Skip to main content

Table 4 Simulation results based on 1,000 replications

From: Efficiency of two-phase methods with focus on a planned population-based case-control study on air pollution and stroke

  

Scenario 2.1***

Number of second-phase subjects:

300 cases

300 controls

Scenario 2.2

Number of second-phase subjects:

400 cases

200 controls

Scenario 2.3

Number of second- phase subjects:

200 cases

400 controls

 

True OR††

OR*

SD*

Ef**

OR*

SD*

Ef**

OR*

SD*

Ef**

Individual-level information on all subjects †

1.50

1.50

0.12

100

1.49

0.12

100

1.50

0.12

100

 

3.00

3.00

0.11

100

3.02

0.12

100

3.02

0.12

100

Method 3‡

1.50

1.50

0.22

29

1.49

0.23

27

1.52

0.22

30

 

3.00

2.98

0.21

29

2.97

0.21

30

3.05

0.20

35

Method 4‡

1.50

1.50

0.20

37

1.51

0.18

46

1.51

0.21

34

 

3.00

3.01

0.17

44

3.02

0.16

51

3.04

0.17

46

  1. * Geometric mean of the OR estimates and the empiric standard deviation of the ln(OR) estimates.
  2. ** Efficiency of the ln(OR) estimates. eff1 = (var(ln(OR1)) + (ln(true OR1))-ln(OR1)))/(var(ln(ORref)) + (ln(true ORref))-ln(ORref))) where ORref is the estimate in the ideal scenario. Efficiencies calculated when varying the number of second-phase cases and controls. The number of first-phase cases and controls are held fixed at 1,200 cases and 1,200 controls.
  3. *** These results are also presented in table 3 (scenario 2).
  4. † Ideal scenario.
  5. ‡ Methods 3–4 are further described in the Methods section and in Table 1.
  6. †† A confounder with OR = 2 is introduced and a positive bias-effect of 20% for OR = 1.50 and 33% for OR = 3.00