CARDIOVASCULAR JOURNAL OF AFRICA • Vol 24, No 1, January/February 2013
AFRICA
249
enables the visual representation of data so as to establish patterns
and trends between patients’ follow-up visits for INR monitoring and
their geographical context. The aim of this study was, therefore, to
investigate these associations by developing a spatial display instru-
ment, mapping the residential addresses of patients against their
respective designated referral clinics.
Methods:
Patient data including residential addresses, referral clin-
ics and concordance patterns were obtained from the REMEDY
database. REMEDY is an RHD registry targeting 3 000 patients in
its pilot phase. Addresses were converted to geographical coordinates
and ArcGIS 9.3.1® was used for mapping and spatial analyses. The
Geocoded addresses were further checked in ArcGIS for errors in
coordinate data. The travel distances between individual residential
addresses and referral clinics were calculated and compared to
concordance patterns of patients.
Results:
RHD patients (
n
=
26) reside between 1.2 and 26.2 km
from their referred clinics (mean 9.68
±
8.1 km). The average period
between clinic visits was 32 days (range 18–43 d). Preliminary
results suggest that concordance to monthly INR monitoring is not
associated with patients’ travel distance.
Discussion:
This is the first attempt at the application of GIS within
this area of RHD. Distance from referred clinic does not impact upon
patients’ attendance at INR monitoring. Nevertheless, GIS presents
as an ideal tool to visual relationships between RHD patients’ follow-
up visits for INR monitoring and their geographical context.
1538: RISK FACTORS ASSOCIATED WITH THE DEVELOP-
MENT OF SYSTEMIC-PULMONARY COLLATERAL FLOW
IN SINGLE VENTRICLE PATIENTS WITH SUPERIOR
CAVOPULMONARY CONNECTIONS
Andrew Glatz, Jonathan Rome, Neil Harrison, Adam Small, Yoav
Dori, Matthew Gillespie, Matthew Harris, Mark Fogel, Kevin
Whitehead
Division of Cardiology, Children’s Hospital of Philadelphia, USA
Background:
Systemic-pulmonary collateral flow (CollF) is common
in patients after superior cavopulmonary connection (SCPC). Risk
factors associated with CollF are unclear. We sought to identify risk
factors for CollF in a cross-section of patients with SCPC.
Methods:
A retrospective chart review of events from birth to the
time of study was performed for SCPC patients who had CollF
quantified by cardiac magnetic resonance imaging (CMRI). CollF
was reported as indexed flow (l/min/m
2
), as a percentage of aortic
flow (CollF/Ao), and as a percentage of pulmonary venous flow
(CollF/PV).
Results:
From 4/08 to 8/11, 96 SCPC patients at 2.6
±
1.1 years of
age and 799
±
400 days from SCPC had CollF measured at 1.6
±
0.7
l/min/m
2
, comprising 33.5
±
11.1% of aortic flow and 48.3
±
15.9%
of pulmonary venous flow. Two of three CollF indices were higher
in patients with a prior BT shunt vs no prior BT shunt (pulmonary
artery band, right ventricle-pulmonary artery conduit, or no initial
palliation) (CollF 1.7
±
0.8 vs 1.4
±
0.4 l/min/m
2
,
p
=
0.04; CollF/
Ao: 35.2
±
11.6 vs 28.6
±
7.6%,
p
=
0.009) and in those with a hemi-
Fontan vs bidirectional Glenn (CollF/Ao 38.5
±
13.5 vs 32.2
±
10%,
p
=
0.02; CollF/PV: 57.6
±
20.1 vs 45.8
±
13.8%,
p
=
0.003). With
Spearman testing, positive correlations exist between CollF indices
and indicators of perioperative morbidity. We did not find associa-
tions between CollF and pulse oximetry, haemoglobin, or haemody-
namics at pre-stage 2 catheterisation.
Conclusion:
CollF occurs commonly in patients with SCPC and is
related to surgical history, chest tube and hospital duration. These
data support the hypotheses that perioperative morbidity and pleural
inflammation may play a role in the development of CollF.
