Dr. Brian Dixon, Director of Public Health Informatics, and Dr. Nimish Valvi, Postdoctoral Fellow, at the Regenstrief Institute, are working on a research project aimed at understanding the impact of COVID-19 on immunosuppressed cancer patients in Indiana. Specifically, they aspire to compare the outcomes of newly diagnosed cancer patients (i.e. since 2019) who become infected with COVID-19 and those who do not become infected with COVID-19. Their study, funded by the IU Simon Comprehensive Cancer Center, is using data available in medical records to inform the treatment and care for patients with cancer. They are not exposing cancer patients to COVID-19 or experimenting on individuals who have cancer. They are simply using Big Data to better understand cancer patients and their risks from COVID-19 infection.
The Slate Project environment is perfect for analyzing the larger datasets necessary to infer populations trends and outcomes. Storage helps with large datasets and performance helps us do our work quickly and efficiently.
Brian E. Dixon
Given that cancer treatment compromises the body’s immune response, Dixon and Valvi felt cancer patients would be at higher risk for COVID-19 infection as well as complications from infection, including death. With limited data on cancer patients’ experiences with COVID-19, Dixon and colleagues from the Fairbanks School of Public Health and the IU School of Medicine seek to explore the clinical and sociodemographic factors associated with COVID-19 infection and complication risk.
This is why Dixon and Valvi are looking at how COVID-19 is impacting this vulnerable population in Indiana. While a lot of emphasis has been placed on the effect that COVID-19 has on older adults, those 65 years and up, the average age of a cancer patient is 44 years old. Complications from COVID-19 affect younger populations with cancer the same as older populations that do not have cancer. Dixon’s and Valvi’s research has shown that when cancer patients do get COVID-19, they are more likely to have severe consequences.
Computation speed is always a challenge with large retrospective observational studies like the one being conducted by Dixon and Valvi. Increasingly, team science requires analysis of large population data to guard against selection bias in smaller trials and studies. Dixon attests, “The Slate-Project environment is perfect for analyzing the larger datasets necessary to infer populations trends and outcomes. Storage helps with large datasets and performance helps us do our work quickly and efficiently.”
Slate-Project allows the entire team to access the large datasets and contribute to developing, testing, and running the scripts they need for analysis on them. During the pandemic, they were all working from home and needed a remote platform with solid performance in which to do their work. Slate-Project provided this platform. Even with unstable or slow internet connections from their homes, team members could log in, work on the project, then log off and pick it up later.
The Biomedical Informatics field is under increasing pressure to analyze large-scale health data quickly to inform policy and develop populations-level interventions. Traditional methods are not sufficient to rapidly analyze data and deliver results to policymakers and public health leaders. Research methods and tools need to continuously evolve to keep up with the private industry to advance not only science, but also population health. Drs. Brian Dixon and Nimish Valvi are committed to improving their research methods and tools in order to better serve the population and to support public health.