CARDIOVASCULAR JOURNAL OF AFRICA • Vol 24, No 7, August 2013
AFRICA
249
ventilatory support was discontinued on postoperative day
four, however, hypoxaemia, possibly due to atelectasis and a
ventilation–perfusion mismatch, required re-intubation. Death
occurred due to respiratory failure on postoperative day six.
Eleven patients (45.8%) in the group of 24 deaths were
found to have the cause of death categorised as preventable.
Six were deemed technical errors and five system errors.
Among the preventable deaths due to technical errors, all
six patients suffered graft thrombosis. In five of the patients,
graft thrombosis was encountered due to LIMA thrombosis.
These patients developed low-cardiac output syndrome (LCOS)
in the early postoperative period, requiring cardiopulmonary
resuscitation and re-exploration, which showed dissection of the
LIMA or haematoma at the LIMA wall. Although in two of these
cases, the left anterior descending (LAD) coronary artery was
bypassed with a saphenous vein graft for a second time, death
was inevitable but secondary to peri-operative MI.
In the sixth patient who had graft thrombosis, ventricular
fibrillation occurred on the third day after surgery and was the
cause of death. Emergency cardiopulmonary bypass (CPB) was
initiated after cardiopulmonary resuscitation. During surgical
exploration, two thrombosed saphenous vein grafts were detected.
Five of the patients in the preventable death group were
considered to have suffered a system error. The two patients
who died of renal failure had excessive bleeding in the early
postoperative period. Haemodialysis was initiated in both
patients but they succumbed on the sixth and eighth days after
surgery, respectively. One of the patients had atrial fibrillation
(AF) but there was no attempt at cardioversion, possibly due to
lack of communication between the consultant and the junior
on-call doctor; the final cause of death was stroke.
Another patient, diagnosed with bullous lung disease, had
subcutaneous emphysema and after an episode of sudden
respiratory insufficiency, died in the ward on day 19 after surgery.
A further patient with hypotension was transferred to the ward by
a junior surgeon on postoperative day three, without considering
that the clinical status of the patient required continuous
inotropic support. This system error led to low-cardiac output
syndrome and subsequent cardiac arrest. The patient failed to
respond to cardiopulmonary resuscitation in the ward.
Discussion
During open-heart surgery, patients with a high risk for
cardiac surgery are extensively studied and there is substantial
documentation of successful surgical outcomes in this group
of patients.
8,9
However, there is not enough data on preventable
deaths in low-risk groups of patients. The investigation of
low-risk groups with a EuroSCORE
≤
2 requires a larger group
of patients to demonstrate accurate mortality rates and it is often
more difficult to predict the possible risk factors that influence
outcomes of mortality in a small group.
In a study by Freed
et al
.,
10
in a group of 4 294 patients with
a logistic EuroSCORE
≤
2, a total of 16 patients (0.37%) died.
It has been claimed that over a third of the deaths studied were
potentially preventable, suggesting that further improvement in
outcomes is possible through modification of surgical technique
or a change in the systematic delivery of cardiac surgical care
and training.
In our study group, of a total of nine patients (37.5%) with
peri-operative myocardial infarction and cardiac-related death,
seven (77.8%) were considered to be preventable, and in five
patients, the main problem was identified as LIMA harvesting.
LIMA–LAD bypass is the mainstay of coronary surgery.
12,13
The LIMA is used extensively as a bypass graft, with excellent
patency rates.
13,14
However, careful technical preparation is
needed to prevent occlusion.
In other studies on isolated CABG, mainly CPB, cardioplegic
solution, problems in myocardial protection, and the length of
aortic cross-clamp were considered to be factors that may have
increased cardiac-related mortality.
10-14
Our study found that in
five of the nine patients (55.6%) who suffered cardiac-related
death, technical problems in LIMA harvesting was the most
important cause of cardiac origin.
Two patients also died of renal failure as a result of excessive
blood transfusions. Meticulous haemostasis and early exploration
for postoperative bleeding may help to prevent excessive blood
transfusion and therefore the development of renal insufficiency
that requires dialysis, which is well known to increase mortality.
15
In one patient, stroke was the reason for death, due to
postoperative atrial fibrillation. However, the cause of death was
secondary to a miscommunication between the junior surgeon
on the ward and the senior surgeon, which caused a preventable
system error. Early cardioversion is crucial and potentially
lifesaving in the face of acute rhythm disturbances, which
require immediate intervention. Late detection in the ward,
miscommunication between the surgical team, and inability to
transfer the patient to the ICU on time are all system-related
errors that need to be identified and solutions discussed in order
to prevent further fatalities.
In another patient who was still in the ward under medical
supervision, underestimation of respiratory insufficiency due
to pneumothorax was the cause of death. A further patient with
clinical signs of deterioration of the haemodynamic status was
transferred to the ward by a junior surgeon who underestimated
the clinical status of the patient, leading to a system error and
death. These system errors can be corrected by more established
protocols and closer follow up of patients after surgery. Common
to both the above cases, delay in the treatment of an identified
and potentially reversible problem was recognised. The cause of
this delay in taking appropriate action was the lack of or unclear
communication between senior and junior surgeons.
There are several similar studies but they differed from our
study in several ways, as the current study was designed to
understand the details of mortality in low-risk cardiac surgery
patients. In both the FIASCO study
10
and the Stockholm
experience,
11
all cardiac surgical procedures were included,
whereas our study dealt with only CABG surgery. Another
difference from the FIASCO study was that patients were
included who died within 30 days of surgery. Previous studies
have only included in-hospital mortality, similar to the study by
Janiec
et al
.
11
The Stockholm experience has revealed excellent results
(0.38% mortality), even in patient groups with a EuroSCORE
≤
3. However, the incidence of preventable deaths was only
13% (2/15) of all deaths. In the FIASCO study, the incidence
of mortality was similar to that of the Stockholm study (0.37%)
in a group of low-risk patients with a logistic EuroSCORE
≤
2.
The incidence of preventable causes of death was 43%, markedly
higher than that of the Stockholm experience.
10,11
