Content uploaded by Mustapha Immurana
Author content
All content in this area was uploaded by Mustapha Immurana on Jun 27, 2017
Content may be subject to copyright.
Copyright © 2017 Mustapha Immurana, Arabi. U. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
International Journal of Medicine, 5 (2) (2017) 158-166
International Journal of Medicine
Website: www.sciencepubco.com/index.php/IJM
doi: 10.14419/ijm.v5i2.7830
Research paper
What factors influence the utilisation of all doses of vaccines
with subsequent doses for under-five children in Ghana?
Mustapha Immurana 1*, Arabi, U.2
1 Research Scholar, Department of Economics, Mangalore University, Mangalagangotri – 574119, Karnataka State, India
2 Professor and Research Guide, Department of Economics, Mangalore University, Mangalagangotri – 574119, Karnataka State, India
*Corresponding author E-mail: mustaphaimmurana@gmail.com
Abstract
Pneumonia, Diarrhoea, Measles, Polio and Whooping Cough are major killers of children in the world. However, one of the most effec-
tive ways of preventing these diseases is through utilising completely their respective vaccines which are normally administered in doses.
Therefore given that the 2014 Ghana Demographic and Health Survey (GDHS) report showed a fall in the coverage of the subsequent
doses of the Pentavalent, Pneumococcal, Polio and Rotavirus vaccines for children in Ghana, this study investigated the factors that in-
fluence the Utilisation of all Doses of these vaccines as well as the Measles vaccine. Using Data from the 2014 GDHS and employing the
binary probit model, the study among other findings revealed that, children from the Eastern, Northern and Western regions were found
to be less probable to have received all the doses of the Polio and Pentavalent vaccines. Also unemployed mothers were found to be less
likely to demand for all the doses of the Measles, Pentavalent, Pneumococcal and Rotavirus vaccines for their children. Moreover, rising
age of the household head was associated with falling utilization of all the doses of the Pneumococcal, Pentavalent and Polio vaccines for
children. Furthermore, children with single mothers were found to be less likely to have received all doses of both the Polio and Pentava-
lent vaccines. Also uneducated mothers were found to be less likely to demand for all the doses of the Polio, Pentavalent and Measles
vaccines for their children relative to mothers with secondary education. In addition mothers without health insurance were found to be
less likely to utilize all the doses of the Measles and Polio vaccines for their children. Therefore reinvigoration of regional centeredness
of child health utilization drives, aiding single, unemployed and uneducated women as well as strengthening the free maternal health
insurance registration scheme, could be effective tools in ensuring full utilisation of all doses of these vaccines.
Keywords: Child Health Utilization; Ghana; Immunization.
1. Introduction
Under -five Mortality continues to remain a major concern among
various countries across the globe; hence it is not surprising that
the sustainable development goal (SDG) 3.2 has the target of re-
ducing under-five mortalities to at least 25 per thousand live
births.
According to the World Health Organisation (WHO), in 2015
alone, 5.9 million children under -five died, mostly due to pre-
ventable causes such as pneumonia, diarrhoea, and malaria
(WHO, 2016a). Further among the causes of under-five mortality,
Polio is a viral disease which normally affects children and can
paralyse the entire body in hours by affecting the nervous system
(WHO, 2014). In addition, Whooping cough is also regarded as a
major cause of infant mortality (Loving, 2016), which in 2008
according to the WHO (as cited in Loving, 2016) killed about 195
thousand children. Also, measles is recognised as one of the main
causes of mortalities in young children (WHO, 2016b).
However, all these major killers of children: pneumonia, diar-
rhoea, polio, measles and whooping cough can be prevented using
the pneumococcal, rotavirus, polio, measles and pentavalent vac-
cines respectively.
According to the Expanded Immunization Programme in Ghana,
children are supposed to take three doses each of the pneumococ-
cal and pentavalent vaccines, four doses of the polio vaccine (if
polio 0 which is given at birth is added), two doses of the rotavirus
vaccine and two doses of the measles vaccine. However, the 2014
Ghana Demographic and Health Survey (GDHS) indicates a fall in
the coverage of the subsequent or successive doses of the polio,
pentavalent, pneumococcal and rotavirus vaccines for children in
Ghana (Ghana Statistical Service [GSS], Ghana Health Service
[GHS] & ICF International, 2015) even though children need to
take all these doses in order to be fully protected from these dis-
eases.
Therefore given that under-five mortality rate in Ghana is estimat-
ed to be 60 deaths per thousand live births (GSS, GHS & ICF
International, 2015) with a recent report indicating that diarrhoea
and pneumonia have been significant causes of under-five mortali-
ty during the period 2012-2013 in Ghana (GHS, n.d.), this study
investigated the factors that influence the utilisation of all doses of
vaccines with subsequent doses for under five children in Ghana.
Thus the study brought to light the factors affecting the full utilisa-
tion of all doses of the polio, pentavalent, pneumococcal and rota-
virus vaccines for children in Ghana. This would help in tackling
the falling up-take of the subsequent doses of the pneumococcal
and rotavirus vaccines which could be the reason for the rising
amount of under-five deaths in Ghana attributable to pneumonia
and diarrhoea recently. Also it would help in tackling the factors
that affect the up-take of the subsequent doses of the polio vaccine
which is a great threat to the polio free status of Ghana that has
been achieved since 2008 according to the GHS (2015). Thus In
all, it helped in revealing the factors that influence the up-take of
all the doses of vaccines with subsequent doses individually and
International Journal of Medicine
159
hence would help in achieving the SDG 3.2 target of at least 25
deaths per thousand live births by 2030.
2. Review of literature
On the theoretical literature, since there is a derived demand for
health and individuals are the producers of health according to
Grossman (1999), these vaccines (all their doses) would be uti-
lised for children in order to produce better health. Thus, the vac-
cines become inputs in the health production function. Therefore
the production of better health using these vaccines would en-
hance the performance of children when they grow, in both the
market and non-market sectors.
On the empirical Literature, In Nigeria Cockcroft et al. (2014)
found among other factors that mother’s education was linked
with the likelihood that a child received measles vaccination. Rus-
so et al. (2015) found younger mother’s age, child being the 3 rd
born or beyond and poor parental attitude as some of the factors
influencing incomplete immunization among children in Came-
roon. Further, older maternal and child’s age and lower distance
from the immunization centre among others were found to be
strongly linked with the likelihood of increased influenza vaccina-
tion up take for children in Kenya (Otieno et al., 2014). Also be-
longing in the rich index by mother was found by Lakew et al.
