Graphs show trends in COVID-19 spread in Davis as experts explain how the data is analyzed
Healthy Davis Together has been collecting data from municipal wastewater autosamplers to assess the extent of the COVID-19 outbreak in Davis since Sept. 2020, and recently compiled and analyzed the wastewater data on its website.
As of March 31, the website shows data from seven regions, plotted on graphs of normalized virus concentrations against sampling dates, accompanied by citywide trends.
There are no numbers on the y-axis, which is intentional, according to Hannah Safford, a fourth-year Environmental Engineering Ph.D. student and core member of the wastewater team.
“As the normalized values don’t tend to mean much to people it’s common practice in the wastewater-surveillance world to leave numbers off of the y-axis so people can focus on trends in the data,” Safford said via email.
The data is normalized against a pepper virus called Pepper Mild Mottle Virus (PMMoV), UC Davis Health Director of Business Development Tod Stoltz explained. PMMoV is a virus that infects pepper plants and is commonly found in human feces, which scientists can identify in sewage treatment plants.
“The way they calculate [COVID-19 levels in wastewater] is that they compare the amount of DNA or RNA in the sample to what’s a fairly consistent amount of RNA from a pepper virus,” Stoltz said via email. “There’s a pepper virus that people generally consume and it’s at a fairly constant level in the wastewater, because people are always eating it and excreting it. So if you compare that COVID-19 RNA to how much pepper virus RNA is in the wastewater, then you can start to say whether there’s a higher amount than normal. It’s that ratio which is being presented in the data.”
Safford elaborated that normalizing against the pepper virus reduces the impact of confounding data.
“Normalizing by PMMoV also helps us adjust for factors like rainfall that can dilute wastewater samples,” Safford said.
The graphs on the wastewater testing website show both lines and crossmarks. The crossmarks represent exact values from the samples, while the lines represent the two-week moving average, which is used to smooth out the noise in the data, according to Safford.
“It’s common to use a moving average to assess trends in noisy temporal data such as the data we collect from environmental monitoring or clinical testing,” Safford said.
“We’re using a two-week moving average for [wastewater] data right now in part because we’re getting limited data points (only two per site per week) and in part because there’s an epidemiological reason: because COVID-19 infections typically run for 2–3 weeks, it’s reasonable to expect that we might see trends manifest over a two-week timeframe.”
UC Davis Assistant Professor in the Department of Civil and Environmental Engineering Heather Bischel elaborated via email that noise can be a considerable element in the data because toilets are flushed at all hours of the day.
“Some of the important sources of variability are the “fecal shedding rate” and the randomness of toilet flushing,” Bischel said. “An infected individual can shed a wide range of virus in their excrement… People will flush the toilet at different times of the day, creating mixtures of virus concentrations in the sewer system. To try to account for some of this variability in our sampling, we collect composite samples from the sewer system using automated samplers that draw a portion of the flow in the pipes every 15 minutes over a 24-hour period. The virus concentration we measure represents an averaged concentration from many people who flushed during the collection period.”
Bischel anticipated that the wastewater testing infrastructure could still be useful in the future, even after the COVID-19 pandemic has passed.
“Wastewater monitoring can continue to inform public health efforts in the future,” Bischel said. “While we are focused on supporting the pandemic response right now, my research group has been studying enteric viruses in wastewater and working to develop tools for more rapid virus detection and monitoring for our water systems.”
Written by: Rachel Shey — firstname.lastname@example.org