Analyzing inherent biases in SARS-CoV-2 PCR and serological epidemiologic metrics

  1. Monia Makhoul
  2. Farah Abou-Hijleh
  3. Shaheen Seedat
  4. Ghina R. Mumtaz
  5. Hiam Chemaitelly
  6. Houssein Ayoub
  7. Laith J. Abu-Raddad

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Background

Prospective observational data show that infected persons with the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) remain polymerase chain reaction (PCR) positive for a prolonged duration, and that detectable antibodies develop slowly with time. We aimed to analyze how these effects can bias key epidemiological metrics used to track and monitor SARS-CoV-2 epidemics.

Methods

An age-structured mathematical model was constructed to simulate progression of SARS-CoV-2 epidemics in populations. PCR testing to diagnose infection and cross-sectional surveys to measure seroprevalence were also simulated. Analyses were conducted on simulated outcomes assuming a natural epidemic time course and an epidemic in presence of interventions.

Results

The prolonged PCR positivity biased the epidemiological measures. There was a lag of 10 days between the true epidemic peak and the actually-observed peak. Prior to epidemic peak, PCR positivity rate was twofold higher than that based only on current active infection, and half of those tested positive by PCR were in the prolonged PCR positivity stage after infection clearance. Post epidemic peak, PCR positivity rate poorly predicted true trend in active infection. Meanwhile, the prolonged PCR positivity did not appreciably bias estimation of the basic reproduction number R 0 . The time delay in development of detectable antibodies biased measured seroprevalence. The actually-observed seroprevalence substantially underestimated true prevalence of ever infection, with the underestimation being most pronounced around epidemic peak.

Conclusions

Caution is warranted in interpreting PCR and serological testing data, and any drawn inferences need to factor the effects of the investigated biases for an accurate assessment of epidemic dynamics.

Article activity feed

  1. Version published to 10.1186/s12879-022-07425-z
  2. ScreenIT

    SciScore for 10.1101/2020.08.30.20184705: (What is this?)

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

    Table 1: Rigor

    NIH rigor criteria are not applicable to paper type.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    All analyses were conducted on the MATLAB R2019a platform.
    MATLAB
    suggested: (MATLAB, RRID:SCR_001622)

    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:
    This study has limitations. PCR testing was assumed random, but in reality this depends on the actual testing policy that is enforced in any setting. For instance, administering PCR testing to only those presenting with clinical symptoms will differentially bias detection towards those who acquired the infection within the last 5-10 days. The above results thus need to be complemented with further analysis for each specific setting to factor the actually-enforced testing policy. While the prolonged PCR positivity and the time delay in development of detectable antibodies are well-established in the literature [7, 8], it is still not sufficiently known how these are distributed in the population by age and sex, factors that may influence these effects and their impact on epidemiological measures. Epidemiological metrics can also be biased by other aspects of PCR and antibody testing, such as assay sensitivity and specificity, which are not investigated in the present study.

    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.

    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.

  3. Version published to 10.1101/2020.08.30.20184705 on medRxiv