(2015) to be one of the determinants of full immunization of chil-
dren in Ethiopia.
On Ghana, it has been found that children membership of National
Health Insurance Scheme (NHIS), mother’s education and geo-
graphical location determined full immunization of children
(Dwumoh et al. (2014). Also antenatal care was found to increase
the odds of immunization among children (McGlynn, 2012). In
addition, similar studies have been done by Bosu et al. (1997),
Matthews and Diamond (1997) and Duah-Owusu (n.d.).
Moreover, Immurana and Arabi (2016a) among other findings
revealed that, mother’s marital status influenced the utilisation of
all the successive doses of the rotavirus and pneumococcal vac-
cines for children in Ghana. Further, mother’s employment was
found to influence the utilisation of the initial and successive dos-
es of the rotavirus and pneumococcal vaccine. Immurana and Ara-
bi (2016b) among other results revealed that, non-wealthy house-
holds, mothers without health insurance and uneducated mothers
were less probable to demand the 2nd dose of the measles vaccine
for their children in Ghana. Further, Immurana and Arabi (2016c)
also found employed mothers to be more probable to utilise all the
successive doses of both the polio and pentavalent vaccines for
their children.
Among all the works on Ghana above, apart from Immurana and
Arabi (2016a, b & c), none of them investigated the factors that
affect the uptake of individual vaccines. However, this work is
novel because it did not look at the determinants of demand for
each dose of a particular vaccine but rather investigated the factors
that influence the utilisation of all doses of a particular vaccine (as
a single regression equation). Thus if all doses of a vaccine have
been utilised for a child or not rather than if a specific dose has
been utilised for a child or not as done by Immurana and Arabi
(2016a, b & c).
Thus since the main motive is to ensure that all children utilise all
the doses of each vaccine, finding out the factors that affect the
utilisation of all doses of a vaccine using a single regression equa-
tion could be a better approach.
3. Data and methods
This paper used data from a cross-sectional survey (2014 GDHS)
undertaken in early September to mid-December, 2014 by the
GSS, GHS and other partners. During the survey, data were cap-
tured concerning the immunisation status of children as well other
socio-economic, maternal, child and household characteristics.
Thus on immunisation, if all the doses of a particular vaccine has
been given to a child, this study treated it as 1 and if otherwise 0.
Thus concerning the three doses each of the pneumococcal and
pentavalent vaccines, four doses of the polio vaccine, two doses of
the rotavirus vaccine and two doses of the measles vaccine, it
means we have Five (5) separate regressions, each representing
the utilisation of all doses of a vaccine.
Since children are too young to make decisions for themselves, it
is assumed that these vaccines would be utilised for them based on
the mothers expected utility. Thus if a mother’s expected utility
for utilising all the doses of a vaccine is greater than not utilising,
then all the doses would be utilised for the child and if a mother’s
expected utility for utilising all the doses of a vaccine is less than
not utilising, then all the doses would not be utilised for the child.
Therefore given that the various dependent variables (utilisation of
all doses of each vaccine) were binary or dichotomous, the study
adopted the binary probit regression model. Also since Williams
(2012) posits that marginal effects help in given more intuitive
understanding to results from non-liner models such as the probit
model, we reported the results using the average marginal effects
(AMEs). We therefore used a simplified model of estimation as
Dv= ϗE +ΩF +ϦG + μ (1)
Where Dv is the dependent variable, which shows the likelihood
that a mother would utilise all the doses of a vaccine for the child
or not. E is mother’s and partner’s features such as age, education,
employment e.t.c, F represents characteristics of the child such as
sex, birth order and age, whiles G indicates the remaining socio-
economic characteristics such as region, household wealth status,
ethnicity, residence type, e.t.c., with ϗ, Ω, and Ϧ being vector of
parameters of the explanatory variables and μ representing the
error term. In this study all categorical explanatory variables were
treated as dummy variables. Further, wealth status, mother’s mari-
tal status and religion were recoded from how they appeared in the
original data to their current state in this study.
4. Results and discussion
4.1. Descriptive statistics with Pearson chi square
This section as shown in Tables 1 and 2 revealed how the utilisa-
tion of all doses of the respective vaccines were distributed among
the various categorical independent variables. Also, the Pearson
chi square was employed to show the strength of association be-
tween various categorical independent variables and the utilisation
of all doses of the respective vaccines. From Table 1, it can be
seen that there were statistically significant relationships between
utilisation of all doses of the Pentavalent, Measles and Pneumo-
coccal vaccines and region, religion, ethnicity, mother’s educa-
tion, money to seek medical care by mother, and partner’s educa-
tion. Also there were statistically significant relationships between
utilisation of all doses of the Measles and Pneumococcal vaccines,
and residence and household wealth status. Further, the results
also showed statistically significant relationships between utilisa-
tion of all doses of the Pentavalent and Measles vaccines and
mother’s employment.
Also in Table 2, the results showed that there were statistically
significant relationships between utilisation of all doses of the
Rotavirus and Polio vaccines and region, residence, religion, eth-
nicity, mother’s education, mother’s health insurance, money and
distance to seek medical care by mother, partner’s education, sex
of household head and household wealth status. Further, descrip-
tive statistics on variables can also be seen in Tables 1 and 2. Giv-
en that, the Person chi square only shows the association between
utilisation of all doses of each vaccine and individual categorical
independent variables without controlling for other factors that
may affect the utilisation of all doses of each vaccine, this study
went a step further to employ a multivariate probit model by in-
cluding other variables.
