Skip to main content
. 2020 May 18;9(5):1515. doi: 10.3390/jcm9051515

Table 1.

Summary of studies included in the meta-analysis (Note: TP = No. of True Positive Cases; FP = No. of False Positive Cases; FN = No. of False Negative Cases; TN = No. of True Negative Cases; Se. = sensitivity; Sp. = specificity; PPV = predicted positive value; PNV = predicted negative value; Cohen’s Kappa values should be interpreted as follows: 0.0–0.20 no agreement, 0.21–0.39 minimal agreement, 0.40–0.59 weak agreement, 0.60–0.79 moderate agreement, 0.80–0.90 strong agreement, > 0.90 almost perfect agreement; POCT = Point-of-Care Test).

Reference Characteristics of the Samples Commercial Test No. of Samples TP FP FN TN Se. Sp. PPV PNV Accuracy Cohen’s Kappa
Imai et al. 2020 [27] 139 samples from 112 COVID-19 patients, 48 negative patients from Saitama Hospital Artron laboratories 187 60 1 79 47 42.2% 97.9% 98.4% 37.3% 57.2% 0.268
Lassaunière et al. 2020 [8] 30 COVID-19 confirmed cases admitted to intensive care unit vs. 10 samples from healthy donors + patients with previous history of coronarivus infection (N = 5), non-coronavirus respiratory infection (N = 45), dengue (N = 9), CMV (N = 2), EBV (N = 10). Artron Laboratories 47 25 0 5 17 83.3% 100% 100% 77.3% 89.4% 0.783
Acro Biotech 20 4 3 1 12 80.0% 80.0% 57.1% 92.3% 80.0% 0.529
AutoBio Diagnostics 62 28 0 2 32 93.3% 100% 100% 94.1% 96.8% 0.935
Dynamiker 62 27 0 3 32 90.0% 100% 100% 91.4% 95.2% 0.903
CTK Biotech 62 27 0 3 32 90.0% 100% 100% 91.4% 95.2% 0.903
Cassaniti et al. 2020 [7] 30 negative healthy controls vs. 30 COVID-19 cases tested with both POCT and RT-PCR VivaDiag 60 19 0 11 30 63.3% 100% 100% 73.2% 81.7% 0.633
50 subjects with suspected COVID-19 sequentially tested with POCT and RT-PCR VivaDiag 50 7 1 31 11 18.4% 91.7% 87.5% 26.2% 36.0% 0.054
Virgilio Paradiso et al. 2020 [10] 191 cases with suspected COVID-19 sequentially tested with POCT and RT-PCR VivaDiag 191 21 13 49 107 30.0% 89.2% 61.8% 68.6% 67.4% 0.215
Perez-Garcia et al. 2020 [11] 45 healthy controls vs. 55 SARS-CoV-2 cases AllTest Biotech 100 26 0 29 45 47.3% 100% 100% 60.8% 71.0% 0.447
Döhla et al. 2020 [22] 39 cases randomly selected among subjects referring to a German COVID-19 screening center Undisclosed manufacturer 49 8 3 14 27 36.4% 88.9% 72.7% 63.2% 65.3% 0.265
Bendavid et al. 2020 [24] 30 orthopedic patients, samples collected before COVID-19 pandemics; 37 PCR confirmed COVID-19 samples Premier Biotech 67 25 0 12 30 67.6% 100% 100% 71.4% 82.1% 0.651
Xiang et al. 2020 [25] 35 healthy individuals; 91 RT-PCR confirmed plasma samples from COVID-19 patients Zhuhai Livzon Diagnostics 126 75 0 16 35 82.4% 100% 100% 68.6% 87.3% 0.723
Adams et al. 2020 [23] Random samples from a pool of 40 SARS-CoV-2 positive blood samples, and 50 SARS-CoV-2 negative blood samples Undisclosed manufacturers 93 18 0 15 60 54.5% 100% 100% 80.0% 54.5% 0.608
128 23 1 14 90 62.2% 98.9% 95.8% 86.5% 62.2% 0.682
93 21 2 12 58 63.6% 96.7% 91.3% 82.0% 63.6% 0.647
98 25 1 13 59 65.8% 98.3% 96.2% 81.9% 65.8% 0.681
91 19 2 12 58 61.3% 96.7% 90.5% 82.9% 61.3% 0.629
91 20 1 11 59 64.5% 98.3% 95.2% 84.3% 64.5% 0.682
93 23 3 10 57 69.7% 95.0% 88.5% 85.1% 69.7% 0.679
92 18 0 14 60 56.3% 100% 100% 81.1% 56.3% 0.626
212 22 4 48 138 31.4% 97.2% 84.6% 74.2% 31.4% 0.340
Liu et al. 2020 [26] 179 consecutive patients, grouped by results of RT-PCR in SARS-CoV-2 positive (No.90), and negative ones (No.89) Undisclosed manufacturer 179 77 13 8 81 90.6% 86.2% 85.6% 91.0% 88.3% 0.765