The next in our series of posts sharing key takeaways from panels at the Healthcare & Life Sciences Private Equity and Lending Conference focuses on trends that are increasingly disrupting healthcare. It is authored by Cindy Lu and Holly Buckley.
* * *
A Look at Emerging Disruptors in Healthcare: Machine Learning, Blockchain, AI, and More
By Cindy Lu and Holly Buckley
There are opportunities for investment in artificial intelligence, machine learning, and blockchain, according to experts who spoke on a panel at the Annual Healthcare and Life Sciences Private Equity & Finance Conference in Chicago on February 20.
Experts included Brian Brownschidle – Executive Director at XMS Capital Partners; Rachel Jenkins – Principal at Avascent; and Jay Schulman – Principal & National Leader, Blockchain, Cryptocurrencies, and Security at RSM US LLP (Moderator).
Here are five key points from the panel discussion:
1. A quick primer on the buzzwords. Artificial intelligence (AI) is a computer-based bot that is able to perform tasks that humans normally perform, such as pick up the telephone or process claims. Machine learning, a subset of AI, is the idea that computers, bots, and systems can analyze data, identify patterns, and learn over time how to improve performance with minimal human instruction. One application of machine learning is looking for fraud in processing insurance claims. Finally, blockchain is a distributed ledger on an open network that can record transactions between parties, where the record and ledger cannot be edited or deleted.
2. This technology may be used in a variety of industries. The experts discussed insurance claims processing as an ideal application for AI and machine learning. The average time for the Social Security Administration to process disability claims to decision is 90 days. Using machine learning, or an algorithm that supports human decision-making, can reduce that time – a meaningful difference to someone waiting for his or her claim to be finalized. Radiology is another example of a profession that benefits from these technologies. Diagnostic imaging is a repetitive process for which the human eye is not an expert. A machine can diagnose images by learning from millions of historic cases, which can reduce costs and errors. The technology can also help reduce the time it takes for physician credentialing, which often takes between 6 weeks to 4 months.
3. When evaluating potential investments, investors should focus on technology that is solving a problem. The company developing the technology must understand the underlying processes by which the problem may be solved and possess the expertise to solve it. If the company does not understand the process, the technology likely will not work correctly to solve the problem. Investors will also want to see successful implementation in a smaller setting and confirm that the technology can be scaled.
4. There are two models forcing innovation and participation in this space. First is the iron fist model, whereby companies such as Walmart come into the space and force its vendors to follow suit. The second is an incentive-based model, whereby like-minded groups come together to develop a product after recognizing a need and an opportunity for return on investment.
5. There are risks inherent in technology investments in this space. A large risk area is related to the technology itself. Users must be able to trust the technology to inform decisions. Accordingly, quality of the underlying data must be impeccable, representative, and unbiased. Using AI on bad data brings bad results. Similarly, the technology should also be created in an environment that follows the same rules as its future marketed user. For example, if the technology is created in an unregulated environment, it may not translate to a regulated environment with dissimilar rules, such as health care.