CARDIOVASCULAR JOURNAL OF AFRICA • Vol 24, No 4, May 2013
AFRICA
115
with in-hospital coronary revascularisation, or the earlier use of
β
-blockers, or both) was the cause of the difference in mortality,
as previously discussed. Referring to earlier literature,
10,19,22,23
the
rates of angiography in the TBH (94%) and SH cohorts (51%)
were similar to and less than the rates of angiography in the
early invasive arms of these trials (96–98%),
10,19,22,23
respectively.
However, the conservative arms of these trials had lower rates
of coronary angiography (11–51%) with a lower mortality
rate.
10,19,22,23
This implies additional factors contributing to the
poorer survival in the South African state hospital setting.
The conservative arms in these earlier international trials
10,19,23
included other anti-platelet agents in addition to aspirin (ADP
receptor antagonists or glycoprotein IIb IIIa inhibitors), as
recommended by the ESC.
1
At the time of writing both of these
agents were only available to patients undergoing angiography
in the state hospital setting in the Western Cape. They are not
available for medical management of an NSTE-ACS, even for
those at high risk with NSTEMIs. Whether this was the cause
of the higher-than-expected mortality rate in this study is not
addressed, and further research into this question needs to be
performed.
Examining specifically those cases in the SH cohort who
died during the index admission, a 23.6
±
21.98% probability
of in-hospital mortality was predicted by the Grace score.
14
This
suggests that the recorded mortality rate was high in comparison
with previous studies. Only half of these 10 cases were referred
to TBH, and four of the five referred patients were accepted. It
appears from this that lack of referral (50%) of those patients
who subsequently died may in part be responsible for a poor
outcome. This discrepancy in referral rate and acceptance rate is
reflected in the cohort as a whole as well (72% referral rate and
94% acceptance rate). The lack of referral for tertiary care was
likely to have been a contributing factor to the relative lack of
in-patient angiography and coronary revascularisation in the SH
group, and hence the higher in-patient mortality rate.
This study did not investigate the reasons for referral or lack
thereof, as this information was difficult to obtain retrospectively.
Further research into this should be done in order to fully address
the high mortality rate in the SH cohort.
When looking at coronary revascularisation rates in the
two groups, it would appear that the rates of PCI were low
in both groups (45% for SH and 68% for TBH). This can be
explained by the fact that a high percentage of patients (40%)
had triple-vessel disease, and when coronary revascularisation is
considered in total [via PCI or coronary artery bypass grafting
(CABG)], the rates of revascularisation increased to 82 and 94%
for the SH and TBH groups, respectively, suggesting that lack of
coronary revascularisation was not a contributing factor in the
high mortality.
Limitations
This study had several significant limitations. The retrospective
nature of the study design left the investigators dependent on
clinical records for all data capturing. Record keeping is often
less than optimal and significant information such as patients’
use of medications prior to admission was often not available in
the records. This could render the baseline characteristics and
clinical risk stratification of the patients unreliable, limiting the
comparison between the two groups.
The sample size also limited the study findings. A number of
factors contributed to a small sample size. The study intentionally
targeted a high-risk group of patients with NSTEMIs. Due to
the significant budget constraints experienced in the South
African state healthcare system, many cases had only a single
cardiac troponin value taken. Although it is part of the current
best-practice guidelines that should only one troponin value be
requested, it is taken six to 12 hours after the index event. There
was no record in the clinical notes documenting that this was
practiced. It was not possible to ascertain when the troponin
samples were taken, and therefore cases may have been missed.
There were also a number of cases that were excluded
due to the inability to obtain the patient’s folder (four for the
SH cohort and 13 for the TBH cohort). No information was
available for these cases and while it is not known whether these
patients suffered an NSTEMI, it does raise concerns regarding
the validity of the data. As the investigators were particularly
interested in access to in-patient coronary angiography, there
were strict exclusion criteria which also contributed to the small
number of cases.
The small sample in the WH and KBH groups limited
the data analysis in that the two groups had to be combined
into a single cohort for analysis. Although sub-analysis was
done comparing the WH and KBH groups when statistically
significant differences were found and this did not reveal any
differences between these two groups, the lack of differences
may have been a factor of the sample size. The combination
of WH and KBH into a single cohort may also have masked
differences between these groups had they been independently
compared to the TBH cohort.
Conclusion
Despite these limitations, the study did reveal some significant
differences in the in-patient management of patients presenting
to secondary services at a tertiary centre (TBH) compared
to patients presenting to secondary-level centres (KBH and
WH). These included less initial
β
-blocker use, less in-patient
invasive management and a higher in-patient mortality rate in
the SH group. The lack of difference between the KBH and WH
groups in this regard suggests that geographical proximity to (or
remoteness from) the tertiary centre (TBH) was not a significant
factor determining access to coronary angiography. Clearly
factors other than geography and distance, specific to the level
of service, were influencing both access to in-patient coronary
angiography and in-patient mortality rates.
While it is well established that early angiography has a clear
mortality benefit in patients suffering a NSTEMI,
2-6
and current
research is investigating the optimal timing for angiography;
access to in-patient angiography remains problematic in the
state hospital setting in the Western Cape, South Africa. The
relative lack of access to coronary angiography for patients
presenting to secondary-level hospitals, regardless of their
geographical proximity to the tertiary centre TBH, results in
an adverse mortality outcome for these patients. Coronary
angiography remains a scarce resource, with three cardiologists
in the Division of Cardiology, Department of Medicine at TBH
providing a tertiary service to a population of 2.64 million.
12
This
inequality in access to in-patient invasive management needs to
be addressed as a priority.
