Skip to main content
. 2017 Aug 31;12(8):e0183037. doi: 10.1371/journal.pone.0183037

Table 1. Model parameter values used in the stochastic tree scenario analysis to assess the sensitivity (SSe) of the French bovine brucellosis surveillance system.

Parameter Input value [95% confidence interval] Sources
Median number of cows (over 24 months-old) per beef herd (excluding herds with ≤ 5 reproductive animals) 36 Dataa
Median number of cows (over 24 months-old) per dairy herd (excluding herds with ≤ 5 reproductive animals) 52 Dataa
Within-herd prevalence (PU*) Uniform(1/n—0.10)b [4, 3337]
Proportion of brucellosis-related abortions (%) 63 [30–96] [2831]
Proportion of abortions detected (%) 30 [18–42] [38]
Proportion of abortion notifications in beef cattle (%) 21 [12–29] [10]
Proportion of abortion notifications in dairy cattle (%) 41 [31–50] [10]
Probability of infection in purchased bovine max(PU*× PH*, 10−5)c -
Annual proportion of herds (in the high- and low-trade groups) with ≥ 1 introduction investigated (%) 47 [29–66]d [11], Dataa
Annual proportion of introductions investigated per herd (%) 22 [13–30]d [11], Dataa
BPAT sensitivity (%)  95.4 [89.2–100.0] [39]
CF test sensitivity (%)  89.0 [64.4–100.0] [39]
Bacteriology sensitivity (%)  46.1 [28.0–64.2]  [39]
Bulk milk ELISA test sensitivity (%) 97.9 [92.3–100.0] [39]
Allergic skin-test sensitivity (%) 78.3 [65.6–91.0] [39]
Milk ring-test sensitivity (%) 89.5 [65.6–100.0] [39]

a Data from the French national cattle register for 2010–2014

b Uniform distribution ranging from one infected animal per herd (1/n) to 0.10

c PU*, within-herd prevalence; PH*, design prevalence

d Uncertainty interval based on a coefficient of variation of 20%