12 / 72
CARDIOVASCULAR JOURNAL OF AFRICA • Volume 27, No 2, March/April 2016
66
AFRICA
Electrocardiographic predictors of peripartum
cardiomyopathy
Kamilu M Karaye, Krister Lindmark, Michael Y Henein
Abstract
Objective:
To identify potential electrocardiographic predic-
tors of peripartum cardiomyopathy (PPCM).
Methods:
This was a case–control study carried out in three
hospitals in Kano, Nigeria. Logistic regression models and a
risk score were developed to determine electrocardiographic
predictors of PPCM.
Results:
A total of 54 PPCM and 77 controls were consecu-
tively recruited after satisfying the inclusion criteria. After
controlling for confounding variables, a rise in heart rate of
one beat/minute increased the risk of PPCM by 6.4% (
p
=
0.001), while the presence of ST–T-wave changes increased
the odds of PPCM 12.06-fold (
p
<
0.001). In the patients,
QRS duration modestly correlated (
r
=
0.4;
p
<
0.003) with
left ventricular dimensions and end-systolic volume index,
and was responsible for 19.9% of the variability of the latter
(
R
2
=
0.199;
p
=
0.003). A risk score of
≥
2, developed by scor-
ing 1 for each of the three ECG disturbances (tachycardia,
ST–T-wave abnormalities and QRS duration), had a sensitiv-
ity of 85.2%, specificity of 64.9%, negative predictive value
of 86.2% and area under the curve of 83.8% (
p
<
0.0001) for
potentially predicting PPCM.
Conclusion:
In postpartum women, using the risk score could
help to streamline the diagnosis of PPCM with significant
accuracy, prior to confirmatory investigations
Keywords:
peripartum cardiomyopathy, electrocardiogram,
predictors, risk score
Submitted 25/1/15, accepted 7/12/15
Cardiovasc J Afr
2016;
27
: 66–70
www.cvja.co.zaDOI: 10.5830/CVJA-2015-092
Although peripartum cardiomyopathy (PPCM) was
first described in 1880, much remains unknown about it.
1,2
Electrocardiography (ECG) is an inexpensive and important
tool for evaluating cardiac electrical function and is widely
available, even in limited-resource settings. A recent study found
the majority (96%) of PPCM patients had ‘abnormal’ 12-lead
ECGs at presentation and highlighted the usefulness of the ECG
in screening and prognosticating patients at risk in resource-poor
settings.
3
To the best of our knowledge, there is a paucity of
ECG data in PPCM, and no data on its use in the diagnosis of
PPCM in women presenting with clinical features of heart failure
towards the end of pregnancy or during the puerperium.
The aim of this study was to determine potential ECG
variables that predict the diagnosis of PPCM. If proved, such
variables could help to streamline the diagnosis of PPCM prior
to confirmatory investigations, particularly in limited-resource
settings.
Methods
This was a case–control study carried out in the Murtala
Mohammed Specialist Hospital (MMSH), Aminu Kano
Teaching Hospital (AKTH), and a private cardiology clinic
in Kano, Nigeria. The research protocol was approved by
the ethics committees of each of the study centres, and the
study conformed to the ethics guidelines of the Declaration of
Helsinki, on the principles for medical research involving human
subjects.
4
The inclusion criteria for the patients were: (1) confirmed
diagnosis of PPCM; (2) onset of symptoms towards the end of
pregnancy or within the puerperium, and presentation to hospital
within nine months postpartum; (3) age of at least 18 years; and
(4) written informed consent. Patients were excluded if: (1) their
symptoms could be explained by diagnoses other than PPCM;
(2) their symptoms started in early pregnancy or after the first
five months postpartum; (3) they were younger than 18 or older
than 45 years; (4) they denied consent to participate.
To be included, the controls had to satisfy the following
criteria: (1) be apparently healthy; (2) no past history of any
cardiac disease or systemic hypertension (except pregnancy-
induced hypertension); (3) normal ECG (except for flat T
waves in leads III or aVF, and inverted T waves in aVR, V1
or V2, which are considered non-specific);
5
(4) present to
the study centres within nine months postpartum for routine
immunisations for their children; and (5) give written informed
consent. Subjects taking drugs known to affect ECG intervals
were excluded from the study.
6
T-wave inversion with or without ST-segment depression were
considered abnormal in all leads except aVR, V1 and V2.
5
In
addition, flat T waves in leads III or aVF were also considered
non-specific.
5
Controls were excluded if: (1) they presented their children for
immunisation after five months postpartum; (2) they presented
to the hospital as patients; (3) they were younger than 18 or older
than 45 years; (4) they were known or found clinically to have
any cardiac disease; (5) they denied consent. The sample size for
Department of Medicine, Bayero University and Aminu
Kano Teaching Hospital, Kano, Nigeria
Kamilu M Karaye, MB BS, DIC, MSc, FWACP, FESC, FACC,
kkaraye@yahoo.co.ukDepartment of Public Health and Clinical Medicine, Umea
University, Sweden
Kamilu M Karaye, MB BS, DIC, MSc, FWACP, FESC, FACC
Krister Lindmark, MD, PhD, FACC, FESC
Michael Y Henein, MD, PhD, FACC, FESC
Department of Cardiology, Umea Heart Centre, Umea, Sweden
Krister Lindmark, MD, PhD, FACC, FESC
Michael Y Henein, MD, PhD, FACC, FESC