160
International Journal of Medicine
Table 1: Descriptive Statistics with Pearson Chi Square Analyses on Utilisation of All Doses of Pentavalent, Measles and Pneumococcal Vaccines
Variable
Pentavalent
(%)
Chi-square
Measles (%)
Chi-square
Pneumococcal
(%)
Chi-square
No
Yes
No
Yes
No
Yes
Region
129.2249***
65.5539***
188.0073***
Western
29.86
70.14
57.14
42.86
54.81
45.19
Central
18.74
81.26
62.24
37.76
53.20
46.80
Greater Accra
20.85
79.15
59.78
40.22
38.13
61.87
Volta
22.96
77.04
58.06
41.94
48.12
51.88
Eastern
25.93
74.07
53.80
46.20
46.78
53.22
Ashanti
20.04
79.96
63.85
36.15
55.58
44.42
Brong Ahafo
15.61
84.39
56.21
43.79
31.78
68.22
Northern
34.09
65.91
71.34
28.66
62.65
37.35
Upper East
17.26
82.74
58.72
41.28
46.05
53.95
Upper West
15.82
84.18
63.85
36.15
56.20
43.80
Residence
0.9529
4.5172**
20.3739***
Urban
22.05
77.95
62.78
37.22
46.21
53.79
Rural
23.17
76.83
59.94
40.06
52.42
47.58
Religion
20.5658***
16.6899***
13.9908***
Christian
22.29
77.71
59.46
40.54
49.01
50.99
Islam
20.81
79.19
64.04
35.96
49.82
50.18
Traditional/Spiritualist/No religion
30.82
69.18
67.43
32.57
58.07
41.93
Ethnicity
114.2734***
44.1331***
79.9677***
Akan
22.64
77.36
58.49
41.51
48.94
51.06
Ga/Dangme
20.16
79.84
56.45
43.55
42.51
57.49
Ewe
21.30
78.70
59.38
40.62
45.48
54.52
Guan
21.88
78.13
60.94
39.06
42.74
57.26
Mole-Dagbani
19.28
80.72
62.89
37.11
49.97
50.03
Grusi
16.24
83.76
56.36
43.64
45.89
54.11
Gurma
39.02
60.98
72.51
27.49
66.85
33.15
Mande
7.14
92.86
62.35
37.65
41.18
58.82
Other
28.57
71.43
58.10
41.90
50.48
49.52
Mother’s Education
14.7840***
22.0538***
46.4330***
Uneducated
25.12
74.88
65.05
34.95
55.45
44.55
Primary
23.81
76.19
60.47
39.53
50.88
49.12
Secondary
20.40
79.60
58.58
41.42
45.73
54.27
Higher
20.28
79.72
55.19
44.81
40.87
59.13
Mother’s Insurance
2.4793
0.6479
11.1862***
Uninsured
24.06
75.94
61.85
38.15
53.37
46.63
Insured
22.14
77.86
60.71
39.29
48.48
51.52
Marital Status
33.3296***
0.3063
2.3480
Single
27.30
72.70
61.58
38.42
51.41
48.59
Married
20.43
79.57
60.81
39.19
49.22
50.78
Mother’s Employment
35.5548***
22.9819***
0.0132
Unemployed
29.26
70.74
67.21
32.79
50.09
49.91
Employed
21.01
78.99
59.49
40.51
49.90
50.10
Permission to seek medical care/help by
mother
4.8612**
0.5920
0.0087
Big Problem
27.35
72.65
62.93
37.07
49.73
50.27
Not a Big Problem
22.39
77.61
60.93
39.07
49.98
50.02
Money to seek medical care/help by mother
7.6875***
7.7867***
18.5770***
Big Problem
24.33
75.67
62.95
37.05
52.97
47.03
Not a Big Problem
21.22
78.78
59.30
40.70
47.16
52.84
Distance to seek medical care/help by mother
9.1481***
2.6779
3.3605*
Big Problem
25.23
74.77
62.63
37.37
51.78
48.22
Not a Big Problem
21.57
78.43
60.34
39.66
49.13
50.87
Partner’s Education
16.4362***
11.8858***
64.3402***
Uneducated
25.38
74.62
64.18
35.82
57.85
42.15
Primary
21.21
78.79
58.95
41.05
50.16
49.84
Secondary
20.19
79.81
59.77
40.23
46.77
53.23
Higher
20.04
79.96
57.33
42.67
40.70
59.30
Sex of Household Head
0.7029
3.2847*
1.0295
Male
22.46
77.54
61.72
38.28
50.33
49.67
Female
23.57
76.43
58.92
41.08
48.71
51.29
Sex of Child
0.1592
0.0719
0.5532
Male
22.94
77.06
61.24
38.76
50.44
49.56
Female
22.49
77.51
60.89
39.11
49.43
50.57
Wealth Status
0.5212
3.8210*
22.6691***
Non-Rich
22.97
77.03
61.85
38.15
51.92
48.08
Rich
22.06
77.94
58.99
41.01
44.73
55.27
Source: Author’s computation from the 2014 GDHS. Notes: 1. *,**, and *** showing significant difference at 10%, 5% and 1% respectively within the
predictors and utilisation of all doses of Pentavalent, Measles and Pneumococcal vaccines among children in Ghana. 2. Traditional in this study means
traditional/ spiritualist/ no religion.
