Perioperative Stroke in Adult Noncardiac Surgery: Evolving Controversies Across the Perioperative Continuum

Perioperative stroke after adult noncardiac surgery is uncommon but associated with substantial mortality and disability. In a cohort of 523,059 noncardiac, non-neurologic procedures, clinically apparent stroke occurred in 0.1% and was associated with an approximately eightfold increase in 30-day mortality. A separate prospective study detected covert cerebral infarction on postoperative MRI in 7% of older adults. Preoperative assessment can identify prior stroke, atrial fibrillation, vascular disease, frailty, hypertension, and diabetes, informing surgical timing, management of established cardiovascular and metabolic conditions, antithrombotic plans, and postoperative surveillance. This narrative review examines the evidence and remaining controversies across the perioperative continuum. Ischemic stroke may result from embolism, in situ arterial thrombosis, or hypoperfusion; hemorrhagic stroke has distinct mechanisms. Observational studies suggest that excess stroke risk after a recent ischemic stroke declines substantially over the first 3 months, but the benefit of longer surgical delay remains uncertain. Chronic clinically indicated β-blockers should generally be continued, whereas high-dose initiation immediately before surgery may increase hypotension and stroke. Routine heparin bridging is unsupported for most patients requiring anticoagulant interruption, with limited evidence in those at very high thromboembolic risk. Blood pressure trials have not established a stroke-specific target, and no anesthetic technique has demonstrated universal superiority for stroke prevention. Residual anesthesia and delirium can delay postoperative recognition, supporting repeated neurological assessment and urgent evaluation of new deficits or unexpected recovery. Risk-adapted care is warranted, but dedicated stroke-focused trials are needed to determine which interventions improve neurological and functional outcomes.