1540: POSTMORTEM IMAGING IN PAEDIATRIC CARDI-
OLOGY
Francesca Romana Pluchinotta
1,4
, Prashob Porayette
1
, Sanjay P
Prabhu
2
, Lisa Teot
3
, Stephen P Sanders
1
1
Department of Cardiology, Boston Children’s Hospital, Boston,
USA
2
Department of Radiology, Boston Children’s Hospital, Boston, USA
3
Department of Pathology, Boston Children’s Hospital, Boston, USA
4
Pediatric Cardiology Unit, Department of Pediatrics, University of
Padova, Italy
Background:
Postmortem imaging (PMI), used widely in forensic
pathology, could be substituted for autopsy in paediatric cardiology
patients, increasing effective autopsy rate and facilitating quality
improvement.
Materials and methods:
Paediatric cardiology patients who died
during the study period and in whom an autopsy was planned were
eligible. PMI included magnetic resonance imaging (MRI) of heart,
brain and abdomen (table), and multi-detector computed tomography
(MDCT) angiography, using iodinated contrast diluted 1:8 with PEG,
injected in the inferior vena cava (8 ml/kg) and aorta (5 ml/kg).
Results:
During a 4-month period 5 patients underwent PMI
followed by autopsy; both MDCT and MRI in 4 and MDCT only in
1 because of metal ECMO cannulas. Cardiac anatomy, condition of
surgical repair, and abdominal anatomy by PMI correlated perfectly
with autopsy. All major intracranial findings (intracranial haemor-
rhage, parietal infarct, white matter atrophy) were diagnosed by both
PMI and autopsy. Cause of death by PMI and autopsy was congruent
in 3 cases: not determined in 2 cases and intracranial haemorrhage
with herniation in 1 case. PMI (MDCT only) showed severe bilateral
pulmonary injury, found to be infarction at autopsy, but missed left
ventricle and pulmonary artery thrombosis. PMI missed diffuse left
and right ventricular infarction seen by histology in 1 case. MRI
and MDCT were complementary, with MRI better for delineation
of cardiac and abdominal anatomy, and brain abnormalities, but
MDCT was superior for delineation of central vasculature, especially
the coronary arteries, and pulmonary disease. Position of lines and
tubes, condition of the body wall and skin, and skeletal anatomy were
demonstrated by the MDCT 3D data set.
Conclusions:
Our preliminary study in human cadavers shows that
PMI is feasible in a paediatric hospital environment. The small
sample size precludes analysis of diagnostic accuracy. Nevertheless,
we are encouraged by the diagnostic performance of PMI.
1542: POSTMORTEM IMAGING OF ANTEMORTEM
MYOCARDIAL ISCHAEMIA
Francesca Romana Pluchinotta
1,5
, Prashob Porayette
1
, Patrick O
Myers
2
, Peter Chen
2
, Eric Feins
2
, Lisa Teot
3
, Sanjay P Prabhu
4
,
Stephen P Sanders
1
1
Department of Cardiology, Boston Children’s Hospital, Boston,
USA
2
Department of Cardiac Surgery, Boston Children’s Hospital, Boston,
USA
3
Department of Pathology, Boston Children’s Hospital, Boston, USA
4
Department of Radiology, Boston Children’s Hospital, Boston, USA
5
Pediatric Cardiology Unit, Department of Pediatrics, University of
Padova, Italy
Background
: Postmortem magnetic resonance imaging (PMRI) has
been used in forensic pathology to detect antemortem ischaemia but
the sensitivity is unknown. We tested the survival time required for
detection of antemortem ischaemia by PMRI in a pig carcass model.
Materials and methods:
Nine pigs (7–35 kg) underwent surgical
ligation of the distal left anterior descending (LAD) (8) and/or right
coronary (RCA) branch (4) and were euthanised 1–6 hours after liga-
tion. PMRI (T1, T2, PD, and spin echo 3D volumetric sequences) was
performed 12–48 hours after euthanasia. Images were inspected, and
signal intensity of 17 myocardial segments was measured serially.
Heart sections were submitted for histology.
Results:
Ligation produced discoloration and dyskinesis of the target
segment(s) in all cases. MRI T2-weighted sequences (TE 102, TR
4000) showed the ischaemic area as hyperintense in 4/4 LAD liga-
tions with
≥
4 hours of ischaemic time and in 0/4 with
<
4 hours.
