“Choose a major that you enjoy, especially if it’s unique.” That’s the advice that Colin Rogerson, MD, MPH, received as an undergraduate preparing for medical school. It was a move, his advisors said, that would help his medical school application stand out. Rogerson heeded the advice, deciding to major in statistics—his favorite subject in high school. Today, he’s using those skills as a physician-scientist focused on applying health informatics and data science to the study of severe pediatric respiratory diseases.
“Critical asthma is one of the most common conditions we care for in the PICU [pediatric intensive care unit], and although most children respond quickly to initial treatments, those that don’t are at risk for severe complications,” said Rogerson, assistant professor of pediatrics and a pediatric critical care physician at Riley Children’s Health. “My research focuses on using commonly collected clinical data to predict asthma phenotypes that may respond differently to different ancillary medications. My career goal is to provide evidence on how to identify which critical asthma patients will respond to which asthma therapies and ultimately improve their clinical outcomes.”
Joining IU School of Medicine in 2014 as a pediatric resident, Rogerson also completed a three-year pediatric critical care fellowship at the school and earned a master’s in public health at IU. An investigator in the Center for Biomedical Informatics at the Regenstrief Institute, Rogerson is using a K01 career development award from the National Heart, Lung and Blood Institute to fund his current research.
What inspired you to combine your role as a critical care physician with work in clinical research?
During my fellowship in 2019, a few of my mentors took me to the Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) research conference. I was introduced to a group of critical care physicians that specialized in informatics and data science research in the pediatric critical care population. One researcher presented his study using statistical algorithms to classify pediatric patients with sepsis into specific phenotypes, or groups of sepsis that had different clinical features and outcomes. After the meeting I was hooked and decided this is what I wanted to do with my career.
What interests you about research in biomedical informatics as it relates to severe pediatric respiratory disease?
In the Riley PICU, we take care of the sickest children in the state. I quickly learned that although these are the patients that need evidence-based care the most, they are the population with the least available evidence. This is due to many factors, including small sample sizes and highly heterogeneous disease states, leaving us with a lot of variability in practice because there isn’t great scientific evidence to tell us the best thing to do in many situations. Rather than trying to answer every clinical question with randomized controlled trials, which are extremely valuable but also expensive and time consuming, data science provides cutting-edge statistical methods to compile and analyze data to try and answer these vital clinical questions. I chose to focus on respiratory disease because it is the most common condition cared for in the PICU, which provides a great deal of data to study, and we have many unanswered questions on the best way to care for these patients.
Do you have any recent developments or findings to share related to your work?
I am currently working on a large project as part of my NIH K01 award that involves using PEDSnet, a national research collaborative of large academic children’s hospitals that share research ideas and granular electronic health record data. Using this database, I have built a cohort of more than 80,000 hospital encounters for pediatric asthma. In this cohort, I have been able to use unsupervised machine learning methods to identify a subgroup of patients who have a more atopic or allergic type of asthma that is more severe. The results are preliminary, but it appears that we can identify these patients using only demographic information, vital signs and a medical history that would all be available shortly after presentation to the hospital. This could potentially alter the treatments we provide and the counsel we give to families.
How has IU School of Medicine supported your research efforts?
I was given salary support that provides more time for me to learn essential research skills and conduct early research. I benefited from tuition assistance to obtain my MPH degree in public health informatics. Additionally, I have received phenomenal formal mentorship through the Regenstrief Institute, the Department of Biostatistics and Health Data Science, the Wells Center and the Children’s Health Services Research Institute, as well as informal mentorship through the I3 program [now the Ascent Scholars Program]. I also received several early grants through the Division of Pediatric Critical Care.
What are your plans for future research?
After completing my K01 project, I plan to apply for an R01 grant to implement my predictive model into an electronic health record (EHR) system and prospectively work to classify asthma patients into clinical phenotypes in real time. I am also interested in securing R01 grant funding to validate a computational asthma phenotype I developed to identify hospitalized asthma patients in EHR data. My plans for future research also include a project that uses a novel data source called the PICU Data Collaborative to further evaluate the efficacy of ancillary asthma medications in PICU patients, as well as a project that evaluates the efficacy of different respiratory support devices commonly used in pediatric asthma.
What do you find most rewarding about your work as a physician-scientist?
I love the logical science involved in conducting research and using innovative approaches to answer clinical questions. I love that these answers don’t simply satisfy curiosity but also improve the lives of critically ill children and their families. Recently, I was asked to review a quality improvement paper in a peer-reviewed medical journal. In this paper, the authors had taken evidence I generated in my previous work and applied it to their PICU treatment approach to critical asthma. After applying these approaches, their clinical outcomes improved. It was extremely rewarding to see that the work I’m doing as a physician-scientist is being recognized and used by people in other institutions to make outcomes better for the children they care for.