International Journal of Medicine
161
Table 2: Descriptive Statistics with Pearson Chi Square Analyses on Utilisation of All Doses of Rotavirus and Polio Vaccines
Variable
Rotavirus (%)
Chi-square
Polio (%)
Chi-square
No
Yes
No
Yes
Region
171.0019***
218.1898***
Western
48.99
51.01
46.26
53.74
Central
48.66
51.34
43.88
56.12
Greater Accra
32.18
67.82
28.86
71.14
Volta
44.81
55.19
43.23
56.77
Eastern
44.42
55.58
51.17
48.83
Ashanti
50.74
49.26
42.22
57.78
Brong Ahafo
27.45
72.55
42.20
57.80
Northern
55.48
44.52
59.59
40.41
Upper East
42.11
57.89
29.21
70.79
Upper West
53.09
46.91
30.74
69.26
Residence
25.6761***
63.8702***
Urban
41.06
58.94
36.51
63.49
Rural
48.01
51.99
47.32
52.68
Religion
11.2040***
43.5330***
Christian
44.27
55.73
41.96
58.04
Islam
45.74
54.26
40.73
59.27
Traditional/Spiritualist/No religion
52.32
47.68
57.20
42.80
Ethnicity
54.8561***
163.5444***
Akan
43.72
56.28
43.40
56.60
Ga/Dangme
40.65
59.35
36.14
63.86
Ewe
41.73
58.27
38.81
61.19
Guan
38.71
61.29
48.44
51.56
Mole-Dagbani
45.70
54.30
36.99
63.01
Grusi
43.04
56.96
36.44
63.56
Gurma
59.45
40.55
65.76
34.24
Mande
41.18
58.82
34.12
65.88
Other
43.81
56.19
56.19
43.81
Mother’s Education
49.0614***
82.3075***
Uneducated
50.92
49.08
49.56
50.44
Primary
45.90
54.10
45.19
54.81
Secondary
41.12
58.88
37.96
62.04
Higher
35.10
64.90
26.89
73.11
Mother’s Insurance
12.7702***
40.5129***
Uninsured
48.89
51.11
49.38
50.62
Insured
43.68
56.32
40.23
59.77
Marital Status
0.0520
37.5859***
Single
45.47
54.53
48.74
51.26
Married
45.15
54.85
40.14
59.86
Mother’s Employment
0.0001
4.4576**
Unemployed
45.22
54.78
45.77
54.23
Employed
45.24
54.76
42.32
57.68
Permission to seek medical care/help by mother
0.2893
2.1705
Big Problem
46.61
53.39
46.67
53.33
Not a Big Problem
45.17
54.83
42.77
57.23
Money to seek medical care/help by mother
17.6247***
25.1693***
Big Problem
48.18
51.82
46.45
53.55
Not a Big Problem
42.54
57.46
39.81
60.19
Distance to seek medical care/help by mother
6.6776***
27.2277***
Big Problem
47.82
52.18
48.11
51.89
Not a Big Problem
44.09
55.91
40.68
59.32
Partner’s Education
44.2999***
75.8325***
Uneducated
51.80
48.20
50.19
49.81
Primary
46.77
53.23
42.74
57.26
Secondary
42.58
57.42
39.54
60.46
Higher
37.89
62.11
30.17
69.83
Sex of Household Head
5.4837**
3.1022*
Male
46.12
53.88
42.38
57.62
Female
42.40
57.60
45.13
54.87
Sex of Child
0.1870
0.5360
Male
45.54
54.46
43.49
56.51
Female
44.96
55.04
42.52
57.48
Wealth Status
27.8295***
80.6605***
Non-Rich
47.43
52.57
46.68
53.32
Rich
39.48
60.52
33.36
66.64
Source: Author’s computation from the 2014 GDHS. Notes: 1. *,**, and *** showing significant difference at 10%, 5% and 1% respectively within the
predictors and utilisation of all doses of Rotavirus and Polio vaccines among children in Ghana. 2. Traditional in this study means traditional/ spiritualist/
no religion.
162
International Journal of Medicine
5. Regression results
Table 3: Probit Regressions on Determinants of Full Utilisation of Vaccines with Subsequent Dose (S)
Dependent Variable
Pentavalent
Polio
Measles
Rotavirus
Pneumococcal
Independent Variable
AME
AME
AME
AME
AME
Region(Ref: Upper West)
Western
-.1107938***
-.1603083***
.0619564*
.0447919
.0339517
Central
.003308
-.1344341***
.0338538
.053704
.05755
Greater Accra
-.0539132
-.0732413*
.0572557
.2016023***
.1734253***
Volta
-.0574622
-.1534051***
.1072518***
.0963826**
.1030397**
Eastern
-.0810036**
-.2244345***
.1212638***
.1052747***
.112036***
Ashanti
-.0241715
-.1535217***
.0164905
.0247428
.0127424
Brong Ahafo
.0212145
-.1091178***
.0934987***
.2589328***
.2543019***
Northern
-.1154858***
-.2080313***
-.0086175
.0310371
-.0007991
Upper East
-.0347185
.0053446
.0580195**
.1098284***
.0914145***
Residence
(Ref:Rural)
Urban
-.0109281
.0304979*
-.0775844***
.0131332
.006396
Religion (Ref: Tradition-
al/Spiritualist/No religion
Christian
.0048738
.0329224
.0092836
-.0103739
-.0052963
Islam
.0161784
.0562507*
.0022238
-.0162111
.0025328
Ethnicity(Ref: other)
Akan
.0269507
.1454865***
-.0647428
.0359911
.0516615
Ga/Dangme
.0826768
.2033152***
-.0575017
.0553396
.0960034*
Ewe
.0667151
.2089311***
-.0979571*
.0449239
.0651385
Guan
.0616773
.0906604
-.053489
.1017865
.1256384**
Mole-Dagbani
.07546*
.1750327***
-.0278927
.0612048
.0843141*
Grusi
.074672
.137086**
-.0067375
.0665237
.1058927*
Gurma
-.0426816
.0605423
-.1128341**
-.0550015
-.0535215
Mande
.1740878***
.2071394***
-.0418908
.047641
.113462
Mother’s Education (Ref: Unedu-
cated)
Primary
.0007086
.0184027
.0321172*
.0009508
-.0191873
Secondary
.0431675**
.0578301***
.05701***
.028404
.0194398
Higher
.0365684
.0743072*
.064055
.0250654
.0059191
Mother’s Insurance(Ref: Unin-
sured)
Insured
.0065101
.0411707***
.0332583**
.0067697
.0005121
Marital Status (Ref: Single Moth-
ers)
Married
.0402833***
.0379729**
.0125784
.0252578
.0352167**
Mother’s Employment (Ref: Un-
employed)
Employed
.0516401***
.0230205
.034099**
.0541613***
.0481056***
Permission to Seek Medical Care
by Mother (Ref: Not a big prob-
lem)
Big Problem
-.013339
.003491
-.0165122
.0017693
.0212443
Money to Seek Medical Care by
Mother (Ref: Not a big problem)
Big Problem
.0052668
.0208162
-.0171736
-.004508
-.0109479
Distance to Seek Medical Care by
Mother (Ref: Not a big problem)
Big Problem
-.0163249
-.0127886
.0089444
.0180564
.0273362*
Partner’s Education (Ref: Unedu-
cated)
Primary
.0079032
.028514
.0080398
.0031536
.028717
Secondary
.0191173
.0430786**
-.0067679
.0113096
.0386271*
Higher
.0071381
.0639698**
.0088649
.024883
.0734837**
Sex of Household head (Ref: Fe-
male Head)
Male head
.0129475
.0251423
-.0070846
-.0371018**
-.0122194
Sex of Child (Ref: Female)
Male Child
-.0039511
-.0105276
-.00013
-.0020021
-.0061394
Wealth (Ref: Non-Rich)
Rich
-.0348546*
.0296453
.0413539*
.0133997
.0010351
Mother’s Age
.0048729***
.0087163***
.0013995
.0010194
.0028369*
Age of Household Head
-.0012891**
-.001808***
-.0001042
-.0007321
-.001878***
Birth Order of Child
-.0133703***
-.0246294***
-.0071245
-.0054823
-.008168
Childs Age
.0345582***
-.0044372
.1227777***
-.1249063***
-.1141308***
N=5126
Prob > chi2 =
0.0000
N=5140
Prob > chi2 =
0.0000
N=5128
Prob > chi2 =
0.0000
N=5066
Prob > chi2 = 0.0000
N= 5073
Prob > chi2 = 0.0000
Source: Author’s computation from the 2014 GDHS. Notes: 1. ***P-value<.01, **P-value<.05, *P-value<.1. 2. Traditional in this study means traditional/
spiritualist/ no religion.
