Kirie Water Reclamation Plant

Interpreting results

Wastewater surveillance data is "noisy," meaning it is highly variable. People infected with SARS-CoV-2 shed the virus at different levels of intensity and for different lengths of time. Samples are also taken from sources (i.e. wastewater treatment plants) that are subject to a variety of environmental impacts, including weather events and industrial activity. For this reason we focus on trends in the data rather than specific measurements, specifically increases and decreases. A significant increase over time may indicate cases are increasing in a community.

Wastewater surveillance data should always be interpreted alongside other public health metrics.

Why are there two different colored data points?

We made a change to our lab methods in February 2022 to improve our ability to detect very small quantities of virus in wastewater. Our new method increases the amount of virus we are able to extract from each sample. Because wastewater contains more than just viruses (chemicals, food scraps, soaps, human waste, etc.), it can be tricky to isolate the viruses we are looking for. The new method helps us do just that.

Data generated with the previous method is not directly comparable with the new method.

What does the line mean?

Since wastewater data is highly variable, we include a trend line that fits a curve to the data to better visualize increases and decreases. We do this by "smoothing" the data using the Locally Weighted Scatterplot Smoothing (LOWESS) method, a local regression analysis. This method is useful because it reduces the influence of data outliers (i.e. sample readings that are much higher or much lower than the other samples taken at the same site during the same general time frame). We also include 95% confidence intervals around the trend line. This means we are 95% certain the actual trend line is within the intervals.

What is the difference between log and linear?

The linear scale uses an equal distance between steps on the vertical axis, adding a certain number for each new step (e.g., 1, 2, 3, 4 or 10, 20, 30, 40). A log scale uses an unequal distance between steps, multiplying a certain number for each new step. Typically, as we do here, this multiple is 10, so the axis might show steps at 1, 100, 1000, 10,000, etc.

Log scales are used to visualize data when the range of values shown is large and when viewing data over a longer time period. Public Health officials and scientists often work with data on these scales, while linear scales are more familiar to the general public.

Limitations of the data

Researchers are still working on understanding many open scientific questions, including being able to tell how many people are sick. Variability between wastewater sources, treatment facilities, and communities makes it difficult to translate the SARS-CoV-2 gene copy (gc) concentration into a measure of how many people are infected in the community. However, an upward/downward trend in SARS-CoV-2 gc/liter generally suggests a similar trend in the number of people infected within a given community.

For similar reasons related to the variability in each sample, comparing data from site-to-site is also discouraged.

Methodology

Wastewater is analyzed using digital polymerase chain reaction (dPCR) to determine the concentration of the SARS-CoV-2 virus in a sample. The nucleocapsid protein (N) gene of the virus is targeted in the assay, and results are reported in gene copies per liter (gc/L) of starting wastewater.

References

• Armbruster, D.A. et al., Limit of blank, limit of detection and limit of quantitation, Clin. Biochem. Rev. 29, Suppl 1, 2008
• Lu X, Wang L, Sakthivel SK, et al. US CDC Real-Time Reverse Transcription PCR Panel for Detection of Severe Acute Respiratory Syndrome Coronavirus 2. Emerging Infectious Diseases. 2020;26(8):1654-1665. doi:10.3201/eid2608.201246.