Fig 1.
Graphic representation of the sequential pooling approach.
The cohort dimension is N = 30, the pool size is 5, and the virus frequency, vf, is 0.10. To begin, the 30 samples are used to create 6 horizontal (H) pools. Since 3 pools’ results are negative (where red icons represent positive subjects), we can exclude 15 subjects. The remaining 15 are used to create 3 vertical (V) pools. As one of these pools yields negative result, only 10 subjects require individual testing. In the end, the total number of tests is equal to 19 (9 pools and 10 validation tests).
Fig 2.
Graphic representation of the Informed sequential pooling approach.
The cohort dimension is N = 30, the pool size is 5, and the vf is 0.10. Panels A and B show two possible scenarios based on available information allowing for the classification of subjects as either “suspect positive” or “suspect negative.” In scenario A, information allows for a concentration of all the positive subjects in the same horizontal pool. Thus after 6H pools, only 5 subjects are left, and vertical pooling is not necessary (globally 11 tests are required). In scenario B, information allows for a concentration of some of the positive samples but not all of them. Among 6 horizontal pools, 4 turn out to be negative. The 10 remaining subjects are then rearranged in 2 vertical pools. Since one pool gives a positive result, 5 validation tests are required (globally 13 tests).
Fig 3.
The curves plotted represent the 1st, 25th, 50th (median), 75th, and 99th percentiles of T/N obtained in the set of 5,000 simulations, for three values of the pool size (s = 3,12,24).
Table 1.
Values of the 1st, 25th, 50th (median), 75th, and 99th percentiles of T/N obtained in the set of 5,000 simulations, for three values of the pool size (s = 3, 12, 24) and some values of the virus frequency (vf = 0.01, 0.10, 0.15, 0.20, 0.25, 0.30).
Fig 4.
Random sequential pooling vs one-step pooling.
The curves plotted represent the 25th, 50th (median), and 75th percentiles of T/N obtained in the set of 5,000 simulations, for 9 values of the pool size (s = 5, 6, 8, 10, 12, 15, 20, 24, 30).
Fig 5.
The curves plotted represent the 50th percentile (median) of T/N obtained in the set of 5,000 simulations, for three values of the pool size (s = 3,12,24). The upper plots were obtained with α = 0.5 and α = 0.6, combined with β = 0.5, 0.6, 0.7, 0.8. The lower plots were obtained with α = 0.7 and α = 0.8, combined with β = 0.5, 0.6, 0.7, 0.8.
Table 2.
Informed sequential pooling, pool size s = 3.
Values of the 1st, 25th, 50th (median), 75th, and 99th percentiles of T/N obtained in the set of 5,000 simulations, for all combinations of α and β in {0.5, 0.6, 0.7, 0.8}, for some values of the virus frequency (vf = 0.01, 0.10, 0.15, 0.20, 0.25, 0.30).
Table 3.
Informed sequential pooling, pool size s = 12.
Values of the 1st, 25th, 50th (median), 75th, and 99th percentiles of T/N obtained in the set of 5,000 simulations, for all combinations of α and β in {0.5, 0.6, 0.7, 0.8}, for some values of the virus frequency (vf = 0.01, 0.10, 0.15, 0.20, 0.25, 0.30).
Table 4.
Informed sequential pooling, pool size s = 24.
Values of the 1st, 25th, 50th (median), 75th, and 99th percentiles of T/N obtained in the set of 5,000 simulations, for all combinations of α and β in {0.5, 0.6, 0.7, 0.8}, for some values of the virus frequency (vf = 0.01, 0.10, 0.15, 0.20, 0.25, 0.30).
Table 5.
Summary of practical indications for pooling creation.