Skip to main content
. 2022 Jan 21;13:811952. doi: 10.3389/fimmu.2022.811952

Table 3.

The accuracy of the model to distinguish severity group 2a or over from severity group 1 in the cases of breakthrough infection.

A. Clinical data alone
Day True Severity Estimated S1 (n) estimate S2a, 2b, 3 (n) Error Rate Accuracy
day 4-7 S1 (n) 3 2 0.40 0.78
S2a, 2b, 3 (n) 0 4 0.00
day 5-8 S1 (n) 3 1 0.25 0.90
S2a, 2b, 3 (n) 0 6 0.00
day 6-9 S1 (n) 1 2 0.33 0.78
S2a, 2b, 3 (n) 0 6 0.00
day 7-10 S1 (n) 1 2 0.67 0.75
S2a, 2b, 3 (n) 0 5 0.00
day 8-11 S1 (n) 1 3 0.75 0.63
S2a, 2b, 3 (n) 1 6 0.14
day 9-12 S1 (n) 1 3 0.75 0.60
S2a, 2b, 3 (n) 1 5 0.17
B. Clinical data + antibody data
Day True Severity Estimated S1 (n) Estimate S2a, 2b, 3 (n) Error Rate (%) Accuracy
day 4-7 S1 (n) 4 1 0.20 0.89
S2a, 2b, 3 (n) 0 4 0.00
day 5-8 S1 (n) 3 1 0.25 0.90
S2a, 2b, 3 (n) 0 6 0.00
day 6-9 S1 (n) 0 3 1.00 0.67
S2a, 2b, 3 (n) 0 6 0.00
day 7-10 S1 (n) 0 3 1.00 0.63
S2a, 2b, 3 (n) 0 5 0.00
day 8-11 S1 (n) 0 4 1.00 0.64
S2a, 2b, 3 (n) 0 7 0.00
day 9-12 S1 (n) 0 4 1.00 0.60
S2a, 2b, 3 (n) 0 6 0.00