Skip to main content

Table 4 Mis-Implementation Predictors among programs overseen by U.S. state health department chronic disease unit staff, 2018 (N=613)

From: Patterns and correlates of mis-implementation in state chronic disease public health practice in the United States

Independent variable

Dependent Variable Category

OR

95% CI

Final Inappropriate Termination among programs overseen by SHD Model

 I have the skills to modify EBIs between one priority population to another

Often/Always

Sometimes

Never/Rarely

3.24

1.25

1 (ref)

1.19, 8.85

0.83, 1.89

 Work unit uses the CDC’s Community Guide in its work

Often/Always

1.56

0.66, 3.72

Sometimes

Never/Rarely

1.73

1 (ref)

1.10, 2.73

 In my agency, the number of layers of authority impede decisions about program continuation or ending

Often/Always

3.23

1.61, 7.40

Sometimes

Never/Rarely

1.25

1 (ref)

0.91, 1.83

 My agency uses quality improvement processes

Often/Always

0.52

0.27, 1.03

Sometimes

Never/Rarely

0.49

1 (ref)

0.34, 0.73

 Leadership in work unit perseveres through the ups and downs of implementing EBIs

Often/Always

0.16

0.07, 0.34

Sometimes

Never/Rarely

0.62

1 (ref)

0.43, 0.89

Final Inappropriate Continuation among programs overseen by SHD Model

 In agency, the number of layers of authority impede decisions about program continuation or ending

Often/Always

Sometimes

Never/Rarely

4.70

1.63

1 (ref)

2.20, 10.04

1.14, 2.34

 Work unit uses economic evaluation in its decision-making about programs

Often/Always

Sometimes

Never/Rarely

0.35

0.62

1 (ref)

0.17, 0.73

0.43, 0.89

 Work unit leaders are competent at managing change

Often/Always

Sometimes

Never/Rarely

0.26

0.57

1 (ref)

0.13, 0.53

0.38, 0.87

 The agency is willing to make changes to enable use of EBIs

Often/Always

Sometimes

Never/Rarely

0.57

0.54

1 (ref)

0.27, 1.23

0.37, 0.78

  1. Boldface indicates statistical significance (p< 0.05)
  2. The survey items for independent variables were asked on a 5-point scale (strongly disagree to strongly agree) and were dichotomized for model specification into strongly agree/agree (1) and neither disagree or agree/disagree/strongly disagree (0). Model fit statistics for model 1 were X2(12)= 83.88, p< 0.001 and for model 2 were X2(8)= 103.11, p< 0.001. OR: odds ratio, CI: confidence interval