(IP205) Risk Score for the Event of Perioperative Myocardial Infarction at the Time of Carotid Endarterectomy

Objectives: The purpose of this study is to create a risk score for the event of perioperative myocardial infarction (MI) following carotid endarterectomy (CEA).

Methods: The VQI CEA module was queried between January 2003 and October 2023 and 192,547 met study inclusion. An internal VQI validation cohort was similarly created with the same exclusion criteria utilizing CEA performed between November 2023 and October 2024 over which time period 17,449 met inclusion. The primary study outcome was perioperative MI at CEA. Univariable analysis was conducted followed by binary logistic regression analysis utilizing significant univariable factors. Regression beta-coefficient was used to created a weighted risk score for MI and internal validation with testing at each risk score was conducted. Mortality rate in long term follow up for those with and without MI was investigated.

Results: The following factors had a significant (P <.05) multivariable association with perioperative MI for CEA: female sex; advancing age; rural home status; diabetes; history (Hx) of CAD; MI or angina pectoris within 6 months of surgery; CABG greater than 5 years ago; CHF regardless of severity; renal insufficiency; positive stress test within two years; anemia; Hx of PAD intervention; prior CEA or carotid artery stenting; dual antiplatelet therapy at time of presentation; modified Rankin score of 2 or greater at time of CEA; and urgent/emergent endarterectomy. Not having had a Hx of prior MI in combination with having no current CAD symptom was protective (P <.01) versus perioperative MI. There was significant escalation noted with increasing risk score as patients with scores of 5 or less experienced MI in just 0.2% of cases whereas patients with risk scores greater than 25 experienced MI at a 20 times higher rate of 4.1% (Figure 1). AUC analysis for the risk score had a value of 0.70. Application of the risk score to the validation cohort resulted in a similar AUC value of 0.72.

Conclusions: A risk score for the event of perioperative MI at the time of CEA has been developed that has good accuracy. The variables with the strongest association with development of MI include known history of CAD, positive cardiac stress test withing two years, MI or angina within six months, CABG greater than 5 years prior, PAD, renal insufficiency, rural home status, urgent/emergent operation and female sex. Patients experiencing perioperative MI have a significantly increased 5 year mortality rate relative to those without (Figure 2).