24 / 60
CARDIOVASCULAR JOURNAL OF AFRICA • Volume 32, No 1, January/February 2021
22
AFRICA
This was a cross-sectional study to determine the prevalence
and spectrum of CHD among neonates in the immediate
postnatal period (first week of life) over a period of two years
from February 2017 to January 2019. The study population
included neonates delivered or attended to within the first week
of life in JUTH and PSSH and their affiliated immunisation
centres. All babies aged one week or younger, delivered or
provided with treatment or immunisation services in JUTH and
PSSH were eligible for the study as long as their mothers or
caregivers provided written informed consent.
A total-population approach to sampling was employed in
this study where all babies who met the inclusion criteria in
the two health institutions were recruited and sampled within
the study period upon consent. The neonates were recruited
on weekdays before their discharge from the hospital or before
routine vaccinations at the immunisation centres.
Approval for the study was obtained from the Institutional and
Health Research Ethics Committee of JUTH, while permission
was obtained from PSSH and the affiliated immunisation centre
before the study commenced. Written informed consent was also
obtained from each neonate’s mother or primary caregiver.
Five residents in paediatrics at JUTH served as research
assistants and were trained by the lead investigator for two days
on the study protocol, administration of questionnaires, the
required clinical examination, and documentation of findings,
including echocardiogram findings.
A semi-structured proforma developed specifically for this
study was used to obtain clinical and demographic information
about the neonates and their parents, from mothers/caregivers.
Physical examination findings of the neonates were also
documented in the proforma, which was pre-tested on 34
babies to identify and address ambiguities, determine ease of
administration, appropriateness of the questions, and to estimate
the average duration of data collection.
Demographic and clinical information about every neonate
was documented by the trained research assistants. Demographic
data such as maternal parity, age and educational status of
parents, the neonate’s gender, postnatal age, as well as gestational
age at the time of delivery, were recorded. The neonates’ birth
weights (taken immediately after birth) were obtained from their
delivery records. Their crown-to-sole lengths and occipitofrontal
circumferences (OFC) were measured in centimetres using an
infantometer and non-stretchable tape, respectively, according to
standard methods.
12
Three measurements of each were taken and
the average (to the nearest 0.1 cm) was determined and recorded.
A transthoracic Doppler echocardiogram was performed
by the lead researcher, who has been trained in and routinely
performs paediatric echocardiography, on all the enrolled neonates
using a Vivid e
®
portable echo machine (GE, China, May 2016).
A diagnosis of CHD was made in any newborn with single or
multiple structural heart defects. The diagnosis of the various
types of CHD was made based on the ICD-10 diagnostic codes.
13
The CHD were classified as cyanotic or acyanotic defects,
based on the presence or absence of cyanosis, and also as mild,
moderate or severe lesions.
5,10-13
Mild lesions included small atrial
septal defects (ASD) 3–5 mm in diameter, and small ventricular
septal defects (VSD) < 3 mm in diameter. Other mild lesions
included pulmonary stenosis (PS) with peak gradient < 30
mmHg, and bicuspid aortic valve without aortic stenosis or
incompetence. Although patent ductus arteriosus (PDA) < 1.5
mm in diameter after the first week of life was also considered a
mild lesion, since the study was carried out in the first week of
life, PDA < 1.5 mm in diameter was not categorised as CHD in
this study. (Small ASD was differentiated from a patent foramen
ovale by the absence of a flap in the latter.
13
)
Moderate lesions were: large ASD > 5 mm in diameter,
moderate-sized PDA and VSD measuring 1.5–3 mm and 3–6
mm, respectively, complex forms of VSD associated with other
CHD, non-critical coarctation of the aorta, moderate pulmonary
stenosis with peak gradient of 30–60 mmHg, and mild-to-
moderate aortic stenosis with ≤ 50 mmHg peak gradient.
Severe lesions included all cyanotic CHD, as determined
by the presence of central cyanosis, which was defined as
oxygen saturation < 95% in the absence of other causes,
14
and
acyanotic CHD such as large VSD > 6 mm, large PDA > 3
mm, atrioventricular septal defects (AVSD), severe pulmonary
stenosis with peak gradient > 60 mmHg, severe aortic stenosis
> 50 mmHg peak gradient, and any critical CHD such as severe
duct-dependent lesions requiring urgent surgical intervention
for survival. These included hypoplastic left heart syndrome
(HLHS), critical coarctation of the aorta (CoA), critical aortic
stenosis, tricuspid atresia without shunt defects, total anomalous
pulmonary venous connection (TAPVC) and severe tetralogy of
Fallot (TOF).
15
Septal defects were further classified based on the location
of the defect in the septum: ostium primum, ostium secundum
and sinus venosus defects for ASDs and peri-membranous, inlet,
outlet and muscular defects for VSDs.
16
Newborns found to have CHD were referred to the Paediatric
Cardiology unit in JUTH for in-patient care or out-patient
follow up, depending on the infant’s clinical condition. Pre-term
neonates with PDA were managed according to the hospital’s
protocol.
Statistical analysis
Data were processed and analysed using STATA
®
statistical
software version 14.0 (Texas, USA). Means and standard
deviations were used as summary indices for numerical data
such as age, weight, length and OFC, while non-numerical data
such as gender, and type and severity of CHD were presented
as frequencies or percentages, or as charts. The unpaired
student’s
t
-test was used to test the difference between means of
continuous variables such age, weight and gestational ages of
neonates with and without CHD.
Analysis of variance (ANOVA) was used to compare means
of more than two groups, while the chi-squared test was used to
compare non-numerical characteristics such as gender, place of
delivery and type of care received by neonates with and without
CHD. Multivariate analysis was also performed to determine
the association of CHD with maternal age, and neonateal
gestational age and weight. A 95% confidence interval was used
in this study as the interval estimate and a
p
-value of ≤ 0.05 was
considered statistically significant.
Results
A total of 3 857 babies were recruited into the study. There were
2 016 males and they comprised 52.3% of the total. The neonates
were born at a mean gestational age of 36.9 ± 9.2 weeks. The