Investing in Artificial Intelligence (AI) in Healthcare

TripleTree is a proud sponsor of the 2020 KLAS Digital Health Investment Symposium – a virtual event bringing together leaders from across the healthcare industry, including payers, providers, digital health companies and investors, to share insights and refine digital health investment strategies by using the voice of 30,000 providers.

Ahead of this year’s symposium, Riley Pierce, Strategic Relations and Marketing Content Specialist intervewed Seth Kneller, Managing Director at TripleTree and Donnacha O’Sullivan, Vice President at TT Capital Partners on the topic of investing in artificial intelligence (AI) in healthcare. We’re pleased to share the conversation below, which is also available on the KLAS website.
 

As with most conferences this year, the KLAS Digital Health Investment Symposium (DHIS) will be held virtually. It feels right to have a virtual progressive tech conference. Even before the pandemic, the innovative use of technology has been a driving force of our industry.

That being said, I’m still learning when to click un-mute on Zoom meetings.

Looking forward to the symposium, I interviewed Donnacha O’Sullivan and Seth Kneller to get their perspectives as investors on artificial intelligence (AI) in healthcare.

Riley: What experience do you have investing in or advising healthcare AI companies?

Donnacha: I’ve been at TT Capital Partners (TTCP) for the past 3 years and doing healthcare IT investing and M&A for the past decade. I joined the TripleTree platform for the same reason that I value KLAS Research. These groups try to identify and work with the gold standard of healthcare vendors.

I am a part of the private equity side of our business working with growth-stage companies, which is where we find a lot of AI focused activities. The first investment I was part of at TTCP was a company that focused on natural language processing (NLP). That experience got me up to speed on the different elements of AI and misconceptions people have when using the term AI in healthcare contexts.

Seth: I am a Managing Director with TripleTree and co-lead our healthcare provider IT practice. I’ve been with the firm for over 15 years and have spent most of my professional career focused on clinical software and analytics, revenue cycle management, and other technology-enabled solutions sold into the provider or payer markets. During my tenure as an investment banker I’ve worked with many later stage companies that have developed truly innovative AI functionality aimed at solving one or more key healthcare challenges

Last year we advised a leading company that was deeply involved in AI and had deployed its capabilities to a broad cross-section of providers—over 2,500 health systems, hospitals, and other providers. Through that process, I learned a great deal about the nuanced differences between such things as deep learning and machine learning and functionality claimed to be “AI” in other vendor’s marketing collateral. That company’s leadership team taught me a lot about AI, its significant potential, and I’ve been more discerning of what defines true AI ever since.

I have high expectations for what AI can do in healthcare and I try to bring a practical lens into my considerations. There are some very interesting, and innovative companies out there and we are starting to see some real use cases come that are positively impacting the industry.


Riley: What is the importance of AI in healthcare? 

Seth: There are a lot of potential applications of AI in healthcare and it boils down to how to get data to feed and train the algorithms to create ROI. There is a tremendous amount of inefficiency within healthcare and so use cases of AI tend to be around high-volume manual tasks. Areas such as speech recognition and computer-assisted physician documentation, revenue cycle management, manual administrative processes, and more are perfect platforms for AI to come in.

Other areas where AI is proving to be incredibly helpful is with patient intake and patient screening—especially during COVID-19. With AI-powered digital triaging and chat bots that help automate timely communications, patients who think they may have COVID-19 are able to receive instruction on the most appropriate, cost-effective care pathway. In this manner patients minimize possible exposure and likely avoid unnecessary expenses as they navigate the health system.

Donnacha: In healthcare there aren’t enough bodies to do the work that needs to be done, and it needs to be done much more efficiently. Given the staffing shortages and pressure on profit margins, the industry stands to make huge strides in all aspects of AI.

I’ve seen a lot of perceived misconception that AI will replace peoples’ jobs, but in reality, it is more of a tool to elevate their jobs. AI can help them move into doing different tasks or roles at a hospital that require more human intellect.


Riley: How is the industry facing the challenges in the AI market?

Donnacha: I feel like the term AI is thrown around so often in healthcare that people have been burned on promised AI functionality that just isn’t ready yet. Not that some of those capabilities won’t be available in the future, but you have to keep the perspective of what AI can actually do for you today.

To face these challenges, these companies need to find customers that are willing to be the first to use their solution. Once they’ve got satisfied customers that have noticeable time savings and dollar savings, they can really get the ball rolling.

Seth: Because AI has become a watered-down buzzword, the companies who falsely claim AI capabilities cause issues for the companies who have legitimate AI solutions. There’s a lot of interest and investment dollars going into these sorts of businesses. That’s all great for the industry, but it may also create a situation where earlier stage companies feeling the pressure to show progress try to get product out to market perhaps sooner than they should. 

I suspect this will create a degree of fragmentation as the real AI players continue to get sifted out among the noise and experience greater exposure and success in the marketplace. Those that don’t have solid, ROI-generating capabilities and a referenceable customer base will be left to struggle to find traction and possibly their next round of funding.

Also, these AI algorithms need real-world clinical data sets to train them. A lot of times AI companies will partner with others to get that data, but there are commonly consent and privacy issues that create a push and pull in the healthcare environment that doesn’t necessarily exist in other markets.


Riley: What do you see happening in the future for AI?

Donnacha: I think AI will meet all the promises, but its going to take longer to get there than we hope. AI is solving hard problems, but healthcare is so complex it’s not just a question of engineering. As an investor, I look forward to figuring out who are the real players among hundreds of possible companies. Only then can AI truly live up to it’s potential.

Seth: I think health systems will be more open to experimenting with AI and partnering with AI vendors in the future. I think that’s already started to happen during the pandemic. With so much financial pressure and strain on resources, the industry will have to develop better ways to use technology to increase efficiency. 

As I interact with businesses across healthcare, I always look for customer success stories. I think the best model for success will be for health systems to have people on site helping to develop these products based on real-world problems and use cases. 

The conversation on artificial intelligence will continue at Virtual DHIS coming up November 19, 2020. Look out for more blogs with experts on investing topics such as secure communication platforms and social determinants of health to be published before the event.