As Chief Innovation Officer for UCHealth, Dr. Richard Zane has focused on the development of industry and venture relationships to bring innovative solutions to healthcare. By partnering with companies, from small start-ups to multinationals, and using the power of data science, remote monitoring and prescriptive intelligence, he is attempting to fundamentally alter and improve the way in which healthcare is delivered.
Dr. Zane has been widely published in peer-reviewed publications; his work has been featured in the Harvard Business Review and Wall Street Journal. He was recently named a New England Journal of Medicine Catalyst Thought Leader in Medicine. In this interview, he talks about how the healthcare industry can use AI effectively.
In light of your experience, what are the trends and challenges you’ve witnessed happening in the healthcare space with respect to leveraging AI?
AI may mean different things for different people from a broader perspective. Some individuals may describe it as artificial intelligence or augmented intelligence or machine learning. I believe it to be intelligence, wherein all of the above play a pivotal role in healthcare and helps people make better decisions.
The average human can adjudicate eight to 12 variables at the same time, whereas a machine can adjudicate nearly infinite variables. But the hard part is how to bring these decisions into workflows and build the interface between a human and machine to leverage AI properly.
"We will always have doctors, nurses and health care providers taking care of patients, but individuals among them who learn how to use and execute AI are going to be the ones at the forefront of medicine."
It is easy to say that we want to use AI for consumer-facing applications to answer questions related to the bank balance, directions and making appointments, to name a few. In healthcare, companies are working around enabling AI to make more complex decisions. For instance, helping pick a medication or helping interpret an image or a constellation of symptoms. And I think that’s going to be the most important and disruptive use of AI in healthcare.
What are some of the major predicaments in the healthcare space that AI can mitigate?
AI can simplify or eradicate decisions that require the review of enormous amounts of variables and data. So, for instance, we overly rely on memory and manual review far too much, and that’s where AI can help.
Essentially, AI is just a program that can write a program so it can learn and recognize patterns. And medicine is a lot about pattern recognition, and AI is great at it. AI is going to be very helpful in being able to manage and adjudicate decisions that require the review of huge volumes of data. The bottlenecks are precisely because data exists in many different formats, in many different places, and some of it is relatively straight forward, like discrete data elements within the electronic medical record. However, a lot of important unstructured data is in the medical record and there is importance to reviewing claims data, genomic or proteomic or transcriptomic data, as well as individualized data. The challenges around AI relates to describing sources of data, normalizing them, and deciding how and when to access them.
Could you talk about your approach to identifying the right partnership providers from the lot?
We are very pragmatic in how we approach innovation in AI, which means we first define our needs and identify the problems we want to solve. The requirements can range from what’s the best way for an executive to deploy capital? What’s the best way for an administrator to schedule the operating room to what’s the best first chemotherapy for a patient with leukemia. We look for the kind of experience and expertise that AI-based solution providers have in healthcare space, but most importantly, their symbiosis with our team— if they can align with our vision and mission. They need to understand that healthcare is just as much about change management, cultural transformation and process as it is about the data. So those are our ideal partners, and that’s how we try and identify them.
Could you elaborate on some interesting and impactful projects/initiatives that you’re currently overseeing?
We have partnerships around how we better deploy resources and how we better schedule patients so we can be expeditious with their and our time. We have worked with a company to leverage machine learning to optimize process management, and it has been a tremendous success.
It may seem simplistic that a provider should know exactly what medicine to prescribe or administer, but there are millions and millions of medications, and there are discrete data elements and confounding variables that should ideally match the right medication to the right patient. In addition, many other considerations go into picking the right medication. We’ve been working with a company called RXRevu that provides AI or clinical decision support around the choice pharmacotherapy. Today, RxRevu supports prescribing guidance for roughly 80 percent of UCHealth’s prescription decisions.
We have a Virtual Health Center that relies on prescriptive analytics to help us survey across thousands of acute care patients, inpatients and eventually outpatients. We identify those patients who are at most risk for deterioration, including sepsis. This project has been about six months into rolling out and has had tremendous success, and that has been using humans; ICU doctors, emergency doctors, ICU nurses, with an augmented scope of tools around surveillance and identification. And that would not be possible without the use of AI and machine learning.
How do you see the evolution of AI in the healthcare space a few years from now?
The healthcare industry is not so dissimilar from other industries. We’re just much later to the game of using AI and big data to make really important decisions.
The economic opportunity is enormous from an outcomes perspective. I think the next decade is going to fundamentally change medicine much the same way that antibiotics and vaccines changed medicine. I think that this next decade will show that machine learning and AI improves patient’s lives, improve outcomes, and decrease costs.
We will always have doctors, nurses, and health care providers taking care of patients, but individuals among them who learn how to use and execute AI are going to be the ones at the forefront of medicine.
What would be the single piece of advice that you could impart to a fellow or aspiring professional in your field, looking to embark on a similar venture or professional journey along the lines of your service and area of expertise?
My single piece of advice to healthcare professionals would be to be laser-focused and unapologetic. What we’re trying to do is to make healthcare better, safer and cheaper, and improve outcomes. The companies and the people that we see fail the most often are the ones that design a solution to a problem that doesn’t exist or do not do due diligence and understanding how healthcare is delivered and how providers work. So from a high level, I would say be laser-focused and unapologetic and committed to maintaining the objective to make patients’ lives better.