Evaluating diagnostic accuracies of Panbio™ test and RT-PCR for the detection of SARS-CoV-2 in Addis Ababa, Ethiopia using Bayesian Latent-Class Models (BLCM)

  1. Abay Sisay
  2. Sonja Hartnack
  3. Abebaw Tiruneh
  4. Yasin Desalegn
  5. Abraham Tesfaye
  6. Adey Feleke Desta

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

Rapid diagnostics are vital for curving the transmission and control of the COVID-19 pandemic. Although many commercially available antigen-based rapid diagnostic tests (Ag-RDTs) for the detection of SARS-CoV-2 are recommended by the WHO, their diagnostic performance has not yet been assessed in Ethiopia. So far, the vast majority of studies assessing diagnostic accuracies of rapid antigen tests considered RT-PCR as a reference standard, which inevitably leads to bias when RT-PCR is not 100% sensitive and specific. Thus, this study aimed to evaluate the diagnostic performance of Panbio ™ jointly with the RT-PCR for the detection of SARS-CoV-2.

Methods

A prospective cross-sectional study was done from July to September 2021 in Addis Ababa, Ethiopia, during the third wave of the pandemic involving two health centers and two hospitals. Diagnostic sensitivity and specificity of Panbio ™ and RT-PCR were obtained using Bayesian Latent-Class Models (BLCM).

Results

438 COVID-19 presumptive clients were enrolled, 239 (54.6%) were females, of whom 196 (44.7%) had a positive RT-PCR and 158 (36.1%) were Panbio ™ positive. The Panbio ™ and RT-PCR had a sensitivity (95% CrI) of 99.6 (98.4–100) %, 89.3 (83.2–97.6) % and specificity (95% CrI) of 93.4 (82.3–100) %, and 99.1 (97.5–100) %, respectively. Most of the study participants, 318 (72.6%) exhibited COVID-19 symptoms; the most reported was cough 191 (43.6%).

Conclusion

As expected the RT-PCR performed very well with a near-perfect specificity and a high, but not perfect sensitivity. The diagnostic performance of Panbio ™ is coherent with the WHO established criteria of having a sensitivity ≥80% for Ag-RDTs. Both tests displayed high diagnostic accuracies in patients with and without symptoms. Hence, we recommend the use of the Panbio ™ for both symptomatic and asymptomatic individuals in clinical settings for screening purposes.

Article activity feed

  1. Version published to 10.1371/journal.pone.0268160
  2. ScreenIT

    SciScore for 10.1101/2022.04.25.22274285: (What is this?)

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

    Table 1: Rigor

    EthicsField Sample Permit: Methods of Sample collection and data collection instruments: We collected the nasopharyngeal respiratory specimens twice from each study participant upon consent using viral transport medium: one for rapid antigen and the other for RT-PCR.
    Consent: Written informed consent on the use of data with full anonymity was obtained from the voluntary participants.
    Sex as a biological variablenot detected.
    Randomizationnot detected.
    Blindingnot detected.
    Power Analysisnot detected.
    Cell Line AuthenticationAuthentication: Convergence was assessed by visual inspection of the trace plots and the potential scale reduction factor (Gelman Rubin statistic) being below 1.1.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    Ag-RDT SARS-CoV-2 testing: The collected nasopharyngeal swab was processed on site using the Panbio™ Ag-RDT (Abbott Diagnostic GmbH, Germany).
    Abbott
    suggested: None
    Data analysis and Interpretation: Descriptive data of the research was entered and analyzed using SPSS statistical software version 23.
    SPSS
    suggested: (SPSS, RRID:SCR_002865)
    Bayesian latent class model (BLCM): With the aim to obtain diagnostic tests accuracies in the absence of a perfect gold standard, Bayesian latent class models (BLCM) were fit to the data following the approach from Hui and Walter for two tests and four populations with MCMC (Markov chain Monte Carlo) simulation to construct posteriors in JAGS version 4.3.0 [32] using the runjags package [33].
    JAGS
    suggested: (rjags, RRID:SCR_017573)

    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:
    However, because of resource limitation, we did not monitor the viral concentration of the clients and readers should consider it while inferring our finding with these studies [12, 15, 43, 44]. Our finding revealed the Panbio™ rapid antigen test have highest level of agreement of the inter-test agreement beyond chance of the established assays (RT-PCR) with an average Cohen’s kappa value of 81% (95% CI: 75.82% to 87.45%) among our study subjects of 438 COVID-19 suspected individuals, which was consistent with similar studies [36, 37] and lower finding compared with Torres et. al., where the defined study population was different form the present study [45]. The diagnostic performance Panbio™ Covid-19 Ag rapid test highly correlate with Ct values and day of clinical onset. We found a superior performance in lowest ct values, which is most probably during at this condition the viral concentration become high and as the clinical onset of the day increase its performance among asymptomatic cases [37, 41]. Most of the false negative Ag-RDT results were among the population with contacts of SARS-CoV-2 confirmed patient within a week range of contact of confirmed case and they have reported no clinical symptoms and chief complaints when arriving at the testing health facilities. Their age ranged from 18 to 68 years old and most of them were female. Also there was a single false positive report from the total non-diseased of 242. The low frequency of false positive report of the AG-...

    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.
    • Thank you for including a protocol registration statement.

    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.

  3. Version published to 10.1101/2022.04.25.22274285v1 on medRxiv