Skip to main content

Table 2 Detailed Parameters of Scenarios 1 to 4 in the Simulation Study

From: A comparison of estimators from self-controlled case series, case-crossover design, and sequence symmetry analysis for pharmacoepidemiological studies

Scenarios

β X

β C1

β C2

β C3

β C1C2

β C2C3

α TR

β TR

Et C (Days)a

1

log(1.0)

log(2.0)

log(3.0)

log(1.0)

log(5.0)

log(1.0)

–

–

15

 

log(3.0)

     

–

–

 
 

log(10.0)

     

–

–

 

2

log(1.0)

log(2.0)

log(3.0)

log(5.0)

log(1.0)

log(1.0)

–

–

15

 

log(3.0)

     

–

–

 
 

log(10.0)

     

–

–

 
 

log(3.0)

log(2.0)

log(3.0)

log(0.2)

log(1.0)

log(1.0)

–

–

15

    

log(0.5)

  

–

–

 
    

log(2.0)

  

–

–

 
 

log(3.0)

log(2.0)

log(3.0)

log(5.0)

log(1.0)

log(1.0)

–

–

5

       

–

–

10

       

–

–

20

       

–

–

30

3

log(1.0)

log(2.0)

log(3.0)

log(5.0)

log(1.0)

log(5.0)

–

–

15

 

log(3.0)

     

–

–

 
 

log(10.0)

     

–

–

 
 

log(3.0)

log(2.0)

log(3.0)

log(5.0)

log(1.0)

log(0.2)

–

–

15

      

log(0.5)

–

–

 
      

log(2.0)

–

–

 
 

log(3.0)

log(2.0)

log(3.0)

log(5.0)

log(1.0)

log(5.0)

–

–

5

       

–

–

10

       

–

–

20

       

–

–

30

4

log(3.0)

log(2.0)

log(3.0)

log(1.0)

log(1.0)

log(1.0)

log(1.001)

log(1.0)

15

       

log(1.0)

log(1.001)

 
       

log(1.001)

log(1.001)

 
  1. aLength of the period in which the time-varying covariate C3(t) has an effect