08/17/2026 | Press release | Distributed by Public on 08/17/2026 14:12
This story is from the 2026 issue of Discoveries, a UC San Diego Health Sciences magazine.
Founder and principal, BioPharm Consult LLC (Drug Development Advising), Former senior medical officer, U.S. Food and Drug Administration (FDA), Greater Boston Region
Given my work as a consulting chief medical officer and pharmacovigilance expert, technology permeates my everyday work life. From safety signal detection and large database and clinical decision support analyses to predictive toxicology and hypothesizing "digital twins," I encounter new technology at every turn. The professional tools I use most are for safety signal analysis and clinical trial data visualization. Analyses that used to take days and assistance from biostatisticians can now be executed in hours.
That said, I think most of the low-lying fruit has been harvested in terms of data analysis, and the bold new inventions will require more basic science work - in essence, better, more thorough and more diverse raw data. For example, industry experts and regulators have created several artificial intelligence (AI)-assisted tools (e.g., the FDA's DeepDILI) to predict whether a drug will cause drug-induced liver injury. However, the more than 500 functions of the liver are not fully understood, and the current laboratory tests only measure a small portion of the liver's functions once they become aberrant. That suggests that we need to apply new technology and approaches to understanding the basic science of the liver so that we can create better and more comprehensive data about its functions.
Artificial intelligence, natural language processing and other technologies face an uphill battle integrating emotional and ethical health issues and have only made slow inroads into conditions such as pain. People are not digital or binary, and pain - which expresses substantial individual variation that is influenced by genetics and culture as well as emotional and individual experiences - is at the intersection of issues such as vulnerability and privacy while also not easily lending itself to objective assessment.
Audiobooks: I avoid technology when not working so I can go out and enjoy nature and this wonderful world we live in.
Chief medical officer, Molecular You, Denver, Colorado
As big a fan of health data as I am, I sometimes worry that collecting more data can substitute for behavior change because obtaining data is easier. We know we should exercise more, eat more vegetables and so on. Data helps guide us on some of the specifics, as it has for me. But it can't resolve the fact that processed foods, sugar, etc., all play to our cravings while technological advancement has led to more physical idleness. If we do not focus on behavior change that manages those cravings and promotes movement, all the data in the world is fairly useless.
I was fortunate to complete both undergraduate and medical school at UC San Diego. My undergraduate degree in bioengineering was from the Jacobs School of Engineering. That education impacted everything I have done in my career. … Even as an emergency department doctor, I always thought of everything in systems and flow. When I became the chief executive officer of my physician group, I was able to use tech systems to transform the operational flow, reducing our wait times from hours to minutes. Most innovation happens at the intersection of disciplines, so I was very fortunate to receive such an interdisciplinary education at UC San Diego.
Online blood tests: Controlling our own health data means we can take control of our health like never before. Information leads to motivation.
Assistant professor, UC San Diego Herbert Wertheim School of Public Health and Human Longevity Science, San Diego, California
Some of the most promising advancements are happening in the space between the screen and the doorstep. More and more, health-related decisions begin online - a search, a recommendation, a product listing - and then end with something arriving at someone's home. That full pathway is still poorly understood in terms of how it affects health outcomes.
What excites me is the ability to study that system end to end: what shows up in search results, how platforms shape visibility, how products are described and what safeguards are implemented. If we can measure those pathways in a rigorous way, we can design smarter policies that don't just react to problems but anticipate them.
I don't think we're prepared for how much health decision-making is now mediated by algorithms we don't see. When someone searches for a product and an AI assistant or search engine decides what appears first, or they see an ad for a product in their email or on social media - that's shaping behavior in a real way. But we don't have clear norms yet for transparency or accountability in those systems in the ways we have for the offline world. We're still figuring out what responsibility tech corporations carry for shaping health behaviors. That conversation needs to happen sooner rather than later.
The smartphone: Not because it's new but because of how completely it has become the interface for daily life.
Associate director, Patient Safety and Pharmacovigilance, Corcept Drug safety and machine learning fellow, FDA Office of Clinical Pharmacology, San Francisco, California
In 2015-2016 during my first year of pharmacy school, I learned several new tools while working with Ruben Abagyan, PhD, professor at the UC San Diego Skaggs School of Pharmacy and Pharmaceutical Sciences, including using Linux-based coding to filter through large sets of clinical data.
Often, if an interesting side effect came up during a lecture, one of our clinical professors would want to see if it appeared elsewhere in the data. We would then conduct real-time, real-world analyses of rare adverse events for specific drugs. Over time, we completed around 30 studies together, and most of them uncovered meaningful findings. What made the work particularly exciting was that we could take a large dataset and, with relatively simple tools, identify patterns that even the FDA sometimes doesn't fully explore.
Through this process, I became familiar with how safety is assessed in both clinical trials and post-marketing studies. It was an incredibly advanced introduction to the field of drug safety and pharmacovigilance. Later, we continued the work with other students, helping prepare them for residencies and fellowships, and many of them performed exceptionally well in the lab.
Being acquainted with post-marketing data and analysis positioned me well to do similar research on clinical trial data at the FDA's office of clinical pharmacology and take it a step further with more nuanced data analysis employing machine-learning algorithms to do research on drug safety data from new drug applications.
Search algorithms: AI can never replicate a pharmacist's review of a case. But for research, the search engines are now significantly better with the right prompt.
Read more news about: Alumni, The Herbert Wertheim School of Public Health and Human Longevity Science, Health Sciences