UW Dashboard Website Provides Latest COVID

Two University of Wyoming professors, with assistance from a UW graduate student, developed a public health tool to track COVID-19 data -- one that provides up-to-date information tailored for rural areas, such as Wyoming.

Pavel Chernyavskiy, in the UW Department of Mathematics and Statistics, and Tim Robinson, a statistics professor and the director of UW’s WWAMI (Washington, Wyoming, Alaska, Montana and Idaho) Medical Education Program, are the co-principal investigators of the project.

They recently produced the “Interactive COVID-19 Data by United States County” dashboard page. The dashboard establishes and maintains data visualization using maps related to COVID-19. UW’s Advanced Research Computing Center (ARCC) -- housed within the Office of Research and Economic Development -- hosts the webpage. The dashboard can be found at https://pathfinder.arcc.uwyo.edu:3839/covidvisualmap/.

Colton Zier, a recently graduated master’s degree statistics student from Basin, assisted on the ongoing project.

As the COVID-19 pandemic continues to spread nationwide, Chernyavskiy and Zier collaborated with the Wyoming Department of Health (WDH) on a tool that will help increase understanding of COVID-19 cases across the country, particularly in each Wyoming county. Earlier this fall, Chernyavskiy and Zier had preliminary discussions with Noah Hull and his team at WDH about the COVID-19 dashboard page.

“These collaborations will carry over beyond the current pandemic because all parties are interested in answering pressing public health questions,” Chernyavskiy says.

He adds that he is a believer in using science -- statistics, more precisely -- to do important things.

“I was happy that, this time, our training and research interests were useful in trying to understand this pandemic,” Chernyavskiy says. “In my field, we often work with data that have already been collected and, so, statistics is used to try to understand what happened in the past. It has been equal parts exciting and humbling to apply my training to live, real-time data, as this pandemic continues to unfold.”

The data for UW’s interactive webpage come from the COVID-19 Data Repository provided by the Center for Systems Science and Engineering at Johns Hopkins University. Data are updated daily and used to compile reports from individual state and county public health offices; reported data online are presented in a unified format, Chernyavskiy says. He adds that the UW dashboard page is different from others that can be found on the main website housing nationwide data.

“We focus on geographic disparities in incidence at a snapshot in time -- using data from a day ago -- rather than trends over time,” he says. “We also provide a statistically ‘smoothed’ map of COVID incidence that helps the reader see geographic patterns more clearly than by using raw data alone.”

He says this is important for rural areas, such as Wyoming and across the Mountain West, where counties tend to be sparsely populated, and incidence rates using raw data can be unrealistic.

UW’s dashboard is composed of three maps:

-- The first, “Three Day Incidence Rate,” provides a snapshot of the current conditions to the number of new cases per 100,000 people.

-- The second, “Trend Evaluation,” reflects whether the trend in new cases per day has improved, gotten worse or stayed about the same over the past week vs. the past month.

“The map can help inform public health officials whether local ordinances have had an impact on flattening the local curve,” Chernyavskiy says.

-- The third map, “Predicted Incidence Rate,” provides a “smoothed” (statistically modeled) version of incidence rates, which helps more easily to view statewide, regional and nationwide geographic patterns of risk.

“Rates computed in rural counties can be unrealistic due to low populations, which results in small denominators for those rates,” Chernyavskiy says. “Our ‘smoothed’ map is designed to help mitigate this issue by borrowing information from neighboring counties to optimally smooth out risk by using a statistical model.”

Chernyavskiy mentored Zier for his undergraduate capstone project in statistics, as well as the dashboard project, which served as Zier’s master’s degree project.

“Colton -- like most successful graduate students -- has a knack for getting into the nitty-gritty details, and I know he pulled more than a few all-nighters to get the website up and running,” Chernyavskiy says. “Colton also is very adept at building relationships -- this was an essential skill for our project because he had to manage stakeholders at the Wyoming Public Health Lab, ARCC and the Department of Mathematics and Statistics. I have greatly enjoyed watching Colton grow in this role and see many open doors in his future.”