Skip to main content

Table 4 Logistic regression model performance of exploratory univariable models and the multivariable main effects model

From: Who gets vaccinated in a measles-rubella campaign in Nepal?: results from a post-campaign coverage survey

Index

Variable name and categories

Got Campaign Dose % (weighted)

N (unweighted)

Unadjusted Odds Ratio

p-value (Same as univariable logistic regression p-value)

Logistic regression c-stat (or area under the ROC curve)

Weighted % predicted vx’d (0.5 cut off)

Weighted % vx’d correct prediction (i.e., given they got campaign dose, what % had phat > 0.5?)

Weighted % not vx’d correctly predicted (i.e., given they didn’t get campaign dose, what % had phat < 0.5?)

1a

Knew about Campaign

   

< 0.0001

0.90

87.5

99.7

80.1

 Yes

96.5

4223

1477.21

     

 No

1.8

647

Ref

     

2a

Urban/Rural

   

0.0087

0.53

100

100

0

 Rural

85.6

2640

Ref

     

 Urban

79.0

2230

0.63

     

3

Earthquake Area

   

0.3524

0.52

100

100

0

 Yes

82.6

1100

Ref

     

 No

85.2

3770

1.21

     

4

Gender

   

0.8098

0.50

100

100

0

 Boy

84.5

2565

Ref

     

 Girl

84.9

2305

1.03

     

5a

Caregiver Marital Status

   

0.2009

0.50

100

100

0

 Married

84.6

4746

Ref

     

 Not Married

97.5

33

7.12

     

 Unknown

85.4

91

1.07

     

6a

Caregiver Education

   

0.0717

0.55

100

100

0

 Don’t Know or less than Primary

81.2

1423

Ref

     

 Primary

88.3

888

1.74

     

 More than Primary

85.6

2559

1.37

     

7a

Caregiver Religion

   

0.1012

0.53

100

100

0

 Hindu

84.6

4195

Ref

     

 Buddhist

91.3

320

1.91

     

 Muslim

76.2

239

0.58

     

 Not Hindu/Buddhist/ Muslim

83.5

116

0.92

     

8a

Ethnic Group

   

< 0.0001

0.65

100

100

0

 Hill Brahmin

87.9

579

Ref

     

 Hill Chhetri

90.7

1100

1.35

     

 Terai Brahmin Chettri

91.8

175

1.54

     

 Other Terai Caste

64.9

451

0.25

     

 Hill Dalit

88.0

360

1.01

     

 Terai Dalit

79.4

236

0.53

     

 Newar

84.8

295

0.77

     

 Hill Janajati

85.8

899

0.84

     

 Terai Janajati

91.4

408

1.46

     

 Muslim

76.0

236

0.44

     

 Others

88.5

131

1.06

     

9a

Source of Income

   

< 0.0001

0.61

100

100

0

 Professional/Technical/Managerial

79.5

101

Ref

     

 Jobber/Service holder

85.1

892

1.47

     

 Sales and services

85.5

914

1.53

     

 Skilled manual

74.8

787

0.77

     

 Unskilled manual

80.5

306

1.07

     

 Agriculture

89.1

1651

2.12

     

 Others

79.8

219

1.02

     

10

Migration

   

0.4184

0.51

100

100

0

 Yes

82.8

889

Ref

     

 No

84.9

3981

1.17

     

11

Migration Time

   

0.4706

0.51

100

100

0

 Didn’t migrate

84.9

3981

Ref

     

 Migrated < 12 months ago

75.2

49

0.54

     

 Migrated 1–5 years ago

83.2

840

0.88

     

12a

Someone in the Family Works Abroad

   

0.1444

0.53

100

100

0

 Yes

87.4

1322

Ref

     

 No

83.5

3548

0.73

     

13a

# Family Members in HH

   

0.1630

0.53

100

100

0

 < =4 members in HH

86.5

2190

Ref

     

 > 4 members in HH

83.5

2680

0.79

     

14a

# Children 9–59 months in HH

   

0.2207

0.52

100

100

0

 < =1 child in HH aged 9–59 month

85.4

3503

Ref

     

 > 1 child in HH aged 9–59 month

83.2

1367

0.85

     

15a

Child Age Group

   

0.2961

0.51

100

100

0

 Child <=36mo

85.3

2343

Ref

     

 Child >36mo

84.1

2527

0.91

     

16a

Province

   

< 0.0001

0.69

100

100

0

 Province 1

87.9

806

Ref

     

 Province 2

70.0

730

0.32

     

 Province 3

81.1

1163

0.59

     

 Province 4

87.4

495

0.96

     

 Province 5

91.4

616

1.46

     

 Province 6

98.0

740

6.76

     

 Province 7

96.0

320

3.31

     

17a

Ecological Belt

   

0.0002

0.60

100

100

0

 Mountain

96.1

885

Ref

     

 Hill

88.3

1923

0.31

     

 Terai (Plain)

80

2062

0.16

     

Multivariable Main Effects Model

84.7

4870

  

0.96

86.8

99.6

80.8

  1. Sections 1–17 of this table describe results from univariable models; only the final row of the table describes performance of the multivariable model
  2. aDenotes a variable that was retained in the main effects model