ABSTRACT
BACKGROUND: The standard test that identifies the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) is based on reverse transcriptase-polymerase chain reaction (RT-PCR) from nasopharyngeal (NP) swab specimens. We compared the accuracy of a rapid antigen detection test using exhaled breath condensate by a modified Inflammacheck® device with the standard RT-PCR to diagnose SARS-CoV-2 infection. METHODS: We performed a manufacturer-independent, cross-sectional, diagnostic accuracy study involving two Italian hospitals. Sensitivity, specificity, positive (PLR) and negative likelihood ratio (NLR), positive (PPV) and negative predictive value (NPV) and diagnostic accuracy with 95% confidence intervals (95% CI) of Inflammacheck® were calculated using the RT-PCR results as the standard. Further RT-PCR tests were conducted on NP specimens from test positive subjects to obtain the Ct (cycle threshold) values as indicative evidence of the viral load. RESULTS: A total of 105 individuals (41 females, 39.0%; 64 males, 61.0%; mean age: 58.4 years) were included in the final analysis, with the RT-PCR being positive in 13 (12.4%) and negative in 92 (87.6%). The agreement between the two methods was 98.1%, with a Cohen's κ score of 0.91 (95% CI: 0.79-1.00). The overall sensitivity and specificity of the Inflammacheck® were 92.3% (95% CI: 64.0%-99.8%) and 98.9% (95% CI: 94.1%-100%), respectively, with a PLR of 84.9 (95% CI: 12.0-600.3) and a NLR of 0.08 (95% CI: 0.01-0.51). Considering a 12.4% disease prevalence in the study cohort, the PPV was 92.3% (95% CI: 62.9%-98.8%) and the NPV was 98.9% (95% CI: 93.3%-99.8%), with an overall accuracy of 98.1% (95% CI: 93.3%-99.8%). The Fagan's nomogram substantially confirmed the clinical applicability of the test in a realistic scenario with a pre-test probability set at 4%. Ct values obtained for the positive test subjects by means of the RT-PCR were normally distributed between 26 and 38 cycles, corresponding to viral loads from light (38 cycles) to high (26 cycles). The single false negative record had a Ct value of 33, which was close to the mean of the cohort (32.5 cycles). CONCLUSIONS: The modified Inflammacheck® device may be a rapid, non-demanding and cost-effective method for SARS-CoV-2 detection. This device may be used for routine practice in different healthcare settings (community, hospital, rehabilitation).