27 / 60
CARDIOVASCULAR JOURNAL OF AFRICA • Volume 32, No 1, January/February 2021
AFRICA
25
those with normal hearts (35.7 ± 10.2 vs 36.9 ± 9.1 years,
p
=
0.19) (Table 5). On multivariate regression analysis, the mother’s
age (
p
= 0.006), lower birth weight (
p
= 0.003), in-patient care (
p
< 0.001) and the presence of dysmorphism (
p
< 0.001) remained
statistically significant predictors of CHD.
The five most common lesions detected were all acyanotic.
Isolated VSDs and ASDs occurred more commonly and
accounted for 9.2 and 6.5 per 1 000, respectively. The other
three common lesions detected were combined ASD and VSD
lesions, mild PS and AVSD, detected in 3.4, 2.3 and 1.3 per
1 000. A comparison of the relative distributions of the five most
common subtypes in other regions is shown in Table 6.
Discussion
The prevalence of CHD in the first week of life in our study
was found to be 28.8 per 1 000, with a hospital birth prevalence
of 27.4 per 1 000 live births. Infants with mild CHD had a
prevalence of 17.9 per 1 000, moderate CHD 6.0 per 1 000
and severe CHD 3.4 per 1 000. In addition, acyanotic CHD
accounted for 91.9% of all the CHD identified, with VSD being
the commonest acyanotic CHD found in 31.2%. Neonates with
CHD were born to older women, had a lower mean birth weight
and were more likely to be admitted for a medical illness.
The prevalence of CHD detected in this study is higher than
the 15 per 1 000 reported from Pakistan,
17
even though the latter
enrolled only sick babies admitted within the first 12 hours of
life, who might have been expected to have a higher prevalence.
In China, a prevalence of 22.9 per 1 000 was detected among
three-month-old infants seen in a tertiary hospital.
18
The older
age of the subjects in the Chinese study could explain their
slightly lower prevalence compared to the present study.
Our finding is however at variance with that of a recent meta-
analysis by Liu
et al
.,
19
who showed that the prevalence of CHD
in Africa was significantly lower than that in other continents.
The authors attributed this to poor access to appropriate
healthcare resources in Africa, leading to low CHD detection
rates. However, only four of the 260 studies (study populations
from birth to six years) reviewed were from the continent of
Africa, highlighting the paucity of data from the continent,
which might have been responsible for the apparently low
prevalence. This emphasises the need for more research on the
prevalence of CHD among infants and young children in Africa.
The present study helps to bridge this gap by contributing to the
pool of data on CHD prevalence from this region.
The birth prevalence of 27.4/1 000 live births obtained in
our study is much higher than the 8.0 per 1 000 reported by
Dolk
et al
. across Europe and the 13.7 per 1 000 documented by
Gillum
et al
. in the United States.
20,21
Both studies were registry-
based studies documenting cases diagnosed after referral for
evaluation for CHD, with the potential to underestimate mild
lesions that are largely asymptomatic. A study in Saudi Arabia
also documented a lower birth prevalence of 14.6 per 1 000 live
births compared with the present study.
22
The latter enrolled only
pregnant women attending antenatal clinics, who were followed
up until delivery when their babies had echocardiography,
unlike in our study where all neonates delivered at the two study
hospitals were enrolled, irrespective of where they had received
antenatal care. In addition to the different methodologies
used, differences in the presence of causal mechanisms that
predispose to CHD may account for the wide variations in the
prevalence seen in different parts of the world.
23
The birth prevalence obtained in this study is also much
higher than the 3.5 per 1 000 (for live and still births) reported
by the only available hospital-based prevalence done by Gupta
and Antia in 1967.
10
While our study used echocardiography as
the diagnostic tool, Gupta and Antia used clinical auscultation
and autopsy of still-born babies to estimate the prevalence.
10
Echocardiography helps to detect even asymptomatic lesions,
which would not be easily detected using clinical auscultation.
Therefore, the absence of echocardiography obviously limited
Gupta and Antia’s detection of many mild and moderate lesions
and possibly also some severe lesions that were asymptomatic in
the immediate neonatal period.
Similar to previous reports, acyanotic CHDs were found to
be more common than cyanotic lesions in our study, with VSD
being the most frequent lesion, followed by ASD.
5,18
Although
the relative distributions of septal defects in this study were
much higher than the global prevalence reported by Liu
et al
.,
19
most of the lesions were mild. The high prevalence of VSDs,
especially small lesions, have been shown to be responsible
for the increase in the occurrence of mild CHD and therefore
overall CHD rates found in the present study.
23
Table 6. Comparison of the relative distributions of
common CHD subtypes with other studies
Study
Variables
Liu
et al
.
Hussain
et al.
Sun
et al
.
Ige
et al
.
Location
Global Pakistan
China
Nigeria
Subtypes (per 1 000)
Ventricular septal defect
3
3.1
0.8
9.2
Atrial septal defect
1.4
2.3
0.9
6.5
Patent ductus arteriosus
1.0
1.5
0.4
–
Tetralogy of Fallot
0.3
0.7
0.05
–
Pulmonary stenosis
0.5
0.6
0.01
2.3
Atrioventricular septal defect
–
–
–
1.3
Combined atrial and ventricular
septal defects
–
–
–
3.4
Table 5. Factors associated with CHD
Characteristics
CHD
(111)
No CHD
(3 746)
p-
value
Gender
Male
62
1 954
Female
49
1 792
0.44
Place of delivery
JUTH/PSSH
64
2 276
Other hospitals
42
1 262
Home
5
250
0.40
In-patient care
Yes
35
353
No
76
3 393
< 0.001
§
Dysmorphism
Yes
15
2
No
96
3 744
< 0.001
§
Maternal age in years (mean ± SD)
30.6 ± 5.6 28.4 ± 6.8 < 0.001
§
Neonatal weight in kg (mean ± SD)
2.8 ± 0.6
3.1 ± 0.6 < 0.001
§
Gestational age in weeks (mean ± SD)
35.7 ± 10.2 36.9 ± 9.1
0.19
§
Statistically significant.