28 / 78
CARDIOVASCULAR JOURNAL OF AFRICA • Volume 26, No 4, July/August 2015
174
AFRICA
Kruskal–Wallis test was used between the three groups, followed
by the Bonferroni correction for two-by-two comparisons.
Post
hoc
comparison of the mean ranks of all groups was performed.
To compare frequencies of different categorical variables among
the three groups, the chi-squared test or Fisher’s exact test were
used when appropriate. Statistical significance was set at
p
<
0.05, and after Bonferroni correction at
p
<
0.016 for two-by-two
comparisons.
Results
Six hundred and thirty patients were randomised to receive either
milk or lemon juice, or no intervention and their characteristics
are shown in Table 2. Three hundred and thirteen patients
received milk, 246 patients received lemon juice and there was
no intervention in 71 patients. There were 304 females (48%)
and 326 males (52%). The method of stress was exercise in 51%
and pharmacological stress in 49% of patients. There was a
statistically significant difference in the gender (
p
=
0.003) and
ethnicity (
p
=
0.0002) indexes among the different study groups.
In all three groups, infra-cardiac activity was present in the
majority of patients, both on the stress and rest studies (Table
3). At stress, infra-cardiac activity was seen in 84.1, 84.5 and 97%
of patients in G0, G1 and G2, respectively. At rest, infra-cardiac
activity was seen in 91.7, 90.1 and 100% of patients in G0, G1
and G2, respectively. The visual assessment for the presence or
absence of infra-cardiac activity showed a statistically significant
difference among the three groups, both in post stress (
p
=
0.005)
and at rest (
p
=
0.0063) (Table 3).
With regard to the quantitative grading, the majority of
the patients had myocardial activity greater than infra-cardiac
activity (at stress 74, 71 and 67% of patients for G0, G1 and
G2, respectively) (Table 4). This finding was more evident in the
groups with intervention, especially for the studies done at rest,
and was more overt in G0 compared to G2 (
p
=
0.013). For the
rest group, the majority of patients in G0 and G1 had less or
equal interfering infra-cardiac activity. At rest, 81, 83 and 73%
of patients for G0, G1 and G2, respectively, had myocardial
activity greater than or equal to bowel activity. It was interesting
to note that just over one-quarter of patients in G2 (27%) had
infra-cardiac activity greater than myocardial activity, compared
to G0 (19%) and G1 (18%) (Figs 4, 5).
The difference in visual grading was also statistically highly
significant for the three groups in post stress (
p
=
0.0002), and
a similar difference was noted at rest (
p
=
0.004) (Table 4). The
analysis of the quantitative assessment of the total counts for all
subjects and their comparisons within and between groups are
shown in Table 5. The median was obtained in each group for the
variable, and their minimum and maximum values are included.
Discussion
The use of
99m
Tc sestamibi for MPI often results in increased
splanchnic activity, which creates a major problem in the visual
and quantitative interpretation of the inferior and infero-
septal walls of the left ventricle. Infra-cardiac activity arises
predominantly from the liver, hepatobiliary system, bowel
and/or gastro-duodenal reflux and can result in either an
apparent increase or decrease in radiotracer uptake in the
myocardium, especially in the inferior and infero-septal walls
after reconstruction.
7
Table 2. Patient characteristics
Character-
istics
Total
Total
p-value*
Lemon juice
group (G0)
Milk group
(G1)
Control
group (G2)
Number
630
246 (39)
313 (50)
71 (11)
Mean age ±
SD (year)
58.21 ± 11.42 62.03 ± 11.43 61.37 ± 9.02
Gender,
frequency (%)
0.003
Male
326 (52)
109 (44)
171 (55)
46 (65)
Female
304 (48)
137 (56)
142 (45)
25 (35)
Stress,
frequency (%)
0.83
Exercise
319 (51)
127 (52)
144 (46)
48 (68)
Pharma-
cological
311 (49)
119 (48)
169 (54)
23 (32)
Ethnicity,
frequency (%)
0.0002
Black
193 (30)
71 (29)
81 (26)
41 (58)
Caucasian 238 (37)
95 (39)
131 (42)
12 (17)
Indian
140 (22)
59 (24)
68 (22)
13 (18)
Coloured 59 (11)
21 (8)
33 (10)
5 (7)
*Total
p
-value represents the
p
-value for the three study groups.
Table 3. Evaluation of infra-cardiac activity
by visual assessment
MPI
Total
Total
p-value
Lemon
juice group
(G0)
Milk
group
(G1)
Control
group
(G2)
Stress, frequency
(%): presence
of infra-cardiac
activity
0.005
Yes
528
201 (84.1) 257 (84.5) 70 (97)
No
86
38 (15.9) 47 (15.5)
1 (3)
Rest, frequency
(%): presence
of infra-cardiac
activity
0.0063
Yes
564
219 (91.7) 274 (90.1) 71 (100)
No
50
20 ( 8.3)
30 (9.9)
0
Presence of infra-cardiac activity was graded as ‘yes’ and absence as
‘no’ *Total
p
-value represents the
p
-value for the three study groups.
Table 4. Visual grading of the intensity of infra-cardiac
activity versus myocardial activity
Grading
MPI
0
1
2
3
Total
p
-value
Stress, frequency (%)
0.0002
Lemon juice (G0)
38 (16) 138 (58) 46 (19) 17 (7)
Milk (G1)
47 (16) 166 (55) 49 (16) 42 (14)
Control (G2)
1 (1)
47 (66) 19 (28)
4 (6)
Rest, frequency (%)
0.004
Lemon juice (G0)
20 (8) 100 (42) 73 (31) 46 (19)
Milk (G1)
29 (10) 137 (45) 84 (28) 54 (18)
Control (G2)
0
24 (34) 28 (39) 19 (27)
*Total
p
-value represents the
p
-value for the three study groups.
0: absent infra-cardiac activity
1: bowel activity < myocardial activity
2: bowel activity = myocardial activity
3: bowel activity > myocardial activity