If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Are antigenic tests useful for detecting SARS-CoV-2 infections in patients accessing to emergency departments? Results from a North-West Italy hospital
Department of Clinical and Biological Science, University of Turin, Regione Gonzole 10, Orbassano 10043, ItalyEpidemiology Unit, Regional Health Service ASL TO3, Grugliasco
Department of Epidemiology and Biostatistics, School of Public Health, Imperial College London, London, United KingdomMRC Centre for Environment and Health, Imperial College, London, United Kingdom
Department of Clinical and Biological Science, University of Turin, Regione Gonzole 10, Orbassano 10043, ItalyEpidemiology Unit, Regional Health Service ASL TO3, Grugliasco
In the article “Clinical application of a rapid antigen test for the detection of SARS-CoV-2 infection in symptomatic and asymptomatic patients evaluated in the emergency department: A preliminary report.”,
Clinical application of a rapid antigen test for the detection of SARS-CoV-2 infection in symptomatic and asymptomatic patients evaluated in the emergency department: a preliminary report.
Turcato et al. presented a study on the use of rapid antigenic tests (Ag-RDTs) instead of the usual real time reverse transcription polymerase chain reaction (RT-PCR) assay to detect the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) in the context of Emergency Departments (ED). They observed a general good sensitivity and specificity, lower in the subgroup of asymptomatic patients. Their conclusion is in favour of the use of Ag-RDTs in EDs as an additional tool to address the challenge of containing the SARS-CoV-2 pandemic.
We agree with the authors that the development of reliable but cheaper and faster point-of-care diagnostic tests was expected to be useful either for population-screening or as first aid tests in the emergency room.
Recommendations for use of antigenic tests in the diagnosis of acute SARS-CoV-2 infection in the second pandemic wave: attitude in different clinical settings.
Data on the sensitivity and specificity of currently available Ag-RDTs derive from studies that vary in design, setting, population and type of specimen, thus strongly limiting the comparability and ability to make general inferences. Sensitivity appears to be highly variable, ranging from 29 to 94% compared to the RT-PCR test, but specificity is consistently high (>97%).
Omi K., Takeda Y., Mori M. SARS-CoV-2 qRT-PCR Ct value distribution in Japan and possible utility of rapid antigen testing kit. medRxiv. 2020:2020.06.16.20131243. 12.
Weitzel T., Legarraga P., Iruretagoyena M., et al. Head-to-head comparison of four antigen-based rapid detection tests for the diagnosis of SARS-CoV-2 in respiratory samples. bioRxiv. 2020:2020.05.27.119255.
Omi K., Takeda Y., Mori M. SARS-CoV-2 qRT-PCR Ct value distribution in Japan and possible utility of rapid antigen testing kit. medRxiv. 2020:2020.06.16.20131243. 12.
which usually happens in the pre-symptomatic (0.5–3 days before symptom onset) and early symptomatic phases of the illness (within the first week from symptom onset) but limited data are available about other possible individual modifiers of the accuracy of the assay. A recent Cochrane review highlighted that patients’ characteristics were not available or poorly detailed in many studies, with only three out of 22 studies coming from an ED setting.
Between October 26th and November 10th 2020, 455 patients accessed the ED of San Luigi Gonzaga University Hospital in Orbassano (Turin, Italy) and 324 underwent both RT-PCR and Ag-RDT testing. This period corresponds to the first two weeks of the second pandemic wave, with a weekly incidence of SARS-CoV-2 infection in the Region of about 500 confirmed cases/100,000 inhabitants. Data were obtained as part of an observational study described elsewhere
Clinical characteristics and management of COVID-19 patients accessing the emergency department in a hospital in Northern Italy in March and April 2020.
and a detailed presentation of methods is available in supplementary material.
The prevalence of SARS-CoV-2 infection in this cohort was 65% measured using RT-PCR as a gold standard. Supplementary Table 1 reports test results: 275 (84%) patients showed concordant results (168 positive and 107 negative), while 49 (15%) showed discordant results (42 patients had a positive RT-PCR and a negative Ag-RDT and 7 vice versa). Cohen's Kappa Statistics (k = 0.68 – 95% CI 0.61–0.77) highlighted substantial agreement. Specificity and sensitivity of Ag-RDT were 0.939 (95% CI: 0.895–0.983) and 0.800 (95% CI: 0.746–0.854), respectively, taking RT-PCR as the reference. Overall, the Ag-RDT positive predictive value was 0.960 (95% CI 0.931–0.989), and the negative predictive value was 0.718 (95% CI: 0.646–0.790). The variation of positive and negative predictive values due to difference in prevalence can be observed in Supplementary Table 2 and Fig. 1. Positive predictive value could vary from 0.12, when the prevalence of the disease is 0.01, to 0.77 when the prevalence is 0.20. The negative predictive value could vary from about 1, considering a low prevalence (0.01) to 0.95, considering a higher prevalence (0.20).
