Cost-Effective Sampling Strategies for Wastewater Surveillance: A Large-Scale Longitudinal Study in Hong Kong and Shenzhen

TL;DR

This study explores cost-effective ways to reduce wastewater sampling while still tracking SARS-CoV-2 trends in Hong Kong and Shenzhen.

Contribution

The paper introduces two new downsampling strategies to optimize wastewater surveillance for cost-effectiveness.

Findings

Halving sampling sites did not significantly affect the detection of citywide SARS-CoV-2 trends.

Sampling three times per week was sufficient to capture virus transmission patterns during the Omicron outbreak.

Factors like flow rate and population size influence optimal site selection and trend representativeness.

Abstract

During the COVID-19 pandemic, wastewater surveillance
undoubtedly
served as a useful public health surveillance tool and catalyzed the
establishment of over 4600 sampling sites and 195 wastewater surveillance
dashboards worldwide. However, in the postpandemic era, the continuous
regular operation of these sites has become labor-intensive and costly.
In this study, we established two downsampling strategies (i.e., Enumerative
Method and Iterative Hierarchical Method) to explore the potential
of reducing the number of sampling sites and sampling frequencies
without significantly affecting the observed SARS-CoV-2 transmission
trends. To evaluate the method’s effectiveness, we comprehensively
applied them to two intensive large-scale wastewater surveillance
data sets from two adjacent major cities, i.e., a 9-month monitoring
across 12 sampling sites in Hong Kong and a 5-month monitoring across
38 sampling sites in Shenzhen. We found that consistent citywide SARS-CoV-2
transmission trends were captured in the two cities before and after
halving the sampling sites. Additionally, we observed that multiple
factors (i.e., flow rate, serving population size, number of sampling
sites, geographical locations, etc.) influenced the site selection
and trend representativeness, which should be comprehensively considered
in the design of a sampling strategy. Moreover, sampling three times
per week could reflect the virus transmission patterns during the
Omicron outbreak. These findings highlight the necessity to optimize
current sampling practices in most wastewater surveillance networks
so that they can maximize the value of wastewater surveillance and
achieve favorable cost-effectiveness for sustainable long-term monitoring.

Figures