On the determinants of utilisation of all the doses of the Pentava-
lent vaccine, the results showed that on average, children from the
Western, Northern and Eastern regions were 11%, 12% and 8%
respectively less probable to have received all doses of the penta-
valent vaccine relative to those in the Upper west region. These
send signals on reinvigoration of regional centeredness concerning
International Journal of Medicine
163
the issue of full utilisation of vaccines for children in Ghana. The
result conflicts that of Logullo et al. (2008) who revealed region
not to be linked with adequate utilisation of measles vaccine in
Sao Paulo City (Brazil) but concurs with that of Matthews and
Diamond (1997) who found region to be essential in determining
immunization of children in Ghana and the findings of Immurana
and Arabi (2016c) on the Western, Northern and Eastern regions
with regards to the demand for the 3rd dose of the Pentavalent
vaccine for children in Ghana. On ethnicity, the results showed
that on the average, children with Mande and Mole-Dagbani eth-
nic backgrounds were 17% and 8% respectively more probable to
have been offered all doses of the Pentavalent vaccine relative to
those from the other ethnic groups. The result on Mande ethnicity
concurs with Immurana and Arabi (2016 c) on the demand for the
3rd dose of the Pentavalent vaccine for children in Ghana but con-
trary in the case of Mole Dagbani since it was found to be insig-
nificant by Immurana and Arabi (2016c) with regards to demand
for the 3rd dose of the Pentavalent vaccine for children in Ghana.
Further, mothers with secondary education were found to be 4%
more likely to have demanded all doses of the Pentavalent vaccine
for their children relative to their uneducated counterparts. This is
not startling since educated mothers can better appreciate the im-
portance of utilising all doses of vaccines for their children rela-
tive to their uneducated counterparts. The finding is similar to that
of Ibnouf et al. (2007) in Sudan, Dwumoh et al. (2014) and Mat-
thews and Diamond (1997) in Ghana and that of Immurana and
Arabi (2016 c) on the demand for the 3rd dose of the Pentavalent
vaccine for children in Ghana.
Further, employed and married mothers were on the average 5%
and 4% more probable to demand all the three doses of the Penta-
valent vaccine for their children relative to the unemployed and
single mothers respectively. The finding on mother’s employment
is similar to that of Immurana and Arabi (2016 c) on the demand
for the 1st, 2nd and 3rd doses of the Pentavalent vaccine for children
in Ghana. Further, the finding on married mothers is similar to that
of Immurana and Arabi (2016 c) on the demand for the 3rd dose of
the pentavalent vaccine for children in Ghana. Thus even though
these vaccines are administered for free, employed and married
mothers could get monetary support from their jobs and husbands
respectively and hence would be more capable to cater for any
indirect costs of immunisation such as transportation cost. Also
married women could be boosted in visiting immunisation centres
in order to utilise these vaccines for their children by enjoying
company from their husbands. Surprisingly, wealthy households’
children were 3% less probable to have been offered all the doses
of the Pentavalent vaccine relative to their non-wealthy house-
holds counterparts. This is similar to the finding of Immurana and
Arabi (2016 c) with regard to the demand for the 3rd dose of the
Pentavalent vaccine for children in Ghana.
Also averagely as the ages of the child and mother increased by a
year, the likelihood that a child received all doses of the Pentava-
lent vaccine increased by 3% and .5% respectively. The result on
child’s age concurs with that of Ibnouf et al. (2007) in Sudan. The
result on mother’s age could be due to more experience on the part
of older mothers than their younger counterparts and therefore
would better know how important utilising all the doses of the
Pentavalent vaccine is to the survival of children. Also the result
on child’s age is not startling since subsequent doses of vaccines
are offered if only the child has attained a certain age and hence
leading to older children being more likely to be offered all the
doses of the Pentavalent vaccine as compared to their younger
counterparts who may not have attained the required age to be
offered all the doses especially the subsequent ones. Conversely,
on average as the age of the household head and birth order of the
child increased by a year, the likelihood that a child received all
doses of the pentavalent vaccine fell by .1% and 1% respectively.
The finding on older household heads could be due to the seem-
ingly negative perceptions some older people have about modern
medicine. The result on birth order could be that caregivers de-
tected some side effects which might accompany certain vaccines
when they utilised for their previous or older children and hence
would not be willing to utilise for their subsequent births (high
birth ranked children). The results on ages of the mother, child and
household head as well as birth order of the child are all similar to
the findings of Immurana and Arabi (2016 c) with regards to the
demand for the 2nd and 3rd doses of the Pentavalent vaccine for
children in Ghana.
On the determinants of utilising of all the four doses of the Polio
vaccine, children from the Western, Greater Accra, Central, Volta,
Eastern, Brong Ahafo, Ashanti and Northern Regions were respec-
tively on average 16%, 7%, 13%, 15%, 22%, 11%, 15% and 21%
less probable to have been offered all the doses of the polio vac-
cine relative to those in the Upper West region. The results on the
less likelihood of utilising all doses of the polio vaccine for chil-
dren from the Western, Northern and Eastern regions are similar
to those revealed under the utilisation of all the three doses of the
pentavalent vaccine for children.
