Skip to main content

Table 1 Estimates with standard error (S.E.) of the parameters from the joint two-component model (S=2) in Eqs. (4) - (6) applied to the daily numbers of postcodes sending test requests to two diagnostic laboratories

From: Methodological challenges to multivariate syndromic surveillance: a case study using Swiss animal health data

Parameter

Laboratory A

 

Laboratory B

 

Estimate

S.E.

 

Estimate

S.E.

Ï•

0.000

0.000

   

δ 1

0.149

0.031

   

γ 1

0.063

0.031

   

δ 2

0.057

0.030

   

γ 2

−0.058

0.030

   

Laboratory specific

     

λ

0.042

0.035

 

0.030

0.036

α t

0.001

0.000

 

-0.001

0.000

α Monday

-0.106

0.107

 

1.293

0.075

α Tuesday

0.165

0.098

 

0.976

0.095

α Wenesday

0.364

0.095

 

0.833

0.093

α Thursday

-0.011

0.110

 

0.882

0.086

α Friday

-0.109

0.107

 

1.068

0.081

ψ

7.388

2.187

 

29.391

23.563

  1. Parameters for Laboratory B in the upper part are empty because they are the same as those for Laboratory A: ϕ is the interactive parameter from the other laboratory; δ 1 to γ 2 are the parameters of sine and cosine function to adjust for the seasonal pattern. In the lower part of the table, parameters are different between Laboratory A and B: λ is the auto-regressive parameter; α t is the time trend; α Monday to α Friday refer to weekday effect and ψ is the overdispersion parameter
Back to article page

BMC Veterinary Research

ISSN: 1746-6148

Contact us