RESUMEN
BACKGROUND: We aimed to assess the effect of surgeons' shunting practice and shunt use on the early outcomes of carotid endarterectomy (CEA) in recently symptomatic patients. METHODS: We conducted a retrospective observational study based on a multicenter national prospective database. The Vascular Quality Initiative database (2010-2019) was queried for CEAs performed within 14 days after an ipsilateral stroke or transient ischemic attack. Surgeons were gauged as routine shunters if they shunted in >95% of CEAs, otherwise were classified as selective shunters. In-hospital stroke and death rates were compared between routine and selective shunters, stratifying by type of index event (transient ischemic attack versus stroke) and timing of CEA (≤2 versus >2 days). RESULTS: Thirteen thousand four hundred sixty-nine CEAs were performed after a transient ischemic attack (43%) or stroke (57%), 3186 (24%) by routine shunters, and 10 283 (76%) by selective shunters. Comparing routine and selective shunters, in-hospital stroke (1.9% versus 2.4%; P=0.09) and death (0.4% versus 0.5%; P=0.73) rates were similar. A lower stroke rate (1.5% versus 4.2%; P=0.02) was achieved by routine shunters for CEA performed <2 days after an ischemic stroke. Among selective shunters, a higher stroke rate occurred in case of shunt use (2.9% versus 2.3%; P<0.01), mainly due to cases presenting with stroke (3.5% versus 2.4%; P<0.01) but not transient ischemic attack (1.8% versus 1.5%; P=0.57). Awake anesthesia was adopted in 7.8% of cases by selective shunters and in 0.8% by routine shunters, without impact on the perioperative stroke rate (1.8% versus 2.3%; P=0.349). CONCLUSIONS: In this large national cohort, the overall outcomes of CEA were similar between routine and selective shunters. A lower postoperative stroke rate was achieved by routine shunters in CEA performed <2 days after an ischemic stroke. Among selective shunters, intraoperatively indicated shunting determined an increased stroke rate, likely due to intraoperative hypoperfusion. These data may guide the decision regarding timing of CEA and shunting intention in symptomatic patients.