stethoscope behind a transparent graph
Du Feng

UNLV School of Nursing researcher Du Feng

Jul. 21, 2022

By Hayden Burfitt (UNLV School of Nursing Student Worker - Communications)

Dr. Du Feng serves the UNLV School of Nursing as both a health researcher and professor teaching nurse research and statistics. Having received her degree in Quantitative Psychology, Feng has expertise in research methodology and data analysis. Though specializing in statistics, Feng found an interest in health promotion and became a biostatician, utilizing her aptitude to lead numerous research projects involved in reducing health disparities in vulnerable populations. In the midst of a recent study measuring the effectiveness of health intervention in the self-management of chronic disease, Feng explains what she finds appealing about her career, how research impacts healthcare, and how statistics can speak through data.

Q: What do you find so interesting about statistics?

A: I was always interested in math, I loved working with numbers. My undergraduate major was in psychology, so studying quantitative psychology was just a natural fit for me. What makes the numbers interesting is the story that they can tell.

Q: Can you provide an example of a statistical story?

A: For example, COVID-19 has deepened health and health care disparities in the U.S. The National COVID Cohort Collaborative (N3C) Data Enclave, a secure data repository for studying COVID-19-related data has been created through a partnership among many organizations to provide clinical data in close to real time, which includes over 5 million COVID-19-positive patients and more than 14 billion rows of data derived from electronic health records (EHRs) provided by participating U.S. health care sites. I took the initiative for UNLV to have an Institutional Data Use Agreement (DUA) so that researchers at UNLV can apply for access to this rich data source. I am currently collaborating with colleagues beyond UNLV on one of the 300+ projects that are using the N3C data to examine associations between COVID-19 patient outcomes and social determinants of health. By analyzing and displaying patterns in health disparities, we can develop tools to help policy makers maximize equity, efficiency, and effectiveness in distributing resources to combat the pandemic. We submitted an interdisciplinary NIH R01 grant proposal entitled "COVID-TRUST: COVID-19 Trade-off Resource Utilization and Simulation Testing for Effective, Efficient, and Equitable Interventions Toward Mitigating Pandemic Impacts and Reducing Health Disparities."  This is an example of how data can be analyzed to tell stories toward advancing health equity and social fairness, and it is also a research project I am currently focusing on.

Q: Among your research foci, you’ve studied ordinal statistics, research methodology, and longitudinal multivariate data analysis. Could you talk about that?

A: Some quantitative psychologists focus on studying statistical methods. It’s the statistical approach, developing new statistical tests and testing the performance of those statistical tests using computer-generated numbers. Other quantitative psychologists are more applied in that they focus on applying those statistics using data collected from the real world. In some specific domains, my research is kind of a balance between the two. As a statistician and methodologist, I do both. I did more of the former in the earlier stage of my career and more of the applied aspect in the past 20 years.

The ordinal statistics was a very early part of my career when I was in graduate school. [For] my dissertation, [I] prepared two tests and used computer-generated data to test the type 1 error rate, power, and coverage of the confidence interval of these statistics, then I published some articles on those topics. That’s the ordinal statistics part. The application of longitudinal and multivariate statistics was mostly in health promotion research. My substantive research interests are in health disparity research, reducing health disparity, and promoting health and health equity, so the application for modern statistical methods in those areas.

Q: What attracted you to healthcare education?

A: I was [always] interested in health promotion and social justice. When I was in graduate school, I worked for a research project known as The Longitudinal Study of Generations: a family study of more than 2,000 individuals over at least three decades. I was the data manager for that project. That drew me into the study of families and working with longitudinal survey research. After I graduated with my Ph.D., I did a one year (postdoctoral fellowship) in gerontology at the University of Southern California. Then I went to Texas Tech University as an Assistant Professor in the department of Human Development and Family Studies. It was there that I started working for the School of Nursing through research collaboration. I had an opportunity to collaborate with faculty in the School of Nursing to develop projects aimed at reducing health disparities and doing health promotion. That specific collaboration drew me to nursing and my substantive research on health disparities began.

Q: Could you tell us about your recent study?

A: The project was called the “Báa nnilah project”, and has since been completed. The study population was Native American adults living on or near the Apsáalooke, or the “Crow” reservation. This project started in 2016 and was a five-year project funded by the National Institute for Minority Health and Health Disparities. This was a community-based participatory research (CBPR) project aimed at improving or promoting self-management of chronic illness. The participants were twenty-five years or older adults who had at least one ongoing chronic illness. The research team recruited mentors from the community, and they were trained on [health intervention] topics. [They] also helped with recruiting participants, data collection, and intervention of these projects. We had ten mentors, and each of the mentors recruited about twenty participants. The intervention group went through seven sessions, and we collected data using what is called a PROMIS health survey along with physical tests. Those are standardized measurement-of-health surveys, measuring self-efficacy management of chronic illness, physical wellbeing, mental wellbeing, physical function, depression, and the study session of social roles and patient activation. 

Q: What are the skills or temperament that make a researcher effective?

A: I would say intellectual curiosity. You have to be really interested and want to know about what you care for. With that curiosity, that’s what motivated me to learn about how to do things. Also the research topic; for me it was social justice, that’s what motivated me to do health disparity research because that really might be the one thing I want to do in the world that will make a difference. So know what your passion is. You need some perseverance in order to acquire the training, but for perseverance, you need motivation.

Q: Are there any research projects that you still want to do?

A: Combining the methodologist and the statistician, I want to develop some skills in applying the systems science approach in health promotion. I’m interested in any health promotion that would help improve health, like chronic illness or infectious disease, but particularly among the underserved, low-income, and minorities. Some populations have multiple vulnerabilities that are due to income, ethnicity or race, or geographic location. The research literature has shown the effect of multiple vulnerabilities is huge. It’s rather exponential when people have not just one, but two vulnerability factors. 

Less than a year ago, I was on a sabbatical and I collaborated with colleagues from the University of Texas at Arlington and other institutions. I want to continue to use cutting-edge technologies, especially statistical methods that are either brand new or existing in other fields. For example, the systems science approach originated from engineering, but has been adopted by other areas, including social science, behavioral science, and nursing health research. I’m interested in using techniques or approaches developed in other disciplines and applying them in the health research domains. 

I’m very interested in mental and physical health, and I have learned a lot from my colleagues, nursing faculty, and nursing students. I think it is very important to use the best statistical methods to maximize the use of collected data and tell a true story. If used inappropriately, the results could be deceiving. It’s like that joke, “There are three kinds of lies: Lies, Damned Lies, and Statistics.” One of my goals is that I want our findings to be truthful as to what’s actually going on in the real world.

Visit our "Rebel Nursing Notes" library for more great content!