There were 0 (0%) deaths in the GA group and 2 (2%)in the LA group.
There were 4 (3%) strokes in the GA group and 2 (2%) in the LA group.
There were 4 (3%) of episodes of death or stroke in the GA group and 3 (3%) in the LA group.
There were 2 (1%) transient ischaemic attacks in the GA group and 1 (1%) in the LA group.
There were 2 (1.4%) myocardial infarctions in the GA group and 0 (0%) in the LA group.
There were 4 (3%) arrhythmias in the GA group and 2 (2%) in the LA group.
There was 1 (<1%) haematoma in the GA group and 1 (1%) in the LA group.
There was 1 (<1%) nerve injury in the GA group and 1 (1%) in the LA group.
None of the differences in these outcomes reached statistical significance.
In terms of operative factors:
80% of the surgeons performing the operation in the GA group were consultants and 20% were surgical trainees, versus 100% consultants and 0% surgical trainees in the LA group, (p<0.001);
shunt use was 50% in the GA group versus 13% in the LA group, (p<0.001);
patch use was 25% in the GA group versus 41% in the LA group, (p=0.01);
intimal suture was 3.6% in the GA group versus 2% in the LA group, (p non significant); and
operation time was 110 minutes (interquartile range: 100 - 130) in the GA group versus 98 minutes (interquartile range: 84 - 112) in the LA group, (p<0.001).
The systolic blood pressure data suggested that hypertension (and subsequent treatment) was significantly more likely in the LA group, while hypotension (and subsequent treatment) was significantly more likely in the GA group.
In terms of hospital stay, the median length of stay was 2 days (range: 1 - 2) in the LA group. In the GA group, before 1996, the median total stay was 4 days (range: 1 - 4) and the stay in the high dependency unit (HDU) was 1 day (range: 1 - 2). After 1996, the medial total stay was 3 days (range: 1 - 4) and no patient required HDU stay.