Skip to main content

Table 2 Multilevel regression results

From: Quality of antenatal care in 13 sub-Saharan African countries in the SDG era: evidence from Demographic and Health Surveys

Variable

Model I

Model II

Model III

Model IV

Empty Model

aOR

[95% Crl]

aOR

[95% Crl]

aOR

[95% Crl]

Fixed effects

Individual-level factors

 Age

  15–19

 

0.99

[0.93,1.07]

  

1.00

[0.95,1.07]

  20–24

 

Ref.

Ref.

  

Ref.

Ref.

  25–29

 

1.04

[0.99,1.08]

  

1.04

[0.97,1.08]

  30–34

 

0.98

[0.94,1.03]

  

0.98

[0.94,1.03]

  35–39

 

1.01

[0.96,1.07]

  

1.01

[0.97,1.06]

  40–44

 

1.03

[0.97,1.10]

  

1.05

[0.96,1.11]

  45–49

 

1.06

[0.96,1.17]

  

1.08

[0.97,1.17]

 Education

  No education

 

Ref.

Ref.

  

Ref.

Ref.

  Primary

 

0.65***

[0.63,0.68]

  

0.66***

[0.62,0.69]

  Secondary/higher

 

1.22***

[1.10,1.35]

  

1.23***

[1.10,1.37]

 Wealth

  Poorest

 

0.79***

[0.74,0.84]

  

1.21***

[1.14,1.27]

  Poorer

 

0.91***

[0.86,0.96]

  

1.48***

[1.30,1.49]

  Middle

 

0.99

[0.94,1.06]

  

1.44***

[1.36,1.52]

  Richer

 

1.09**

[1.03,1.15]

  

1.38***

[1.31,1.46]

  Richest

 

Ref.

Ref.

  

Ref.

Ref.

 Marital status

  Married

 

1.66***

[1.59,1.73]

  

2.04***

[1.94,3.05]

  Cohabiting

 

Ref.

Ref.

  

Ref.

Ref.

  Single

 

1.40***

[1.18,1.66]

  

1.81***

[1.29,2.93]

 ANC attendance

  < 4

 

Ref.

Ref.

  

Ref.

Ref.

  ≥ 4

 

2.33***

[2.17,2.50]

  

2.21***

[2.04,2.38]

 Partner’s education

  No education

 

Ref.

Ref.

  

Ref.

Ref.

  Primary

 

0.67***

[0.64,0.70]

  

0.69***

[0.67,0.70]

  Secondary/higher

 

1.11***

[1.06,1.17]

  

1.14***

[1.09,1.18]

 Frequency of listening to radio

  Not at all

 

Ref.

Ref.

  

Ref.

Ref.

  Less than once a week

 

1.07**

[1.06,1.17]

  

1.08**

[1.03,1.13]

  At least once a week

 

1.01

[0.98,1.05]

  

1.05*

[1.02,1.09]

  Almost every day

 

1.59***

[1.35,1.86]

  

1.61***

[1.37,1.89]

 Frequency of watching television

  Not at all

 

Ref.

Ref.

  

Ref.

Ref.

  Less than once a week

 

1.92***

[1.84,2.02]

  

1.82***

[1.74,1.91]

  At least once a week

 

2.55***

[2.43,2.68]

  

2.26***

[2.19,2.39]

  Almost every day

 

3.09***

[2.62,3.65]

  

2.53***

[2.12,2.97]

 Frequency of reading newspaper

  Not at all

 

Ref.

Ref.

  

Ref.

Ref.

  Less than once a week

 

1.01

[0.96,1.07]

  

1.03

[0.97,1.09]

  At least once a week

 

1.12*

[1.02,1.21]

  

1.14**

[1.05,1.26]

  Almost every day

 

1.06

[0.64,1.76]

  

1.08

[0.65,1.79]

Community level factors

 Residence

  Urban

   

Ref.

Ref.

Ref.

Ref.

  Rural

   

0.32***

[0.31,0.33]

0.49***

[0.48,0.80]

 Socio-economic disadvantage

  Tertile 1 (least disadvantaged)

   

Ref.

Ref.

Ref.

Ref.

  Tertile 2

   

0.80***

[0.74,0.86]

1.06

[0.99,1.14]

  Tertile 3 (most disadvantaged)

   

0.83***

[0.76,0.90]

1.06

[0.99,1.14]

 Country

  Burundi

     

Ref.

Ref.

  Malawi

     

1.28

[0.99,3.11]

  Uganda

     

0.55***

[0.33,0.89]

  Chad

     

0.13**

[0.09,0.92]

  Tanzania

     

0.69

[0.40,1.51]

  Zambia

     

0.29***

[0.13,0.40]

  Angola

     

0.82

[0.70,1.65]

  Senegal

     

0.47

[0.27,1.54]

  Mali

     

0.80*

[0.65,0.87]

  Guinea

     

0.56

[0.11,1.33]

  Nigeria

     

0.71**

[0.50,0.87]

  Benin

     

0.60*

[0.22,0.94]

  Cameroon

     

1.19**

[1.04,2.03]

Random-effect

 Community-level

  Variance [95% CrI]

0.29[0.24,0.33]

0.13

[0.11,0.16]

0.22

[0.18,0.25]

0.15

[0.10,0.17]

  VPC % [95% Crl]

36.2[25.9,48.1]

33.0

[24.4,51.5]

34.2

[25.0,44.8]

35.0

[22.8,52.9]

  MOR [95% CrI]

1.67[1.60,1.73]

1.41

[1.37,2.79]

1.57

[1.50,1.60]

1.44

[1.40,1.49]

  Explained variation (%)

Ref.

55.2

 

25.8

 

52.1

 

Model fit statistics

Bayesian DIC

108,124

99,663

 

104,007

 

99,390

 

Sample size

Community-level

1,395

1,395

 

1,395

 

1,395

 

Individual level

79,725

79,725

 

79,725

 

79,725

 
  1. aOR = adjusted Odds Ratio; CrI = Credible Interval; VPC = Variance Partition Coefficient; MOR = Median Odds Ratio
  2. 1 = reference; *p < 0.05, **p < 0.01, ***p < 0.001