Also children from Islamic background and urban areas were 6%
and 3% more likely to have received all the four doses of the Polio
vaccine as compared to those from traditional/spiritualist/no reli-
gion background and rural areas respectively. The finding on ur-
ban children is similar and contrary to that of Immurana and Arabi
(2016c) in the case of demand for the 1st dose (Polio 0) and 2nd
dose (Polio1) of the Polio vaccine for children in Ghana respec-
tively.
On ethnicity, averagely, children from Ga/Dangme, Akan, Ewe,
Grusi, Mole-Dagbani, and Mande ethnicities were 20%, 15%,
21%, 14%, 18%, and 21% respectively more likely to have been
offered all the doses of the Polio vaccine relative to their counter-
parts from the other ethnic groups.
In addition on average, mothers with higher and secondary educa-
tion, married mothers and mothers who had health insurance were
7%, 6%, 4% and 4% respectively more probable to have demand-
ed all the four doses of the polio vaccine for their children as
compared to the uneducated mothers, single mothers and mothers
without health insurance respectively. The finding on marital sta-
tus contradicts that of Logullo et al. (2008) who revealed marital
status not to be linked with adequate utilisation of measles vaccine
in Sao Paulo City (Brazil) but similar to that of Immurana and
Arabi (2016c) in the case of demand for the 1st dose (Polio 0) and
4th dose (Polio3) of the polio vaccine for children in Ghana respec-
tively.
Moreover, the result on health insurance which is similar to that of
Immurana and Arabi (2016c) in the case of demand for the 1st
dose (Polio 0) of the polio vaccine for children in Ghana was ex-
pected. This is because health insurance gives easy access to
health centres and other health facilities where immunization
drives as well as education on the essence of fully utilising vac-
cines is to child survival are normally embarked upon and hence,
making mothers with health insurance more likely to demand all
the doses of the polio vaccine for their children as compared to
their counterparts without health insurance.
Also, children whose mothers had partners with higher and sec-
ondary levels of education were 6% and 4% respectively more
probable to have received all the four doses of the Polio vaccine
relative to their counterparts whose mothers had uneducated part-
ners. These are similar to that of Immurana and Arabi (2016c) in
the case of demand for the 1st dose (Polio 0) of the polio vaccine
for children in Ghana. The findings are not surprising since the
educated can better comprehend the essence of vaccines to child
survival than the uneducated.
Also on average, a yearly increase in mother’s age was found to
increase the probability that a child was offered all the doses of the
polio vaccine by.9%. On the contrary, on average as the age of the
household head and birth of order of the child increased, the like-
lihood that a child received all doses of the Polio vaccine fell by
.2% and 2% respectively.
On the utilisation of both doses of the Measles vaccine for chil-
dren in Ghana, on average, children from the Volta, Western,
Eastern, Upper East and Brong Ahafo regions were respectively
164
International Journal of Medicine
11%, 6%, 12%, 6% and 9% more probable to have been offered
all doses of the Measles vaccine relative to children in the Upper
West region. The findings on Volta, Eastern, Upper East and
Ashanti regions conflict with those of Immurana and Arabi
(2016b) who found these regions to be insignificant in the case of
demand for the 1st dose of the Measles vaccine for children in
Ghana. However, the findings on Volta, Eastern and Brong Ahafo
regions are similar to that of Immurana and Arabi (2016b) in the
case of demand for the 2nd dose of the Measles vaccine for chil-
dren in Ghana.
Concerning residence, urban children were 8% less likely to have
received both doses of the Measles vaccine relative to rural chil-
dren which is similar to that of Immurana and Arabi (2016b) in the
case of demand for the 1st and 2nd doses of the Measles vaccine for
children in Ghana. This can be attributed to probably the over
concentration on rural areas to the neglect of urban areas in child
health utilisation drives given that rural areas are normally seen to
be more vulnerable. With regards to ethnicity, on average, chil-
dren from Gurma and Ewe households were 11% and 10% respec-
tively less likely to have been offered both doses of the Measles
vaccine relative to those from the other ethnic groups. The results
on ethnicity concur with those of Immurana and Arabi (2016b) in
the case of demand for the 2nd dose of the Measles vaccine but
contrary in the case of demand for the 1st dose of the Measles
vaccine for children from the Ewe ethnicity in Ghana who were
found to be more likely to have received by Immurana and Arabi
(2016b).
On maternal education, mothers with secondary and primary lev-
els of education were on the average 6% and 3% more likely to
demand both doses of the measles vaccine for their children as
compared to the uneducated mothers. This is in line with the find-
ings of Cockcroft et al. (2014) and Cockcroft et al. (2009) who
found mother’s education to influence the probability that a child
received the measles vaccine in Nigeria and Pakistan respectively.
Also employed and insured mothers were both 3% more likely to
demand both doses of the measles vaccine for their children as
compared with the unemployed and uninsured mothers respective-
ly. The findings on mothers health insurance conflicts with those
of Immurana and Arabi (2016b) in the case of demand for the 1st
dose of the measles vaccine because it was insignificant but simi-
lar in the case of demand for the 2nd dose of the measles vaccine.
Also the result on mother’s employment is similar to that of Im-
murana and Arabi (2016b) in the case of demand for the 1st dose
and 2nd dose of the measles vaccine.
Moreover, wealthy households’ children were 4% more probable
to have been offered both doses of the measles vaccine relative to
non-wealthy households’ children. This concurs with the findings
of Cockcroft et al. (2009) in Pakistan and that of Immurana and
Arabi (2016b) in the case of demand for the 2nd dose of the mea-
sles vaccine for children in Ghana.
On average, a yearly increase in a child’s age was found to in-
crease the probability of demanding both doses of the measles
vaccine by 12%. This is in line with the findings of Otieno et al.
(2014) who revealed older children to be probable to have re-
ceived influenza vaccination in Kenya and that of Immurana and
Arabi (2016b) in the case of demand for the 1st dose and 2nd dose
of the measles vaccine for children in Ghana.
On the utilisation of both doses of the rotavirus vaccine, on aver-
age, mothers from the Volta, Greater Accra, Eastern, Upper East
and Brong Ahafo regions were respectively 10%, 20%, 11%, 11%
and 26% more likely to have utilised both doses of the rotavirus
vaccine for their children relative to their counterparts in the Up-
per West region. Also on average, employed mothers were found
to be 5% more likely to demand both doses of the rotavirus vac-
cine for their children as compared with the unemployed mothers.
