In-hospital outcomes after carotid endarterectomy for stroke stratified by modified Rankin scale score and time of intervention.

OBJECTIVE: Although the benefits of carotid endarterectomy (CEA) for treating symptomatic carotid stenosis are well known, the optimal timing of intervention after acute stroke and whether the optimal timing will vary with preoperative stroke severity has remained unclear. Therefore, we assessed the effect of stroke severity and timing of the intervention on the postoperative outcomes for patients who had undergone CEA for stroke. METHODS: We identified all patients in the Vascular Quality Initiative who had undergone CEA from 2012 to 2020 for prior stroke. The patients were stratified using the preoperative modified Rankin scale score (mRS score, 0-5) and time to CEA after stroke onset (≤2 days, 3-14 days, 15-90 days, 91-180 days). After univariate comparisons, the patients were stratified into the following mRS cohorts for further analysis: 0 to 1, 2, 3 to 4, and 5. The primary outcome was in-hospital stroke/death. RESULTS: We identified 15,601 patients, of whom 30% had had an mRS score of 0, 34% an mRS score of 1, 17% an mRS score of 2, 11% an mRS score of 3, 8% an mRS score of 4, and 1% an mRS score of 5. Overall, 9.3% of the patients had undergone CEA within ≤2 days, 46% within 3 to 14 days, 36% in 15 to 90 days, and 8.4% within 90 to 180 days. A decreasing mRS score and an increasing time to CEA were associated with lower rates of perioperative stroke/death (Ptrend < .01). After risk adjustment, with CEA at 3 to 14 days as the comparator group, the mRS score 0 to 1 group had had a higher incidence of stroke/death after CEA within ≤2 days (3.6% vs 2.0%; odds ratio [OR], 1.8; 95% confidence interval [CI], 1.2-2.7). The mRS score 2 group had had a similar incidence of stroke/death after CEA within ≤2 days (4.4% vs 3.9%; OR, 1.2; 95% CI, 0.6-2.3) but a lower incidence after CEA at 15 to 90 days (2.1% vs 3.9%; OR 0.5; 95% CI, 0.3-0.96). The mRS score 3 to 4 group had had a higher incidence of stroke/death after CEA within ≤2 days (8.0% vs 3.8%; OR, 2.4; 95% CI, 1.5-3.9) but a similar incidence of stroke/death after CEA at 15 to 90 days (3.0% vs 3.8%; OR, 0.8; 95% CI, 0.5-1.3). For the mRS score 5 group, the stroke/death rates were ≥6.5% across all the time to CEA groups. However, the low sample size limited meaningful comparisons. CONCLUSIONS: Patients with minimal disability after stroke (mRS score, 0-1) seemed to benefit from CEA within 3 to 14 days. However, those with severe disability (mRS score 5) have a very high risk from CEA at any time point given the poor outcomes. In contrast to the current guidelines, patients with mild disability (mRS score 2) could benefit from delaying CEA to 15 to 90 days, and those with moderate disability (mRS score 3-4) might benefit from CEA within 3 to 90 days given the acceptable in-hospital outcomes. These data should be considered within the context of the clinical situation in the weeks after index event to determine the net benefit of delayed CEA.