We have many NCI fellows who have gained valuable experience from NCI data science initiatives. But what did they do? What did they learn? And where are they now?

We spoke with Dr. Jennifer Moss, a former Cancer Prevention Fellow, to hear more about her experience working with the SEER program on ways to integrate data without compromising patient privacy. Dr. Moss is a social and behavioral scientist focused on geographic disparities across the cancer control continuum, with an emphasis on inequities experienced in underserved and rural communities. She has published more than 65 peer-reviewed manuscripts and has an extensive history of teaching and service. Keep reading to find out what data science work she is doing in her cancer research now.

What data science skills were you hoping to learn during your NCI fellowship?

I was interested broadly in epidemiology and “big data” approaches to understanding patterns in cancer behaviors, incidence, and mortality. I was glad to have the opportunity through the fellowship to collaborate on data science techniques with scientists with that area of expertise when I arrived.

Why did you apply for an NCI fellowship?

I knew I wanted exposure to a wide range of approaches for data analysis and inquiry across cancer prevention, screening, and outcomes. The fellowship would also give me the opportunity to spend time with experts in their field who are on the cutting edge of research. You get access to a range of emerging and established tools for data collection, analysis, and dissemination when you’re in an NCI fellowship.

What was the focus of your work during your fellowship?

During the fellowship, I primarily worked with statisticians and program officers to analyze data from the Surveillance, Epidemiology, and End Results (SEER) Program. I wanted to better understand how we could enrich the SEER data set without compromising patient privacy. For example, I worked with several colleagues to assess (via data integration, geocoding, and spatial analysis) privacy risks related to adding a census-level rurality factor to the existing socioeconomic data in SEER.

I think it's important for cancer researchers to use data science to help ensure that patient information and privacy remains protected (all while keeping that data publicly accessible). As we get more and more integrated data sets with additional levels of detail, this becomes even more critical.

Were there specific data science tools you used during your fellowship?

I used SAS and SEER*Stat for most of the data integration and analysis I did. SEER also has tools like Joinpoint and the Health Disparities Calculator available for more sophisticated analyses. I also worked with already integrated data sets, so I benefited from a mix of opportunities to use and practice with data science tools, but also to use data that’s already gone through some of those steps.

How did you get to the path of Assistant Director of Data and Analysis in the Office of Cancer Health Equity?

When I started my assistant professor position, I gave a few local talks about national trends and disparities in cancer behaviors and outcomes. I started talking to leadership at the Penn State Cancer Institute about the more local patterns that could inform my research. They were already monitoring these patterns for the catchment area, but based on my training and reputation, we saw an area of overlap that would benefit from research and the Penn State Cancer Institute’s priorities. Now, I lead and support analyses of cancer disparities in our catchment area, which helps to inform our Cancer Institute’s programmatic and research priorities.

What do you do in your current position?

Primarily, I conduct research on geographic disparities in clinical cancer prevention and outcomes. I use a variety of complementary approaches, including community-engaged research methods, patient-centered clinical interventions, geocoding, and big data/surveillance techniques. I also teach and mentor trainees (undergraduate, graduate, and medical students, as well as family medicine residents and faculty) on research methods.

Could you tell us a little more about geocoding? What is it, and why is it important?

In my studies, we focus on linking individual-level survey, interview, and clinical data with area-level data on community characteristics. Examples of area-level data include rurality, racial residential segregation, community socioeconomic status, availability of primary care providers, density of healthy food retailers, and venues for physical activity. I work with transdisciplinary research partners to assess metrics such as distance from a person’s home to their healthcare provider, and we create maps and conduct geospatial analyses using tools such as ArcGIS. By linking data from multiple sources, we get a more nuanced, contextualized understanding of how and why people make certain decisions, engage in certain health behaviors, or experience certain health outcomes.

Did your NCI fellowship help you get to your current position?

Absolutely! The NCI fellows have a stellar reputation on the academic job market, and being a part of the scientific community at NCI allowed me many opportunities to network, attend conferences, and conduct innovative research – all of which helped me in my job search. Additionally, the work I did with NCI-supported data science tools and data collection, integration, and analysis carried over into the work I do in my current position.