Description
On episode "account superspreading" (specifically in this section) we interpret proportion_cluster_size as:
estimate the probability of having a cluster of secondary infections caused by a primary case identified by backward tracing of size
$X$ or larger (Endo et al., 2020)
However, for homogeneity with package documentation (manual, vignette), we should interpret it as:
estimate the proportion of new cases originating from a transmission cluster of at least
$X$ cases (Knight et al., 2020). This estimate can give answer to a question like: how many cases are associated with large clusters? which is useful to assess backward contact training efforts.
Current template to update interpretations in working branch:
# Larger clusters are likely to be detected
# through backward tracing in the presence of overdispersion
set.seed(33)
# proportion-cluster size ------------------------------------------------
superspreading::proportion_cluster_size(
R = 0.8, k = 0.1, cluster_size = 5
)
#> R k prop_5
#> 1 0.8 0.1 66.2%
# The proportion of new cases originating from
# a cluster of at least 5 secondary cases from a primary case
# is 66%
# The proportion of all transmission event that were part of
# secondary case clusters (i.e., from the same primary case)
# of at least 5 cases
# is 66%
# compare ----------------------------------------------------------------
superspreading::proportion_cluster_size(
R = 0.9, k = 0.02, cluster_size = c(1, 2, 5, 10, 20)
)
#> R k prop_1 prop_2 prop_5 prop_10 prop_20
#> 1 0.9 0.02 100% 98.1% 92.1% 82.9% 66.6%
superspreading::proportion_cluster_size(
R = 0.9, k = 0.1, cluster_size = c(1, 2, 5, 10, 20)
)
#> R k prop_1 prop_2 prop_5 prop_10 prop_20
#> 1 0.9 0.1 100% 91.9% 70.1% 43.4% 14.9%
superspreading::proportion_cluster_size(
R = 0.9, k = 0.4, cluster_size = c(1, 2, 5, 10, 20)
)
#> R k prop_1 prop_2 prop_5 prop_10 prop_20
#> 1 0.9 0.4 100% 80.9% 34.6% 7.11% 0.146%
# the probability of (finding) new cases originating from
# a cluster of at least 5, 10, 20 secondary cases from a primary case
# increases when there is more overtransmission (lower k value)
# proportion-transmission ------------------------------------------------
superspreading::proportion_transmission(
R = 0.8, k = 0.1, percent_transmission = 0.8
)
#> R k prop_80
#> 1 0.8 0.1 8.47%
# The proportion of cases responsible of 80% of transmission
# is 8.47%
# compare ----------------------------------------------------------------
superspreading::proportion_transmission(
R = 0.9, k = 0.02, percent_transmission = 0.8
)
#> R k prop_80
#> 1 0.9 0.02 2.13%
superspreading::proportion_transmission(
R = 0.9, k = 0.1, percent_transmission = 0.8
)
#> R k prop_80
#> 1 0.9 0.1 8.62%
superspreading::proportion_transmission(
R = 0.9, k = 0.4, percent_transmission = 0.8
)
#> R k prop_80
#> 1 0.9 0.4 19.9%
# The proportion of cases responsible of 80% of transmission
# decreases (more concentrated in a lower proportion of cases)
# when there is more overtransmission (lower k value)Created on 2025-04-01 with reprex v2.1.1