Interlaboratory findings published (2019)

WRF Project #5089

Reproducibility and sensitivity of 36 methods to quantify the SARS-CoV-2 genetic signal in raw wastewater: findings from an interlaboratory methods evaluation in the U.S. is now available as Journal article in Environmental Science: Water and Technology at https://pubs.rsc.org/en/Content/ArticleLanding/2021/EW/D0EW00946F#!divAbstract.

• This interlaboratory assessment evaluated the reproducibility and sensitivity of 36 standard operating procedures (SOPs), divided into eight method groups based on sample concentration approach and whether solids were removed.

• The 32 participating laboratories included 17 academic labs, 6 commercial labs, 4 non-municipal government labs, 3 municipalities, and 2 manufacturers of molecular tests.

• Discussion includes (emphases added):

  This study demonstrated that a diverse set of 36 methods was able to quantify the SARS-CoV-2 genetic signal in raw wastewater with a high degree of reproducibility. 80% of the data from the eight different method groups fell within a band of approximately ±1 log GC/L when corrected for recovery. This finding bodes well for the nationwide interest in tracking SARS-CoV-2 in raw wastewater since a single standardized method may not be critical for obtaining comparable results between laboratories. Access to multiple, reliable methods may also increase the number of labs capable of participating in monitoring efforts and provide resilience against supply chain issues that have beset these efforts during the pandemic.

  The findings also show, however, that methods-related hurdles remain before using the data for watershed-based epidemiology and modeling (e.g., estimating incidence and prevalence). This end use requires obtaining accurate information on the absolute concentration of SARS-CoV-2 genetic material in raw wastewater in addition to other information such as fecal shedding rates as noted below. Unfortunately, the accuracy of the methods—i.e., their ability to correctly quantify the true number of SARS-CoV-2 genome copies—could not be assessed because the actual concentrations in the raw wastewater samples were unknown. Despite the relatively tight band of results (80% within ±1 log), this 2 log range may be too wide for estimating community infection since 2 logs represents the difference between 1% and 100% of the population being infected. Additional data gaps must also be addressed for accurately modeling community infections including information on a) viral shedding rates in feces during different stages of infection,6,20,21b) how the genetic signal changes during travel through the wastewater collection system,22–24 and c) sewershed modeling to estimate travel time and dilution. Multiple efforts should be pursued to address these knowledge gaps.

  The findings are encouraging, however, for tracking changes or trends in virus concentrations. For this purpose, the absolute numbers quantified are not as important as identifying when and to what degree those numbers are increasing or decreasing.25 The collection of SARS-CoV-2 wastewater concentrations could be used in conjunction with clinical data to provide complementary information on the extent of community infection and the effectiveness of public health interventions. The data could also be used to identify “hot spots” within a collection system where higher virus concentrations are measured.7–9 This knowledge could be used to trigger additional investigations of the populations within that sub-sewershed to identify and respond to communities experiencing higher infection rates. One benefit of this type of tracking is that the changes in wastewater concentrations may precede the clinical evidence of infection by multiple days, allowing for more responsive and focused public health interventions. A related use of this approach is confirmation of ongoing low community prevalence of SARS-CoV-2 in areas, such as small rural regions, for which testing rates are low. The use of wastewater surveillance as a sentinel for community infection has been described in Utah and at the University of Arizona.11

  This study's findings would suggest that the same method or laboratory be used to assess the SARS-CoV-2 concentrations over time at a given set of locations. For example, use method A to assess trends within the sewersheds in region X over time rather than switching between methods A, B, and C over the monitoring period. Other regions (e.g., region Y) could select different methods, but should then use the same method over the entire testing period to facilitate the tracking of trends. One exception to this may be cases in which multiple laboratories use a similar SOP and have demonstrated a high degree of reproducibility across labs, such as SOPs 4.1, 4.2, and 4.3. Given the high degree of intra-method reproducibility observed (standard deviation <0.2 log GC/L), many methods have sufficient precision to sensitively detect when changes in virus concentrations are occurring. Collecting samples at multiple locations will also help identify where they are occurring.

• Conclusions include (emphases added):

  • A nationwide interlaboratory comparison of methods for the quantification of SARS-CoV-2 genetic signal in wastewater showed a high degree of reproducibility. 80% of the results from eight method groups (36 different methods) fell within a band of approximately ±1 log GC/L when corrected for recovery. These findings suggest that a variety of methods are capable of producing reproducible results, though the same SOP or laboratory should be selected to track SARS-CoV-2 trends at a given facility.

  • Based on the seven order of magnitude range of recovery efficiencies reported in this study, it is recommended that future methods include matrix spikes to quantify and correct for recovery in order to obtain reproducible numbers between methods.

  • Recovery-corrected results did not show a systematic impact from solids removal or concentration method used. Additional methods steps that were evaluated (e.g., pasteurization, primer set selection, and PCR platform) generally resulted in small differences compared to other sources of variability.

  • Factors leading to greater interlaboratory reproducibility include a) the relative insensitivity of the findings to methodological differences, b) the implementation of strict QA/QC requirements, c) the use of a quality assurance project plan to normalize the findings and account for important sources of variability, and d) implementing a shared SOP among different laboratories.

  • The findings support the use of wastewater surveillance for tracking trends in the concentrations of SARS-CoV-2 within communities. They also highlight methodological challenges related to modeling incidence and prevalence.

  • Additional metrics should be used to select the best methods for future efforts including method sensitivity, cost, equipment requirements, and simplicity.

This 2019 Journal article grows out Trussell Technologies’ December 16, 2020, report under Water Research Foundation Project #5098. See https://www.waterrf.org/research/projects/interlaboratory-and-methods-assessment-sars-cov-2-genetic-signal-wastewater.