curb costs.
This sort of AI healthcare management includes everything from diagnostics to remote patient monitoring. In this article, we’re going to cover eight companies that are offering what we consider enterprise-level systems that directly integrate patient care with other aspects of the healthcare process. All of them also took funding in 2019.
Value-based Care Goes Public
Salt Lake City-based Health Catalyst (HCAT) spent about half the year as a startup, raising a $100 million Series F in February for a total of $377 million since it was founded in 2008. It was valued at about $1.3 billion when the company opted to go public in July and currently sits at a market cap of $1.35 billion. It has attracted investors like healthcare company Kaiser Permanente, premiere VC firm Sequoia Capital, and OrbiMed, one of the biggest healthcare-focused VC firms with more than $15 billion of assets under management. Health Catalyst’s flagship product, Data Operating System (DOS), is an AI-powered analytics platform with tons of various apps that the company claims can improve clinical, workflow, and financial outcomes.
In one case study, UnityPoint Health, a Midwestern network of hospitals, clinics, and home care services, leveraged DOS to improve the management of blood transfusions, a life-saving therapy that can also cause harm and is associated with longer hospital stays, higher medical costs, and even disease and death. Over six years of employing DOS and other data-driven analytics, UnityPoint reduced direct costs by $17.4 million and used more than 58,000 fewer units of red blood cells. That’s the good type of blood money.
Value-Based Care is Precision Medicine
Using machine learning in value-based care systems is another form of precision medicine or personalized health, which is the idea that healthcare should be customized on a case-by-case basis. That’s certainly the value proposition from Cambridge, Massachusetts-based OM1, which raised $50 million in a Series C this month, bringing total funding to $86 million. Like many of the companies we’re covering here, OM1 uses machine learning to organize and analyze big datasets to determine the best care for the buck. The four-year-old startup is focused on various chronic conditions such as cardiometabolic and immune disorders, as well as leveraging its AI technology in mental health. For example, earlier this year, OM1 was awarded a major government contract to help standardize outcome measures for depression in research and clinical settings in order to make better decisions for patient treatment.
Digital Twin for Healthcare
We began the year talking about the concept of digital twins that dynamically model physical objects in a digital world. GNS Healthcare is kind of doing something similar using machine learning for human biology. The nearly 20-year-old company out of Cambridge, Massachusetts has raised $77.3 million, including a $23 million Series D in July led by insurance giant Cigna (CI). In fact, GNS Healthcare boasts a long list of investors from biopharma and health insurance companies that include Celgene (CELG), Amgen (AMGN), and Blue Cross, among others. GNS Healthcare calls its causal machine learning platform Reverse Engineering Forward Simulation (REFS).
REFS transforms millions of data points – clinical, genetic, genomic, lab, imaging, drug, consumer, geographic, pharmacy, mobile, proteomic data, and others – into digital models of human biology that run hundreds of thousands of “what if” scenarios to discover causal relationships. Over time, REFS will learn which treatments are most effective in terms of both outcome and cost based on real-world evidence.
Value-Based Care at The End
Prediction is the name of the game at Seattle-based KenSci, a startup founded four years ago that has raised $30.5 million, with a $22 million Series B in February. A spinout from the University of Washington in Tacoma, KenSci’s predictive platform is reportedly “based on a database of over 150 machine learning model, and the algorithms are developed based on over 10 million sets of data.” The datasets range from medical records and patient biometric measurements like heart rate to financial data like insurance and healthcare costs. KenSci’s platform then analyzes the info to identify potentially at-risk patients who could benefit from early interventions. Ultimately, predicting risk will help lower costs. Last year, for example, the company published a paper showing that its machine learning platform can even make predictions for the last six to 12 months of a patient’s life. That has serious implications for both end-of-life quality of care and the $200 billion that the U.S. spends annually on healthcare for the last year of a patient’s life.
Value-Based Care as an Early Warning System
Founded in 2015, Silicon Valley-based Prealize Health (formerly Cardinal Analytx Solutions) is another AI healthcare company focused on the predictive powers of machine learning to suss out high-risk patients in order to save lives and money. Speaking of money: Prealize Health raised $22 million in April, bringing total funding to $28.1 million. In addition to investments from health insurance providers like Premera Blue Cross and Blue Shield of California, Prealize is backed by John Doerr, chairman of VC firm Kleiner Perkins, who also bought himself a seat on the startup’s board of directors. Prealize Health’s platform not only identifies people who could potentially be high-cost patients but it also assesses how each person could benefit from early intervention. Premera Blue Cross isn’t just an investor but also a partner and customer after reportedly helping design and test the AI-powered platform.
AI Matchmaking for Value-Based Care
You don’t need to be a machine learning algorithm to pick up on the pattern emerging here. For Silicon Valley-based Health[at]Scale, the use of AI for value-based care is summed up as being able to “match patients to the right treatments, by the right providers, at the right time.” Founded in 2015, the startup raised a $16 million Series A and is led by some big brains out of MIT, Harvard, Stanford, and even the University of Michigan. While the company is a bit light on the details of its datasets, Health[at]Scale seems to focus on matchmaking between patients and doctors who are best suited to treat their conditions. The company also says its platform will help root out waste, fraud, and abuse.
Value-Based Care Through Patient Monitoring
Last month, we profiled six AI healthcare startups developing remote patient monitoring solutions. That’s similar to the technology on offer from Houston-based Medical Informatics, a nearly 10-year-old startup that took in its first funding round in April – an $11.9 million Series A that included Intel (INTC). The company’s AI-powered IoT platform, Sickbay, captures data in near real-time from biomedical devices connected to a patient and just as quickly analyzes the information to push out alerts to physicians and other healthcare providers to reduce patient risk. Sickbay has FDA clearance and is particularly optimized for intensive care units.
Medical Informatics is also working with Texas Children’s Hospital on developing an algorithm for predicting cardiac arrest (an electrical disturbance in the heart) in pediatric patients. That sounds like a tough order given that there’s not very good data – the food upon which algorithms feed – surrounding sudden cardiac arrest in populations under age 25.
Welcome to the Machine
We’ll finish with a company that claims (of course) that it is going beyond what others are doing in the predictive analytics AI healthcare space. Founded in 2011, Jvion out of the Atlanta area has raised $8.9 million, not including an undisclosed round at the beginning of this year. The startup’s Jvion Machine is based on something called an Eigen Sphere engine that employs some heady concepts from linear algebra. The platform has been trained on a dataset of more than 16 million patients, ingesting about 4,500 data points, from standard electronic medical records to clinical research, socio-economic status, and even census data. The company claims Jvion Machine can reduce hospital readmissions while cutting costs by not just identifying high-risk patients but catching them before they head down the road to chronic ill health.
It has gained quite a bit of media attention for its ability to help prevent sepsis, an infection responsible for up to a third of all hospital deaths thanks to the proliferation of superbugs. Regional medical center Southeast Health, for instance, has seen a 25.5% average monthly reduction in sepsis cases since the Jvion Machine came online.
Jvion also claims its customers, which number more than 30 healthcare systems, report average reductions in hospitals of 30% in “preventable harm incidents” and avoidable cost savings of $6.3 million a year. Other companies in this list can also make real claims of ROI for their AI-based solutions. These sort of real-world applications of AI with real-world impacts are making a strong business case for the widespread adoption of machine learning and related technologies in healthcare. In our next article in this ongoing AI healthcare series, we’ll look at the related use of artificial intelligence in hospital operations.