In addition, children with male household heads were on average
4% less likely to have been offered both doses of the rotavirus
vaccine relative to children with female household heads. The
results on region, mother’s employment and male household heads
are similar to the findings of Immurana and Arabi (2016a) on the
utilisation of the 1st dose and 2nd dose of the rotavirus vaccines for
children in Ghana.
Further, on average, a yearly increase in a child’s age decreased
the probability of demanding both doses of the rotavirus vaccine
by 12%. The finding on child’s age could be that older children
were too old to be offered all doses of the rotavirus vaccine at the
recommended ages (6 weeks for the 1st dose and 14 weeks for the
2nd dose) since the vaccine was introduced in 2012 and the study
used data on children born five years preceding the survey
year/month of 2014 /September. This result is similar to that of
Immurana and Arabi (2016a) with regard to the utilisation of the
1st dose and 2nd dose of the rotavirus vaccines for children in
Ghana.
Moreover, results on the utilisation of all the three doses of the
pneumococcal vaccine showed that on average, children from the
Volta, Greater Accra, Eastern, Upper East and Brong Ahafo re-
gions were respectively 10%, 17%, 11%, 9% and 25% more likely
to have received all doses of the pneumococcal vaccine relative to
those in the Upper West region. The results on region are similar
to the findings of Immurana and Arabi (2016a) on the utilisation
of the 1st dose, 2nd dose and 3rd dose of the pneumococcal vaccine
for children in Ghana. On ethnicity, on average, children with
Guan, Ga/Dangme, Grusi and Mole-Dagbani backgrounds were
13%, 10%, 11% and 8% respectively more probable to have been
offered all doses of the pneumococcal vaccine relative to children
from the other ethnic groups. These concur with the findings of
Immurana and Arabi (2016a) on the utilisation of the 2nd dose and
3rd dose of the pneumococcal vaccine for children in Ghana.
Further, employed and married mothers were 5% and 4% more
likely to demand all the doses of the pneumococcal vaccine for
their children relative to unemployed and single mothers respec-
tively. These findings are similar to the findings of Immurana and
Arabi (2016a) on the utilisation of the 1st dose, 2nd dose and 3rd
dose of the pneumococcal vaccine for children with employed
mothers and on the utilisation of the 2nd dose and 3rd dose of the
pneumococcal vaccine for children with married mothers in Gha-
na.
Further, mother’s with big problems concerning distance in seek-
ing medical care for themselves were found to be 3% more likely
to demand all doses of the pneumococcal vaccine for their chil-
dren as compared to mothers who did not see distance as a big
problem in seeking medical help. This contradicts the findings of
Immurana and Arabi (2016a) who found distance not to be a sig-
nificant determinant on the utilisation of the 1st dose, 2nd dose and
3rd dose of the pneumococcal vaccine for children. The reason
could be that since mothers have challenges with regards to seek-
ing care, they resort to more preventive measures for themselves
and their children.
In line with the expectation of the study, children whose mothers
had Partners with Higher and Secondary education were 7% and
4% respectively more probable to have received all the three doses
of the pneumococcal vaccine relative to children whose mothers
had uneducated partners which are in line with the results of Im-
murana and Arabi (2016a) on the utilisation of the 3rd dose of the
Pneumococcal vaccine for children in Ghana. Also on average, a
yearly increase in mother’s age was found to increase the proba-
bility of children receiving all doses of the Pneumococcal vaccine
by .3%. This concurs with that of Immurana and Arabi (2016a) on
the utilisation of the 3rd dose of the Pneumococcal vaccine for
children in Ghana
Last but not the least, averagely, as the ages of the child and that
of the household head increased by a year, the probability that
children received all doses of the pneumococcal vaccine fell by
11% and .2% respectively. Similar results were obtained by Im-
murana and Arabi (2016a) on the utilisation of the 3rd dose of the
pneumococcal vaccine for children in Ghana. The finding on
child’s age could be that older children were too old to be offered
all doses of the pneumococcal vaccine at the recommended ages
International Journal of Medicine
165
(6 weeks for the 1st dose, 10 weeks for the 2nd dose and 14 weeks
for the 3rd dose) since the vaccine was introduced in 2012 and the
study used data on children born five years preceding the survey
year/month of 2014 /September. The findings are further similar to
that of Immurana and Arabi (2016a) with regard to the utilisation
of the 1st dose and 2nd dose of the rotavirus vaccines for children
in Ghana.
6. Conclusion
From the findings, the study can conclude that residence, region,
ethnicity, mother’s health insurance, education, mother’s em-
ployment, partner’s education, religion, household wealth, birth
order, marital status of mother, child’s age, distance to seek medi-
cal care by mother and sex of household head are significant fac-
tors when it comes to determining the utilisation of all doses of
vaccines. It must however be stressed that, it does not necessarily
mean all these factors individually had significant effects on the
utilisation of all doses of the individual vaccines the study consid-
ered. Therefore from the findings, the study would recommend the
following:
1) Reinvigoration of the regional child health utilization drives
should be highly prioritized. This is because children from
the Eastern, Northern and Western regions were found to be
less probable to have utilized all the doses of the polio and
pentavalent vaccines whiles children in all the remaining
other regions except the Upper East region, were also re-
vealed to be less probable to have been offered all the doses
of the polio vaccine.
2) Affirmative action and initiatives that would enhance job
opportunities for women should be encouraged since em-
ployed mothers were more likely to demand for all the dos-
es of the measles, pentavalent, pneumococcal and rotavirus
vaccines for their children as compared to their unemployed
counterparts. As a short term strategy, unemployed women
can be aided with regards to the immunization of their chil-
dren.
3) Educating mothers and household heads through the media
and any other means available in order for them to under-
stand the need for all children irrespective of their birth or-
der or nature to be immunized should be embarked upon.
This is because rising birth order was revealed to be associ-
ated with the less likelihood that children received all doses
of the pentavalent and polio vaccines. Also children with
male household heads were less likely to receive all the
doses of the rotavirus vaccine and rising age of the house-
hold head was associated with falling utilization of all the
doses of the pneumococcal, pentavalent and polio vaccines
for children.
