Recurrent stroke: Of the 795,000 people who suffer an acute stroke in the U.S. each year, 185,000 are recurrent (Circulation 2023;147:e93-621). Prognosis of disability, physical and intellectual morbidity, and mortality rates are higher with recurrent stroke (Stroke 1999;30:338-49). In a large meta-analysis that included 14,455 patients, predictors of recurrent stroke within one year of initial stroke are: hypertension, diabetes mellitus, atrial fibrillation, and coronary artery disease (J Clin Neurosci 2019;60:24-30). Optimization and secondary treatment and prevention of these comorbidities should be prioritized in patients undergoing elective surgery.

In perioperative patients, the risk of stroke can be difficult to predict. Overall, the risk of acute ischemic stroke after noncardiac, non-neurologic surgery does not appear high, ranging between 0.2%-1.9% (Anesthesiology 2009;110:231-8). However, the risk of silent cerebral infarcts may be underrepresented in this data. In the NeuroVISION study of 1,114 patients undergoing noncardiac surgery with preoperative and postoperative magnetic resonance imaging, 78 (approximately 7%) had new covert acute brain infarcts without any clinical symptoms (Lancet 2019;394:1022-9). Of those patients, 29 (42%) suffered cognitive decline within one year of surgery. Moreover, the presence of a covert stroke was associated with a higher risk of delirium and overt stroke or transient ischemic attack within one year. While these findings have not necessarily been translated to our patient population globally, the risk of new or recurrent stroke in the perioperative period should not be underestimated.

Withholding anticoagulation is an important determination in this patient population. Often these patients are on antiplatelet agents to prevent in-stent thrombosis or recurrent cerebrovascular thrombotic events. In nonoperative patients with atrial fibrillation, the risk of stroke increases with age, and continuing anticoagulants in patients who are older (>66 years) may need to be prioritized, though this has not been studied in the surgical population.

Perioperative neurocognitive disorders (PND): At baseline, patients with recent stroke may be at higher risk for cognitive dysfunction and decline after their cerebral injury. Several studies have shown a high prevalence of cognitive dysfunction after initial stroke, and there are large, prospective studies ongoing to identify the determinants of cognitive decline after stroke (Stroke 2021;52:e499-516). Patients with clinical or subclinical cognitive dysfunction are at higher risk for PND, including postoperative delirium, and short- and long-term cognitive decline.

Professional societies and brain health experts around the world have published several guidelines over the past decade. Recommendations from eight of these publications were summarized and compared in a recent scoping review (J Neurosurg Anesthesiol 2023;35:10-8). A few key considerations in the preoperative, intraoperative, and postoperative settings are highlighted below.

Preoperatively, we should begin by identifying patients who are at risk for stroke or perioperative neurocognitive disorders. A brief screening tool such as the Mini-Cog© can not only establish a baseline but also detect some cases of cognitive impairment that would otherwise go unnoticed. As part of the informed consent process, patients and their families should be counseled on their risk of stroke and PND. Providing a copy of the ASA’s Perioperative Brain Health Initiative handout encourages them to take an active role in safeguarding their brain health (asamonitor.pub/47m8GI8). For high-risk patients, it may be helpful to engage a multidisciplinary team before surgery and throughout the perioperative period.

Moreover, as recommended by the ASA Perioperative Brain Health Initiative, informed consent should be a pivotal part of the preoperative workup (Anesth Analg 2018;127:1406-13). Patients with a history of stroke have a high risk of underlying cognitive dysfunction or decline. Patients at risk of cognitive dysfunction should be informed of the risk of postoperative neurodysfunction after surgery, specifically postoperative confusion, inattention, and memory problems.

In patients with a recent history of stroke, the risk of perioperative stroke decreases over time but will never return to the baseline risk of those without prior stroke. Current guidelines recommend postponing elective surgeries for nine months after a stroke, based on data from a 2014 Danish national cohort study (J Neurosurg Anesthesiol 2020;32:210-26; JAMA 2014;312:269-77). That study reported an odds ratio of 24 for patients with a stroke three to six months before surgery, but the inclusion of low-risk patients in the cohort may have overstated the magnitude of the elevated risk. A recent study of over 5 million Medicare patients reported an odds ratio of 8 for patients with a stroke within 30 days before surgery and reported that risk leveled off after three months (JAMA Surg 2022;157:e222236). These data have not yet been incorporated into guidelines, so clinicians must weigh the risks and benefits of delaying surgery.

