Impact of practice patterns in shunt use during carotid endarterectomy with contralateral carotid occlusion

Abstract

Purpose: This study investigated the association between surgeon practice pattern in shunt placement and 30-day stroke/death in patients undergoing carotid endarterectomy (CEA) with contralateral carotid occlusion (CCO).

Methods: Among 6379 CEAs performed in the Vascular Study Group of New England (VSGNE) between 2002 and 2009, we identified 353 patients who underwent CEA with CCO and compared the 30-day stroke/death rate with 5279 patients who underwent primary, isolated CEA with a patent contralateral carotid artery. Within patients with CCO, we examined the 30-day stroke/death rate across the reason for shunt placement and two distinct surgeon practice patterns in shunt placement: surgeons who selectively used a shunt (≤95% of CEAs) or routinely used a shunt (>95% of CEAs). We used observed/expected (O/E) ratios to provide risk-adjusted comparisons across groups.

Results: Of 353 patients with CCO, 118 (33%) underwent CEA without a shunt, 173 (49%) underwent CEA using a shunt placed routinely, and 62 (18%) had a shunt placed for a neurologic indication. Rates of 30-day stroke/death across categories of reason for shunt use were no shunt, 3.4%; routine shunt, 4.0%; and shunt for indication, 4.8% (P = .891). The risk of 30-day stroke/death was higher for surgeons who selectively placed shunts (5.6%) in all their CEAs and lower for surgeons who routinely placed shunts (1.5%, P = .05). The risk of 30-day stroke/death was >1 in patients undergoing selective shunting (O/E ratio, 1.4; 95% confidence interval [CI], 1.1-1.7) and <1 for surgeons who placed shunts routinely (O/E ratio, 0.4; 95% CI, 0.2-0.9). Stroke/death rates were lowest when individual surgeons' intraoperative decisions reflected their usual pattern of practice: 1.5% stroke/death rate when "routine" surgeons placed a shunt, 3.4% when "selective" surgeons did not place a shunt, and 7.6% stroke/death rate for "selective" surgeons who placed a shunt (P = .05 for trend).

Conclusions: The risk of 30-day stroke/death is higher in CEA in patients with CCO than with a patent contralateral carotid artery. Surgeons who place shunts selectively during CEA have higher rates of stroke/death in patients with CCO. This suggests that shunt use for CCO during CEA is associated with fewer complications, but only if the surgeon uses a shunt as part of his or her routine practice in CEA. Surgeons should preoperatively consider their own practice pattern in shunt use when faced with a patient who may require shunt placement.

Figure 1

Cohort formation.

Figure 2

Surgeon practice patterns in shunt utilization.

Figure 3

30-day stroke and death rate in patients with a CCO, by shunt type and surgeon practice pattern

Appendix 1

30-day stroke and death rate in patients with a CCO, by shunt type and surgeon practice pattern, excluding selective surgeons who shunt >50% of patients . Findings were similar fodr 30% and 90% cutpoints.