4) Special attention should also be given to vulnerable women
such as less educated or uneducated mothers and single
mothers with regards to the health utilization of their chil-
dren. This is because children with single mothers were
found to be less likely to have received all doses of both the
polio and pentavalent vaccines. Also uneducated mothers
were found to be less likely to demand for all the doses of
the polio, pentavalent and measles vaccines for their chil-
dren relative to mothers with secondary education. Moreo-
ver, enhancing girl child education could be an effective
long term strategy.
5) Also further strengthening the free maternal health insur-
ance registration under the National health insurance
scheme should be encouraged given that insured mothers
were more likely to utilize all the doses of the measles and
polio vaccines for their children relative to the uninsured
mothers.
6) Also in our attempt to help vulnerable rural dwellers, care
must be taken not to ignore urban dwellers with regard to
child health utilization drives since urban children were less
likely to have received both doses of the measles vaccine.
7) Helping younger mothers could be relevant since rising ma-
ternal age was found to be linked with the utilization of all
doses of the polio, pentavalent and pneumococcal vaccines
for children in Ghana.
Acknowledgement
We are thankful to the DHS Program for the data
Conflicts of interest
None declared
References
[1] Bosu KW, Ahelegbe D, Edum-Fotwe E, Bainson AK & Turkson
KP (1997) Factors influencing attendance to immunization sessions
for children in a rural district of Ghana. Acta Tropica 68, 259–267,
Elsevier. http://dx.doi.org/10.1016/S0001-706X (97) 00094-6.
[2] Cockcroft A, Usman UM, Nyamucherera FO, Emory H, Duke B,
Umar AN & Andersson N (2014) Why children are not vaccinated
against measles: a cross-sectional study in two Nigerian States. Ar-
chives of Public Health, 72(48). Retrieved from
http://www.archpublichealth.com/content/72/1/48.
https://doi.org/10.1186/2049-3258-72-48.
[3] Duah-Owusu M (n.d.) Social Determinants and Immunisation in
Ghana: Is there an Association? Unpublished MPhil Thesis, Uni-
versity of Bergen research centre for health promotion and devel-
opment, faculty of psychology.
[4] Dwumoh D, Essuman EE & Afagbedzi KS (2014) Determinant of
factors associated with child health outcomes and service utilization
in Ghana: multiple indicator cluster survey conducted in 2011. Ar-
chives of Public Health, 72(42). Retrieved from
http://www.archpublichealth.com/content/72/1/42.
https://doi.org/10.1186/2049-3258-72-42.
[5] Ghana Health Service (2015) Ghana Health Service 2014 Annual
Report.
[6] Ghana Health Service (n.d.) 2014 Family Health Annual Report.
[7] Ghana Statistical Service (GSS), Ghana Health Service (GHS) &
ICF International. (2015). Ghana Demographic and Health Survey
2014. Rockville, Maryland, USA: GSS, GHS, and ICF International.
[8] Grossman M (1999) the Human Capital Model and the Demand for
Health. NBER Working Paper No. 7078.
https://doi.org/10.3386/w7078.
[9] Immurana M & Arabi U (2016a) Determinants of demand for sub-
sequent doses of pneumococcal and rotavirus vaccines for children
less than five years of age in Ghana, International Journal of Health,
4 (2), 120-127. doi: 10.14419/ijh.v4i2.6114
[10] Immurana M & Arabi U (2016b) Demand for measles and yellow
fever vaccines for children in Ghana: are socio-economic, demo-
graphic and Geographic factors relevant?, International Journal of
Accounting and Economics Studies, 4 (2),136-141.
https://doi.org/10.14419/ijaes.v4i2.6526.
[11] Immurana, M & Arabi U (2016c) Socio-economic determinants of
successive polio and pentavalent vaccines utilisation among under-
five children in Ghana (Unpublished Manuscript).
[12] Lakew Y, Bekele A, & Biadgilign S (2015) Factors influencing full
immunization coverage among 12–23 months of age children in
Ethiopia: evidence from the national demographic and health sur-
vey in 2011. BMC Public Health, 15(728).
https://doi.org/10.1186/s12889-015-2078-6.
[13] Loving S (2016) Vaccine Knowledge Project, Authoritative Infor-
mation for All. Oxford Vaccine group, University of Oxford. 2016.
Accessed from: www.ovg.ox.ac.uk/pertusis-whooping-cough on
4th April, 2016
[14] Matthews Z & Diamond I (1997) Child Immunisation in Ghana:
The Effects of Family, Location and Social Disparity, Journal of
Biosocial Science, 29(03), 327 – 343. DOI: null, Published online:
04 April 2001.
[15] McGlynn KN (2012) Antenatal Care as a Determinant of Immun-
ization, and Appropriate Care for Fever and Diarrhoea in Ghanaian
Children. Unpublished MSc Thesis in Epidemiology & Biostatistics,
166
International Journal of Medicine
the School of Graduate and Postdoctoral Studies, the University of
Western Ontario, London, Ontario, Canada.
[16] Otieno AN, Nyawanda OB, Audi A, Emukule G, Lebo E, Bigogo G,
Ochola R, Muthoka P, Widdowson M-A, Shay KD, Burton CD,
Breiman FR, Katz AM & Mott AJ (2014) Demographic, socio-
economic and geographic determinants of seasonal influenza vac-
cine uptake in rural western Kenya, 2011. Vaccine 32, 6699–6704.
Retrieved from https://doi.org/10.1016/j.vaccine.2013.10.089.
[17] Russo G, Miglietta A, Pezzotti P, Biguioh MR, Mayaka BG, Sobze
SM, Stefanelli P, Vullo V & Rezza G (2015) Vaccine coverage and
determinants of incomplete vaccination in children aged 12–
23months in Dschang, West Region, Cameroon: a cross-sectional
survey during a polio outbreak. BMC Public Health, 15(630).
https://doi.org/10.1186/s12889-015-2000-2.
[18] World Health Organization (2016a). ‘‘Children: Reducing Mortali-
ty.’’ Fact Sheet updated September, 2016. Accessed from:
http://www.who.int/mediacentre/factsheets/fs178/en/ on 22/05/2017.
[19] World Health Organisation (2016b). Measles: Fact Sheet reviewed
March, 2016. Accessed from:
http://www.who.int/mediacentre/factsheets/fs286/en/ on 02/06/2016.
[20] World Health Organization (2014). Poliomyelitis, Fact Sheet NO
114. 2014