Intraoperatively, routine use of high-risk medications such as benzodiazepines and anticholinergics should be avoided for patients at risk of PND. In terms of anesthetic technique, there is no clear advantage of regional anesthesia as compared to general anesthesia techniques. Some guidelines recommend the use of processed EEG monitors to titrate anesthetic drugs, but recent randomized controlled trials have had mixed results (Anesth Analg 2018;127:1406-13). With regard to stroke risk, normocapnia is preferred. In terms of hemodynamics, common sense would argue for the avoidance of hypotension. Studies to date have not identified precise blood pressure targets to decrease the risk of overt stroke, let alone covert stroke.

Best practices for the postoperative period begin with routine screening. For stroke, any new focal neurological deficits should prompt emergent brain imaging. For delirium, screening should be performed with a validated instrument such as the CAM-ICU (Confusion Assessment Method for the ICU) or the Nu-DESC (Nursing Delirium Screening Scale). When possible, hearing aids and glasses should be available to patients in the recovery room. Throughout the postoperative period, frequent orientation, early mobilization, adequate nutrition, and sleep hygiene are beneficial. Systems-based interventions such as the Hospital Elder Life Program (HELP) have a strong evidence base and may also improve outcomes beyond brain health (Am J Geriatr Psychiatry 2018;26:1015-33).

There is an exponential increase in the use of glucagon-like peptide-1 receptor agonists (GLP-1RAs) in recent years due to their numerous clinical benefits.  However, GLP-1RAs delay gastric emptying with recent reports of aspiration of gastric contents during sedation and general anesthesia despite traditional, preoperative fasting requirements. The optimal approach to mitigation of aspiration risk remains controversial because of sparse and inconsistent evidence. In 2023, the American Society of Anesthesiologists (Schaumburg, Illinois) published guidance for management of patients on GLP-1RAs with the primary aim of informing clinicians and patients of the heightened risk of aspiration and emphasizing shared decision-making.  Subsequently, other professional societies have also published recommendations for perioperative or periprocedure management of patients on GLP-1RAs.

A multisociety consensus for the management of patients on GLP-1RAs was developed by representatives from the American Society of Anesthesiologists, the American Gastroenterological Association (Bethesda, Maryland), the American Society for Metabolic and Bariatric Surgery (Gainesville, Florida), the International Society of Perioperative Care of Patients with Obesity (Lynnwood, Washington), and the Society of American Gastrointestinal and Endoscopic Surgeons (Los Angeles, California).  This practice guidance emphasizes that the approach to managing patients on GLP-1RAs should be based on shared decision-making of the patient, the prescribing care team, the proceduralist or surgeon, and the anesthesiologist.  One of the first steps is preprocedure assessment of aspiration risk, performed with enough time in advance to identify patients at risk and allow for adjustments in periprocedural care if needed. Factors that may increase the risk of aspiration include the escalation phase of dosing, long-acting GLP-1RAs, higher doses, the presence of gastrointestinal symptoms of delayed gastric emptying (e.g., nausea, vomiting, and abdominal discomfort), and medical conditions other than GLP-1 RA usage that might delay gastric emptying. 

The multisociety clinical practice guidance suggests that patients without risk factors may continue GLP-1RA therapy as usual in the perioperative period. Bridging therapy is not a viable option for most patients. For patients identified with an increased risk profile, use of a liquid diet for at least 24 h before the procedure to decrease the risk of retained gastric contents on the day of the procedure (as usually performed in patients undergoing colonoscopy and bariatric surgery) allows for the continuation of GLP-1RA therapy in the perioperative period.  If an unacceptable safety profile exists to continue GLP-1RA, patients may be asked to withhold therapy; however, this should be balanced with the risk of inducing a metabolic disease state, like hyperglycemia, that may complicate postprocedure patient care. The duration of holding should follow the guidance of the American Society of Anesthesiologists consensus-based guidance (i.e., holding the day of surgery for daily formulations and a week before surgery for weekly formulations).

On the day of procedure, if GLP-1RAs are continued, patients should again be assessed for symptoms suggestive of delayed gastric emptying. A point-of-care gastric ultrasound could also be considered to assess for retained gastric contents but presents technical and logistical difficulty. For patients determined to be at higher risk of aspiration on the day of procedure, rapid sequence induction of general anesthesia may be considered, if appropriate. These recommendations apply to all patients receiving GLP-1RAs, irrespective of the indication for GLP-1RA therapy.

In summary (table 1), this multisociety consensus provides guidance for management of patients on GLP-1RAs; however, it is not an evidence-based guideline. This approach emphasizes shared decision-making and provides recommendations for balancing continuation of GLP-1RA therapy perioperatively for surgery and procedures while minimizing the risk of aspiration.