Interpreting Wastewater SARS-CoV-2 Results using Bayesian Analysis

  1. Kyle Curtis
  2. Raul A. Gonzalez

Reviewed by

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

Wastewater surveillance of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has proven a practical complement to clinical data for assessing community-scale infection trends. Clinical assays, such as the CDC-promulgated N1, N2, and N3 have been used to detect and quantify viral RNA in wastewater but, to date, have not included estimates of reliability of true positive or true negative. Bayes’ Theorem was applied to estimate Type I and Type II error rates for detections of the virus in wastewater. Conditional probabilities of true positive or true negative were investigated when one assay was used, or multiple assays were run concurrently. Cumulative probability analysis was used to assess the likelihood of true SARS-CoV-2 detection using multiple samples. Results demonstrate highly reliable positive (>0.86 for priors >0.25) and negative (>0.80 for priors = 0.50) results using a single assay. Using N1 and N2 concurrently caused greater reliability (>0.99 for priors <0.05) when results concurred but generated potentially counterintuitive interpretations when results were discordant. Regional wastewater surveillance data was investigated as a means of setting prior probabilities. Probability of true detection with a single marker was investigated using cumulative probability across all combinations of positive and negative results for a set of three samples. Findings using a low (0.11) and uniformed (0.50) initial prior resulted in high probabilities of detection (>0.95) even when a set of samples included one or two negative results, demonstrating the influence of high sensitivity and specificity values. Analyses presented here provide a practical framework for understanding analytical results generated by wastewater surveillance programs.

Article activity feed

  1. ScreenIT

    SciScore for 10.1101/2021.08.17.21262165: (What is this?)

    Please note, not all rigor criteria are appropriate for all manuscripts.

    Table 1: Rigor

    Ethicsnot detected.
    Sex as a biological variablenot detected.
    Randomizationnot detected.
    Blindingnot detected.
    Power Analysisnot detected.

    Table 2: Resources

    No key resources detected.

    Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    Limitations and Assumptions: Applying a Bayesian framework for interpreting analytical findings is fundamentally driven by the sensitivity and specificity values assigned to the assays in question, along with the assumed prior. As a result, the accuracy of these values is highly consequential for the analysis. For this preliminary exercise, in vivo specificity values similar to those reported by Lu et al. (2020) were used in conjunction with a novel in situ approach to determining sensitivity values from a regional SARS-CoV-2 wastewater monitoring program. As noted, specificity values of 0.999 were used in place of the 1.0 specificities reported by Lu et al. (2020) for the N1 and N2 assays. The goal of this approach was to assume a false positive rate of 1/1000 in order to conservatively examine the reliability of analytical results when neither sensitivity nor specificity are perfect. Where possible, future studies should conduct in-house specificity testing using their own analytical methodology performed on the matrix being studied to derive lab-specific specificity values. A similar recommendation is suggested for deriving lab-specific sensitivity values and is likely more impacted by wastewater matrix effects such as inhibition. After establishing assay sensitivity and specificity, prior probability is the final value required to conduct the simple conditional probability analyses described above. Prior probability is another critical piece of this framework as it also e...

    Results from TrialIdentifier: No clinical trial numbers were referenced.

    Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

    Results from JetFighter: We did not find any issues relating to colormaps.

    Results from rtransparent:
    • Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
    • Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
    • No protocol registration statement was detected.

    Results from scite Reference Check: We found no unreliable references.

    About SciScore

    SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

  2. Version published to 10.1101/2021.08.17.21262165 on medRxiv