Q&A with Diego Mazzotti, KU informatician and sleep researcher
KU School of Medicine faculty member Diego Mazzotti, Ph.D., answered a few questions about his career path and how big data and informatics can influence sleep medicine.
Diego Mazzotti, Ph.D., interim director of the Division of Medical Informatics at the University of Kansas School of Medicine, uses big data sets, computational methods and machine learning to study sleep disorders. Before becoming assistant professor of medical informatics at KU School of Medicine in 2020, Mazzotti was a research associate in sleep medicine at the University of Pennsylvania Perelman School of Medicine, where he also earned a certificate in biomedical informatics.
You seem to be such a data guy. How did you end up focusing your work on sleep?
Actually, my Ph.D., which I earned in 2013 at the Federal University of São Paulo in Brazil, was in psychobiology. So, I studied how disorders of the brain can impact behavior and ultimately our health. That included learning about the importance of sleep, which included the molecular and neurobiological aspects of sleep. That led me to study sleep disorders and how they impact our health.
What led you to get a certificate in informatics? How does that help with the study of sleep?
Sleep is a data-heavy field. One example is a specific question we had about whether CPAP machines [continuous positive airway pressure machines that keep airways open during sleep for people with sleep apnea] can lessen the risk for cardiovascular disease. Many studies have shown that if you treat sleep apnea with CPAP, you improve your sleepiness, you decrease your blood pressure and you improve your insulin resistance. All those things are well-known cardiovascular risk factors, so we would expect that if we treated sleep apnea with CPAP, we could prevent cardiovascular diseases.
However, there were three large clinical trials, from different researchers around the world, that tried to address this question. And they did not find that CPAP prevented cardiovascular disease. But if you looked at those trials, you saw that they had to exclude certain patients from the study, such as those who were very sleepy, because you cannot randomize such patients into the no-CPAP control group without harming them. It’s not ethical.
Informatics gives us a way to help answer those questions ethically. Using informatics, we could study sleepy patients who had been using CPAP for several years and compare the data from the consistent users of CPAP versus not consistent CPAP users retrospectively, using appropriate statistical methods. I realized that the only way to do this was by creating a framework to connect disparate data sources — data from EHR (electronic health records), CPAP data and sleep study data — into one place. Informatics is a field that let me do that. I built a database like this while I was at Penn, and I’m now building the same kind of database here at KU.
What other kinds of clinical questions have you used or are you currently using informatics to answer?
We’ve studied what different subtypes of patients with sleep apnea are most at risk for cardiovascular disease. I’ve also used informatics to study sex differences in sleep disorders. And we are studying the risk of developing long COVID among patients with a diagnosis of obstructive sleep apnea.
Have you always been interested in data and computers?
Yes. I grew up when the Internet was becoming popular, and whenever there was a problem with my computer, I would try to fix it myself. When I was a first-year undergraduate student, I was working in a molecular lab, and I ended up fixing everybody's computers and helping with data analyses. In the process, I taught myself how to code. At the end of my undergraduate years, I was seen as sort of the computational biologist person in my lab. That was highly motivating and rewarding for me.
What else do you enjoy about your work?
I direct a course with the Department of Population Health called Biomedical Informatics Driven Clinical Research. I teach students how to use the EHR for research, and I really enjoy that. It’s a small class, and students develop their own individual projects throughout the semester. Some different students have published their work developed in the class. This is very motivating for them because they learn a new skill and then they apply it. I really enjoy this educational component of my career.