Skip to main content

Table 2 Sensitivity analyses: Number of deaths prevented after one year of policy implementation and their 95% confidence interval (CI) under Scenario 3 (lagged level and slope change) being analyzed with a misspecified model

From: Estimating effects of health policy interventions using interrupted time-series analyses: a simulation study

Year of implementation

Approach

Number of deaths preventeda

95% CI

Number of deaths preventeda

95% CI

Number of deaths preventeda

95% CI

Scenario 3, using Model 1

Effect size

  

−5

−10

− 15

 

True values

1

 

17

 

33

 

5

Estimated

45

(−12, 101)

85

(21, 148)

107

(9, 205)

 

Predicted

−1

(−93, 91)

16

(−77, 108)

32

(− 60, 124)

 

True values

1

 

17

 

33

 

9

Estimated

28

(− 36, 91)

61

(−8, 130)

91

(11, 171)

 

Predicted

1

(−73, 75)

16

(−59, 90)

34

(−41, 109)

 

True values

1

 

17

 

33

 

13

Estimated

−7

(−63, 50)

33

(−24, 91)

74

(14, 134)

 

Predicted

0

(−68, 67)

15

(−55, 85)

32

(−38, 102)

Scenario 3, using Model 2

Effect size

  

−5

−10

−15

 

True values

1

 

17

 

33

 

5

Estimated

13

(−67, 94)

58

(−27, 142)

85

(−20, 191)

 

Predicted

−1

(−93, 91)

16

(−75, 108)

32

(−60, 124)

 

True values

1

 

17

 

33

 

9

Estimated

13

(− 49, 75)

51

(−16, 117)

83

(1, 164)

 

Predicted

0

(−74, 75)

16

(−57, 89)

34

(−41, 109)

 

True values

1

 

17

 

33

 

13

Estimated

8

(−50, 66)

38

(−24, 99)

68

(3, 132)

 

Predicted

0

(−68, 67)

15

(−55, 84)

32

(−38, 102)

  1. CI Confidence interval
  2. aIn the 12 months following the intervention