Clinical validation of rapid instrument-free SARS-CoV-2 detection by the pH-EVD method. A) Procedures of the instrument-free detection method and the routine RT-qPCR assay for testing clinical NP samples and contrived saliva samples. B,C) _C_ q values of the 33 saliva samples and 30 clinical NP samples by RT-qPCR following spin column-based extraction, respectively. D,E) Visual detection results of the instrument-free pH-EVD method for the saliva and NP samples, respectively. F) Confusion matrix describing the overall performances of the two assays between positive and negative samples. The cutoff of _C_ q was 40. The spin column-based extraction and RT-qPCR are considered as the standards. The samples marked with red were tested to be negative by the pH-EVD method, while positive by RT-qPCR. Each experiment was repeated three times. Figure 7B,C displays the _C_ q values for contrived saliva samples (_n_ = 33) and clinical NP swab samples (_n_ = 30) by RT-qPCR, respectively. As per the cutoff (_C_ q = 40) suggested by the CDC,[ 36 ] 31 samples were tested to be positive (17 for saliva and 14 for the NP swab) and 32 samples were negative (16 for both saliva and the NP swab). Figure 7D,E shows the visual detection results of the pH-EVD method on testing the saliva and NP swab samples, respectively. By contrast, most of the pH-EVD methods’ results conform with those of RT-qPCR, except the saliva samples of 3, 16, and 23, and the NP swabs of 7, 10, 19, and 24. For these seven samples, the _C_ q values ranged from 36.22 to 39.48, very close to the cutoff, therefore considering them to be very weak positives. With spin column-based extraction and RT-qPCR set as the standards, the confusion matrix indicates 77.4% sensitivity, 100% specificity, and 100% precision for our pH-EVD assay (Figure 7F). The reduced sensitivity is likely attributed to a small volume of samples (28 μL) supplied for pH-paper-based RNA extraction, compared to that (140 μL) for spin column-based extraction. However, if focusing on the samples with the cutoff of _C_ q = 35, our pH-EVD platform shows 100% sensitivity and 100% specificity (Figure S5, Supporting Information). Therefore, as a simple, low-cost, instrument-free diagnostic approach, our pH-EVD-based method provides a promising diagnostic alternative for rapid SARS-CoV-2 diagnostics as a preliminary screening tool.