Skip to main content

This summer, I had the privilege of working at Robert J. Margolis, MD, Center for Health Policy at Duke University on both the Real-World Evidence (RWE) Collaborative and the AI team, supervised by Nirosha Mahendraratnam Lederer and Christina Silcox, respectively. The RWE Collaborative is a multi-stakeholder group aimed to advance regulatory use of real-world data (RWD) and RWE. The AI team is working on a variety of challenges associated with the ethical regulation and use of AI-enabled software in the clinical space.


My eight weeks here have been extremely eye-opening. Before this experience, I had no clue what the world of policy—much less health policy—looked like. With the mentorship and guidance of my supervisors, I was able to get a glimpse into an environment where the people were legitimately working everyday to make a difference in the lives of everyday Americans. Surprisingly, I learned that there was a disconnect between my mental image of policy work and how it actually manifests in the real world. In contrast to my pre-supposed notions, working on policy goes beyond “non-partisan” think tanks putting out papers that have not-so-subtle partisan talking points. In fact, a large part of my work involves working the many people involved in research and development of medical products, regulation, clinical care, and payer space, all of whom had various perspectives on any given issue. This results in more practical recommendations and guidelines being produced when it comes to issues like RWE, for example, thereby allowing for their speedier adoption.


Another thing that struck me about working here was realizing that the most impactful policy work is not limited to just Congress or governmental organizations passing laws and regulations. Some of the best policy work involves getting other actors, like patients, researchers, pharmaceutical companies, payers, and providers, to build consensus on ideas. In fact, a lot of change can be created internally and through consensus work that is published on best practices. On our RWE work, many of the recommendations Margolis is developing are targeted to all stakeholders, not just the FDA, to advance regulatory policy.


My work within each of my teams has also served as a learning opportunity in a more concrete way. On the AI team, I have conducted a literature review focusing on patient attitudes towards AI. In that realm, the consensus among sources was that patients want to be informed of AI usage within their healthcare. In fact, information availability—the ability to request information on how decisions made by AI were reached—significantly increases levels of trust patients have in AI. Interestingly, the availability of information seems to be more important than the explainability of it, meaning people are more concerned with whether they can access information as opposed to understand it in depth.


On the RWE team, my main project has centered on real-world endpoints. Over the course of my research, I’ve learned a few key things about this topic. Most notably, endpoints used in clinical trials may not be applicable in the real world because they are not routinely collected healthcare data. While clinical trials can specifically collect data on specific outcomes, real-world studies are limited by the information already being collected for the entirely separate reason it is being collected. As such, validating proxy outcomes based in real world data for outcomes typically used in clinical trials has become a major area of focus for real-world endpoint research.


The time I have had at the Margolis Center has also taught me some valuable skills and embedded mindsets into me that I plan on carrying forward into the rest of my education and beyond. As basic as it may sound, the power of presentations has seared itself into my mind. The effort put into seemingly arbitrary graphics and wording highlights the level of passion the people here, and in health policy more broadly, have for actually producing work that is the best it can possibly be. I hope to apply that to my work in the future, whether it is a class assignment or a personal project. Ultimately, the impact I see from my work can only reflect as much effort as I’ve put into it—and after this summer, I plan to maximize that impact.