Dynamic network measures reveal the impact of cattle markets and alpine summering on the risk of epidemic outbreaks in the Swiss cattle population

Table 4 Comparison of the effectiveness of different surveillance schemes. IQR: interquartile range

Sentinel selection	Detection rate % (median)	IQR	Outbreak size (median)	IQR
outgoing contact chain	83.5	72.4–92.9	50	35–89
in-degree	54.7	32.2–67.2	104	56–173
out-degree	52.6	40.7–66.4	132	49–214
degree prod	56.0	42.4–77.6	114	54–158
betweenness	54.6	31.7–68.7	91	34–159
in-closeness	64.5	38.7–81.2	85	47–151
out-closeness	52.1	35.9–70.3	118	41–208
random	70.3	57.7–85.6	112	57–143

Epidemic outbreaks were simulated using stochastic simulations with 1000 repetitions for each months and each scenario. Surveillance was based on 100 holdings which were either randomly selected from all holdings with an outgoing contact chain larger than 100, from all holdings in the top 5% percentile for one of the static network measures in the respective month, or from all active holdings

ISSN: 1746-6148