Variation in surgical and anaesthetic technique and associations with operative risk in the European carotid surgery trial: implications for trials of ancillary techniques

Objectives: several ancillary surgical techniques, such as shunting and patching, are used in association with carotid endarterectomy. However, the balance of risks and benefits of these techniques is uncertain because of the lack of large randomised controlled trials (RCTs). To assess the potential for further trials, we studied the variation in use of these techniques by surgeon and by country in the European Carotid Surgery Trial (ECST).

Methods: use of each ancillary technique was assessed by surgeon and by country. For each technique, the relationships between the use of the technique and baseline patient characteristics, use of other techniques, and the 30-day operative risk of stroke and death were determined.

Results: there was considerable variation between surgeons in the use of ancillary operative techniques both within (p<0.001 for shunting and patching), and between countries (p<0.001 for shunting and patching). Some surgeons used techniques selectively, and so the characteristics of patients differed depending on which techniques were used. Use of each technique was also significantly associated with the use of other techniques. Multiple regression analysis, taking into account all these factors, found no statistically significant associations between operative risk and the use of shunting, patching, intra-operative EEG monitoring, or type of anaesthetic. The only surgical technique significantly associated with an increased operative risk was not using intra-operative anticoagulation (hazard ratio=2.33, 95% CI=1.4-4.2). Other factors associated with an increased risk were an operation time of less than 1 h, or greater than 1.5 h, and the surgeons' subjective assessment that the operation was difficult.

Conclusions: in the ECST, operative risk was more closely related to patient characteristics, length of surgery, and the surgeons' perception of the difficulty of the operation, than to the use of particular ancillary operative techniques. The considerable variation between surgeons, and between countries, in the use of ancillary techniques is in keeping with the lack of convincing data from RCTs, and suggests that there should be sufficient uncertainty to make large pragmatic trials possible.