Fig. 1Positive and negative predictive value estimates in relation to prevalence, using the sensitivity and specificity of the test found in our population (sensitivity = 0.800; specificity = 0.939).
No difference in patients’ characteristics between true positive and false negative tests was observed (Supplementary Table 3). On the contrary, false negative patients were significantly younger and they were tested significantly later after symptoms onset compared with true negative patients (Table 1). Moreover, fever (64.3% vs 19.6%, p < 0.0001) and cough (42.9% vs 15.0%, p = 0.0003) were significantly more frequent in false than true negatives, while chronic obstructive pulmonary disease was more frequent in true than false negatives, with a borderline significance (16.5% vs 4.8%, p = 0.06). Few true negative patients had bilateral pneumonia (n = 10, 9.4%), that was highly present in false negative patients (n = 25, 61.0%, p-value for difference < 0.0001) and multivariable analysis confirm these results, suggesting that wrong group allocation for negative patients occurred more frequently in patients with fever, cough, and pneumonia, while it was less likely in patients with COPD.
Table 1Ag-RDT negative patients: comparison of patients’ characteristics between true negative and false negative patients. Wilcoxon sum rank test (quantitative variables) and chi-square or Fisher's exact test (qualitative variables) are used and multivariable logistic model (including significant variables) to evaluate the association between being a false negative and patients’ characteristics.
True negative (n = 107)Mean (SD), medianor Frequency (%)
False negative (n = 42)Mean (SD), medianor Frequency (%)
: the ideal test in a crowded ED context should help in identifying asymptomatic patients arriving to the ED for reasons other than COVID-19, who are concurrently found COVID-19 positive.
Our results suggest that a negative Ag-RDT test should not exclude COVID-19 in patients that clinically have symptoms that are strongly suggestive of COVID-19. Ag-RDTs alone had a low negative predictive value (we cannot trust a negative result of the test), thus they need to be evaluated in association with clinical judgement. A high level of suspicion should be maintained in patients with fever, cough or pneumonia notwithstanding a negative Ag-RDT. Since the predictive value is strictly related to the prevalence of disease, and then to the pre-test odds, Ag-RDTs are not really useful in settings where the prevalence of disease is high or in patients with high pre-test odds. On the contrary, in periods with low prevalence of the disease or in patients with a low pre-test odds (asymptomatic) or with symptoms probably related to a known COPD, Ag-RDTs can be used alone and we can trust a negative result.
In conclusion, our results confirm the limits of antigenic tests as first line screening tests in settings with high prevalence of disease or in patients with high pre-test odds, where a negative test is not informative (i.e. in ED in a pandemic period). This suggests that in these situations the antigenic test should be integrated in a clinical algorithm.
Declaration of Competing Interest
All authors declare no conflict of interests.
Acknowledgments
Authors would like to thank Marco Alvich, Rebecca Tasca, Sara Roetti, and Selene Demaria for their help in collecting data.
Funding
The authors received no specific funding for this work.
Clinical application of a rapid antigen test for the detection of SARS-CoV-2 infection in symptomatic and asymptomatic patients evaluated in the emergency department: a preliminary report.
Recommendations for use of antigenic tests in the diagnosis of acute SARS-CoV-2 infection in the second pandemic wave: attitude in different clinical settings.
Omi K., Takeda Y., Mori M. SARS-CoV-2 qRT-PCR Ct value distribution in Japan and possible utility of rapid antigen testing kit. medRxiv. 2020:2020.06.16.20131243. 12.
Weitzel T., Legarraga P., Iruretagoyena M., et al. Head-to-head comparison of four antigen-based rapid detection tests for the diagnosis of SARS-CoV-2 in respiratory samples. bioRxiv. 2020:2020.05.27.119255.
Clinical characteristics and management of COVID-19 patients accessing the emergency department in a hospital in Northern Italy in March and April 2020.
Fisher's exact test.
00255-3&id=gr1.jpg){kind=link}