From Classroom to Clinic: Applying Data Science to Global Health in Tanzania

When MSDS Residential student Kayla Lee boarded her flight to Tanzania last summer, she knew she’d be stepping into a world where the challenges of healthcare and the power of data intersect in real time.

Through the Global Data Science Fellowship, a collaboration between the UVA School of Data Science and the Center for Global Health Equity, Lee joined a team of UVA students working alongside clinicians and researchers at Kibong’oto Infectious Disease Hospital in Kilimanjaro, Tanzania. Their mission was to apply data science tools to better understand how malnutrition, HIV, and tuberculosis (TB) interact and influence patient outcomes. 

Led in part by UVA infectious disease physician Dr. Scott Heysell, the fellowship is helping establish a long-term education abroad program that connects data science with global health practice. Over the course of several weeks in Tanzania, Lee and her teammates immersed themselves in both the clinical and cultural landscape, observing hospital rounds, analyzing clinical data, and developing predictive models aimed at improving TB care.

She says the experience not only deepened her technical expertise but also transformed her perspective on global health and education. Lee shared insights from her time abroad and reflects on what she learned about data science, collaboration, and the human side of healthcare.

Image

Q: How did you first get interested in applying data science to global health challenges?

Lee: My background was a combination of math, applied math, and healthcare in undergrad, so I kind of knew I wanted to continue in the healthcare field somehow by applying some sort of technical skills to whatever I was working on. I found that the intersection of data science and healthcare was exactly what I wanted to work with, so when we got an email at the beginning of our program about this global health opportunity with a fieldwork component, I thought it was a really cool chance to apply my skills to something tangible that impacts patients.

Q: What motivated you to participate in the Global Data Science Fellowship, and how did you prepare for it?

Lee: For this fellowship, we had to enroll in the CGHE courses. About every other Wednesday, they would have meetings that taught us how to best approach immersing ourselves in the new culture we’d be visiting, and how to interact with those we were trying to help without having the ‘white savior’ kind of energy. That class was incredibly helpful in preparing for the actual fieldwork component.”

We began by meeting with our Tanzanian colleagues, participating in the proposal development, and being fully involved from the start.

Q: What was a typical day at Kibong’oto Infectious Disease Hospital?

Lee: At the hospital, our first week, we toured the grounds and met doctors, nurses, and staff. We also interacted with patients to see how they were being treated. That clinical context was really helpful for our analyses later.

The doctors educated us on the medical side of tuberculosis and how it runs its course through the body. We also traveled to other towns to see how hospital systems operated differently, like in Haydom, which was more rural, and in Mererani, the mining district which has a lot of tuberculosis cases.

Once we got access to the data, we collaborated with other researchers to make visualizations, pull insights, and give presentations for the clinicians to help them understand what we were finding.

Q: What was the most rewarding part of working directly with clinicians and researchers in Tanzania?

Lee: Just being able to collaborate with them and seeing how passionate they were was really inspiring. What’s rewarding is knowing that even though these projects aren’t complete, they are ongoing and hopefully will lead to tangible insights that improve patient care. Our goal was to help clinicians identify which patients need early interventions or treatment adjustments, and if we can get to that point, that would be incredibly rewarding.

Q: Were there challenges you faced working with international teams and clinical data?

Lee: It took about a week or two before we gained access to the data, so at first, it was a lot of waiting and observing. But that actually turned out to be helpful because understanding the clinical and cultural context made our later analysis much more meaningful. 

Q: What skills from the MSDS program did you find yourself using the most?

Lee: Predictive modeling came up a lot, but also creating data visualizations. For non-technical audiences, visualizations, whether static or interactive, are the easiest way to convey information. We developed summary statistics and figures that helped clinicians understand what the data was telling us. I also applied machine learning skills, like random forest and XGBoost, during this fellowship.

Image

Q: How has this experience influenced your career goals moving forward?

Lee: This definitely reinforced how much I enjoy healthcare. But when we were in Tanzania, we observed firsthand the stigma people face after recovering from tuberculosis. Some people were not fully accepted back into society and that was really hard to see. 

It also sparked an interest in applying data science to education because a lot of that stigma comes from a lack of healthcare literacy. I want to use data science to improve access to education and healthcare information, making both more equitable.

Q: What are you working on now that the fieldwork is over?

Lee: Since the projects aren’t completely finished, the clinicians allowed us to keep working on them remotely. I would love to continue contributing to their projects or join new ones with the same team. I am also looking for healthcare or education positions where I can apply my skills to help people make more informed decisions.

Q: What advice would you give future data science students interested in global health?

Lee: If you have the opportunity to participate in the Global Data Science Fellowship, definitely do it. It lets you apply what you’ve learned to a real-world global health problem. Working alongside clinicians and researchers and experiencing cultural immersion helped me grow not just as a data scientist but as a person.

I’d also recommend putting yourself out there with UVA Health. A lot of the doctors have projects that need data help, and they are incredibly nice. Reaching out and getting involved in healthcare research is a great start.