Regardless of sampling plan or purpose, any new sewage surveillance strategy should use a documented procedure and follow it consistently. Additionally, regardless of the type, location, or frequency of samples collected, all sampling strategies should consider the below characteristics of their samples:
- The representation of desired sources in samples being collected, which will strongly depend on the question being addressed, sample type, and wastewater system characteristics. Ensuring that samples collected are representative of the population being sampled is critical to generating results that can be interpreted appropriately. However, there are also tradeoffs between obtaining a perfectly representative sample and logistical barriers to collecting this sample efficiently and effectively. Important considerations to ensure sample representation include:
- How many people are contributing fecal material to the sample location?
- How well-mixed is the sample location?
- Confounding factors. Small differences in measured SARS-CoV-2 concentrations could simply result from analytical precision issues, uneven sampling of the population (i.e. sample is not well-mixed & representative of the population), or variation in the composition of the sewage itself. Therefore, samples collected over the course of several weeks could vary in virus concentration by 2- to 3-fold, even with consistent COVID-19 prevalence in the population. However, with the proper controls and knowledge, many of these confounding factors can be appropriately considered in order to obtain meaningful and informative data.
A recently published white paper from the Water Research Foundation also provides useful information on the logistics of sample design and collection schemes: https://www.waterrf.org/sites/default/files/file/2020-06/COVID-19_SummitHandout-v3b.pdf
