CARDIOVASCULAR JOURNAL OF AFRICA • Vol 24, No 4, May 2013
114
AFRICA
from the tertiary centre or other differences in management
between the secondary hospitals. Patients presenting to the
secondary-level service provided by the Division of Cardiology at
TBH were more likely to receive invasive in-patient management
with coronary angiography than were those presenting to the SH
(94 vs 51%,
p
<
0.0001). This difference was due to the fact that
the secondary service at TBH is provided by the sub-specialist
Division of Cardiology with immediate access to angiography.
The difference in physical proximity to TBH betweenWH (94 km
away) and KBH (4.6 km away) did not influence the accessibility
of in-patient angiography, with cases being equally likely to be
referred to and accepted by the Division of Cardiology at TBH
from WH and KBH.
The difference in the TIMI risk score for the two groups was
a potential confounder in the analysis of why the TBH cohort
received more angiography than the SH cohort, however both
groups fell in the intermediate risk group, so the significance
of this difference is unclear.
11
The TIMI score was calculated by
the investigators from the case records, as it was not uniformly
documented. The difference in TIMI score between the TBH
cohort and the SH cohort can be accounted for by the difference
in frequency of documented dyslipidaemia and prior aspirin
use in the two groups. This difference may be true, or it may be
only an apparent difference due to the retrospective nature of
the study.
As the baseline data were captured from patient records,
only those cardiovascular risk factors documented could be
captured. If there was no record of medications taken prior to
admission, prior aspirin use could not be assumed. The fact that
the difference between the two cohorts in the TIMI risk score
11
was not reflected by the Grace risk score suggests that this might
be a factor of documentation rather than one of clinical risk.
14-16
This study also showed a difference in in-patient mortality
between the two cohorts, with a higher mortality in the SH
cohort. This was despite the fact that the TBH cohort had a
greater risk for mortality within the first 14 days, as assessed
by the TIMI risk score.
11
As the only difference in management
between the two groups was the initial use of
β
-blockers
(although
β
-blocker use at discharge was similar in both groups),
and in-patient invasive management, one must consider that one
of these is responsible for the difference in mortality.
While
β
-blockers are well known to have benefit acutely in
patients with ST-elevation myocardial infarctions (STEMIs),
17,18
this is not known for NSTEMIs. On the other hand, it is well
documented that early invasive management in patients suffering
NSTEMIs improves survival.
2-6
It is therefore reasonable to
assume that the difference in access to in-patient coronary
angiography between the TBH and SH groups was at least in
part responsible for the difference in mortality between the two
cohorts.
Comparing the study patients to the literature, both the TBH
and SH cohorts had a higher mortality rate than expected, both
in-hospital (6 and 23%, respectively) and at six months (10
and 27%, respectively). The literature predicts a mortality of
1.4–4.4% in hospital and 1.9–5.9% at six months to one year.
9,10,19-
22
Whether this was due to differences in in-patient angiography
rates among our cases (51% for the SH group and 94% for
TBH) compared to the literature remains unclear, as in-patient
angiography was performed in 10–98% of patients in these
trials.
10,19-22
The predicted mortality rates for the two cohorts as calculated
by the Grace risk score
14
(in-patient mortality: 8.79
±
14.60%
and 5.49
±
0.09%; mortality at six months: 15.57
±
22.49% and
11.03
±
12.68% for the SH group and TBH groups, respectively)
was also higher than the mortalities expected from the literature,
as quoted above.
9,10,19-22
The patients in this study were a high-risk
group as they all suffered an NSTEMI, whereas the trials quoted
above looked at all non-ST elevation ACS (NSTE-ACS), and
therefore included patients with UAP.
This study also demonstrated a high rate of cardiac failure.
As the presence of cardiac failure was elicited from descriptions
in the records, this may have been a true reflection of the study
population or it may have been due to documentation. This may
explain the high mortality rate in this study, both in reality and
as predicted by the Grace risk score.
14
When examining the two groups separately, the patients in the
TBH cohort came closer to the mortality rate predicted by the
Grace risk score,
14
both in-hospital and at six months (6 vs 5%
in-hospital and 10 vs 11% at six months, respectively) than did
those in the SH cohort (23 vs 9% in-hospital and 28 vs 12% at
six months, respectively), which would imply that the difference
in management (either the increased frequency of angiography
TABLE 5. OUTCOMESAT DISCHARGEANDAT 6 MONTHS
Secondary
hospitals
n
=
43
(%)
TBH
n
=
33
(%)
p
-value
At Discharge
Mortality
10 (23)
2 (6)
0.0326*
Discharge medications (% of
survivors)
Asprin
31 (94)
30 (97)
0.5918
β
-blocker
29 (88)
30 (97)
0.1851
ACE inhibitor
25 (76)
26 (84)
0.4201
ARB
1 (3)
2 (6)
0.5175
Statin
30 (91)
29 (94)
0.6942
Spironolactone
3 (9)
2 (6)
0.6942
Clopidogrel
8 (24)
9 (29)
0.6646
Days in hospital (
±
SD) (% of
survivors)
6.14 (
±
4.33) 5.82 (
±
5.26) 0.2100
At 6 months
Survived (expressed as % of
entire group)
29 (73)
27 (90)
0.0614
Survived (expressed as % of
survivors at discharge)
29 (97)
27 (96)
0.9247
Chest pain (% of survivors at 6
months)
None
17 (59)
16 (59)
0.6206
Occasional
9 (31)
10 (37)
0.8540
CCS 2
2 (7)
1 (4)
0.5960
CCS 3
1 (3)
0.(0)
0.3302
Readmission to hospital (% of
survivors at 6 months)
9 (31)
4 (15)
0.1461
Subsequent angiography (% of
survivors at 6 months)
6 (21)
3 (11)
0.3248
Time from admission (months) 3.86 (
±
2.07) 1.33 (
±
0.58) 0.0121*
Coronary revascularisation
within 6 months (
n
=
43 and 33)
19 (44)
28 (85)
0.0001*
CCS: Canadian Cardiovascular Society angina classification
*
p
-values calculated comparing TBH and secondary hospital groups,
p
<
0.05 was statistically significant.