CARDIOVASCULAR JOURNAL OF AFRICA • Volume 25, No 1, January/February 2014
36
AFRICA
velocity (PVd) were measured from the apical four-chamber
view by placing a sample volume in the right upper pulmonary
vein using Doppler echocardiography (Fig. 1A).
Isovolumic contraction time (IVCT), isovolumic relaxation
time and ejection time (ET) were assessed by simultaneously
measuring the flow into the LV outflow tract and mitral
inflow using Doppler echocardiography (Fig. 1B). The index of
myocardial performance (IMP or Tei index) was calculated by
dividing the sum of IVRT and IVCT by ET.
The pulsed-wave tissue Doppler imaging (TDI) was
performed by activating the tissue Doppler function in the
same echocardiographic machine. Mitral annulus velocities
(myocardial diastolic velocities) were measured using a pulsed-
wave TDI technique by placing a 1–2-mm sample volume
at the level of the septal and lateral annulus. Early diastolic
and late diastolic (Aa) velocities of the mitral annulus were
determined from the septal and lateral aspects, and the average
was calculated.
In addition, ratios such as E/Ea, E/Vp, IVRT/Tei, and PVs/
PVs
+
PVd were calculated. The time intervals between the peak
of the R wave and the onset of the mitral E velocity, as well as
the time interval between the peak of the R wave and the onset
of Ea at the lateral mitral annulus were also measured (Fig. 1C).
All Doppler measurements were obtained a maximum of three
hours before cardiac catheterisation.
Haemodynamic measurements were done by placing a 6-F
fluid-filled catheter in the LV from the right femoral approach
under fluoroscopic guidance. The fluid-filled pressure was
balanced and calibrated with the external pressure transducer
positioned at the mid-axillary level. All recordings were
performed before the injection of contrast agent. LV end-diastolic
pressure was measured at the nadir of the atrial contraction wave
before the onset of rapid LV systolic pressure rise or at the peak
of the R wave in a simultaneous ECG if the atrial contraction
wave did not exist.
In MS patients undergoing percutaneous commissurotomy,
the mean left atrial pressure (LAP) was also recorded.
Haemodynamic data were collected at end-expiration by an
investigator unaware of the echocardiographic measurements
and represented the average of five and 10 cycles in sinus and
AF rhythm, respectively.
Statistical analysis
We described continuous variables as mean
±
standard deviation
(SD) and categorical data are expressed as frequencies and
percentages. Two variables (right atrial area and LV diameter)
had
>
15% missing data and were omitted from further analysis.
Missing values in other variables were imputed using a multiple
imputation technique. The first set of imputations was used for
further analyses.
Due to the small sample size, we chose to perform a
univariate pre-selection of clinically relevant predictors with a
p
-value threshold of 0.3. We then applied a backward selection
procedure to develop the final prediction model using linear
regression. Model performance was quantified with regard to
discrimination [area under the receiver operating curve (AUC)].
The AUC ranges from 0.5 to 1.0 for sensible models. Statistical
analyses were done with SPSS for Windows (SPSS Inc, Chicago,
Ill), and R for Windows (Version 2.11.1).
Results
All patients had moderate to severe MS, and 20 (58.8%) had
severe MS (MVA
≤
1 cm
2
). The mean MVA was 0.89
±
0.19 cm
2
.
Less than moderate AI and MR were seen in 60 and 66.7% of
patients, respectively. Recording of PV flow was feasible in 30
out of 33 patients (90%). Echocardiographic and haemodynamic
characteristics of the patient population are reported in Table 1.
The mean LVEDP for the 33 patients was 9.9
±
5.3 mmHg
and ranged from 3–25 mmHg. The results of the univariate
analyses are presented in Table 2. In univariate analysis, the only
significant relationship was noted with left atrium area (LAA)
Table 1. Summary of haemodynamic and echocardiographic
measurements in patients with mitral stenosis
Mean
±
SD (
n
=
33)
Heart rate (bpm)
83.4
±
20.2
Mean arterial pressure (mmHg)
83.2
±
10.1
Mean pulmonary pressure (mmHg)
44.3
±
20.2
LVEF (%)
46.4
±
7.7
Left atrial area (cm
2
)
28.4
±
12.2
Average annular Ea (cm/s)
5.5
±
1.9
Average annular Aa (cm/s)
5.3
±
1.5
Average E/Ea
38.0
±
17.5
IVRT (ms)
55.1
±
10.3
Tei index
0.3
±
0.1
PVs/PVs + PVd
0.5
±
0.1
TE–Ea (ms)
23.0
±
53.0
Velocity propagation (cm/s)
61.0
±
15.6
E/velocity propagation
0.1
±
0.01
IVRT/TE–Ea
1.1
±
4.8
SD, standard deviation; LVEF, left ventricular ejection fraction; Ea, peak
early diastolic velocity of mitral annulus; Aa, peak late diastolic velocity
of mitral annulus; E, mitral inflow peak early diastolic velocity; IVRT,
isovolumic relaxation time; PVs, pulmonary vein systolic flow velocity;
PVd, pulmonary vein diastolic flow velocity ; TE–Ea, interval between
the onset of mitral E and annular Ea.
Table 2. The results of univariate and multivariate linear
regression for the prediction of lvedp
Univariate model
Multivariate model
Characteristic
Coefficient
(SE)
R
2
p
-value
Coefficient
(SE)
p
-value
Intercept
–
–
– –49.51 (6.31) 0.94
LAA
0.38 (0.19)
0.12 0.05 0.43 (0.18) 0.14
Ea
–0.76 (0.50)
0.07 0.14 –0.89 (0.46) 0.02
Tei index
10.95 (8.9)
0.04 0.23 12.30 (8.08) 0.06
E/Ea
10.51 (6.32) 0.08 0.11
IVRT/TE–Ea
0.33 (0.28)
0.04 0.26
TE–Ea
0.02 (0.02)
0.03 0.30
VP
0.06 (0.07)
0.02 0.42
IVRT
–0.07 (0.1)
0.02 0.46
E/VP
49.00 (115.34) 0.01 0.67
PVs/PVs + PVd –2.73 (13.36) 0.01 0.84
SE, standard error; LAA, left atrium area; Ea, peak early diastolic veloc-
ity of mitral annulus; E, mitral inflow peak early diastolic velocity; IVRT,
isovolumic relaxation time; TE-Ea, interval between the onset of mitral E
and annular Ea; VP, mitral inflow propagation velocity, PVs, pulmonary
vein systolic flow velocity; PVd, pulmonary vein diastolic flow velocity.
