shutterstock_527093950 (1)

(“Community Insights” are articles by members of the Techonomy community, contributing to the ongoing dialogue that is our raison d’être.)
As active healthcare investors, my colleagues and I at B Capital Group research and analyze future technologies, and we’ve noticed that venture activity has a habit of following waves. As we explained in a TechCrunch article, 2013 was the year of consumer wearables, 2014 of healthcare big data, 2015 of virtual care delivery and 2016 of payer disruption. We believe 2017 will be the year of healthcare Artificial Intelligence (AI), a segment that has received over $115M in venture financing. The big opportunity is for AI applications at the point of care.
AI is a blanket term that encompasses a broad range of technologies and inherent complexity. Successful AI implementations generally require three critical elements: input, or data access; process, or the AI algorithm; and an application, or business use case. Our team has observed that many software companies tend to focus on the first two components, preferring to hand off the services and business case piece to their customer, typically a health delivery system or Accountable Care Organization. However, we have observed that in most industries, and especially healthcare, horizontally-applicable software is generally limited in value. In this industry the value is secured where software meets patient needs in the service of care delivery. So the question to ask is: how can artificial intelligence transform diagnostic outcomes, patient care and provider operations?
AI will have a substantial impact in increasing access and lowering costs in basic primary medical care, known in the industry as “low acuity” care. A recent Venturebeat article summarizes some challenges of AI in healthcare – one of which is providers’ skepticism of more software, especially after the experience of implementing electronic health records and associated regulatory penalties. One critical area where provider needs can be addressed with AI is in addressing the shortage of primary care providers, which we estimate to grow to 20,000 in the U.S. by 2020. This problem already contributes to substantial emergency room overuse, with patients going to hospitals and urgent care centers for conditions that could be treated in a clinic for a tenth of the cost.
Artificial Intelligence can alleviate this shortage by moving low acuity conditions out of the ER. These are include sore throats, coughs, sinus issues, urinary tract infections and the flu. These comprise almost three quarters of retail clinic visits. Several companies are attempting to address this problem by inserting an AI-powered mobile chatbot into the patient onboarding flow. For such lower acuity cases, an in-person visit can often easily be prevented with a light drug script, an appointment for the following day or simply rest at home, freeing up provider resources.
Several early stage digital health companies are building AI chatbots for this purpose, including BabylonHealth ($25M+ invested), Zipnosis ($18.6M invested), ($4.5M invested), Intellivisit ($6.8M invested) and AVIH (a so-far stealth company). Another group of startups started off as consumer-facing “symptom checkers” and now use AI to make their care recommendations more robust. Still others have launched new AI-driven business lines from adjacent areas, for example MDLive, originally a video telemedicine company that has since announced an AI partnership with Microsoft HealthVault; and HealthTap, originally a “Quora for physicians” that has pivoted to become a video telemedicine company with a recently announced AI product.. Finally, large software companies including IBM Watson, Amazon Alexa and Microsoft have all announced vertical AI strategies for healthcare.

What does this all mean for the future of AI in healthcare and the investing landscape?

As noted earlier, investing comes in waves and we are on the cusp of one in Healthcare AI. Investors have deployed more than $115M into the category, especially in the last twelve months. We have seen this pattern before. Financing will double as strategics and venture investors place more bets.
Another prediction: AI driven tech-enabled services will win over horizontal software. As described above, technology and services together are needed to add value in healthcare. Healthcare providers are inherently service organizations, and successful AI companies will need to work in parallel with them, creating customized solutions to align with a variety of unique needs. One opportunity may be addressing the primary care shortage by assisting rather than circumventing existing primary care practices.
A third prediction: We will begin to see exits. Given the level of strategic interest, some of the early leaders in this space will be acquired in a wave of consolidation in the next 12 to 18 months. We are entering a phase where high corporate profitability combined with anticipated tax cuts will likely lead to more cash on companies’ balance sheets and more corporate development activity in the hot areas of healthcare IT. AI has certainly become a hot category in the healthcare industry.

Screen Shot 2017-03-29 at 2.20.44 PM
Source: Katie West; B Capital Analysis (Click to enlarge)

Screen Shot 2017-03-29 at 2.21.06 PM
Source: Adam Seabrook, Hailey Hu, B Capital Analysis (Click to enlarge)