GLP-1 drugs: A gastric ultrasound signal is not an aspiration rate

GLP-1 medication creates two distinct perioperative questions: whether gastric contents are more likely to remain despite fasting, and whether clinically important aspiration becomes more frequent. These questions require different evidence. An imaging study can identify a physiologic concern without measuring the complication itself; a large administrative cohort can examine a recorded complication while missing the precautions clinicians took to prevent it. The growing literature is most useful when those differences remain visible.

The original ultrasound signal

Sen and colleagues prospectively enrolled 124 patients presenting for elective procedures after guideline fasting. Increased residual gastric content appeared in 35 of 62 weekly GLP-1 receptor agonist users, compared with 12 of 62 nonusers: 56% versus 19%. The adjusted prevalence ratio was 2.48 (95% CI, 1.23–4.97). The ultrasound definition included solids, thick liquids, or more than 1.5 mL/kg of clear fluid. This was a cross-sectional association, using convenience sampling, rather than a randomized medication experiment. Ultrasound findings were disclosed to clinicians, and data collection stopped before subsequent airway decisions and cancellations were captured. The study therefore supplies no aspiration rate and no validated interval that reliably normalizes risk. Sen et al., Methods, Results, and Limitations.

That distinction changes how the numbers should be communicated. Saying that more than half of exposed patients had increased gastric content describes the measured imaging outcome. Saying that more than half were likely to aspirate would substitute an unmeasured event for that outcome. Even the fluid-volume threshold is an operational research definition, rather than a patient-specific probability of harm.

What the larger surgical cohort adds

Chen and colleagues used commercial insurance claims for 366,476 adults undergoing 14 common operations. They classified preoperative exposure using prescriptions and assessed aspiration pneumonia during the following 30 days. The adjusted odds ratio was 0.78 (95% CI, 0.57–1.06), without a statistically significant association. Claims cannot establish what patients actually took immediately before surgery, their fasting adherence, gastric ultrasound results, or case cancellations. Patients whose operation never proceeded also do not contribute a postoperative event. Chen et al., Methods, Results, and Limitations.

The cohort is reassuring about the outcome it could observe, but it cannot isolate the medication's effect from the surrounding system of care. The ultrasound and claims findings can coexist: residual content, clinical precautions, and a coded pulmonary complication occupy different positions along the same causal pathway. Neither estimate should be used to erase the other.

New randomized evidence in 2026

The OCULUS trial randomized patients undergoing upper endoscopy to continue treatment or hold one dose. At its prespecified interim analysis, clinically significant residual gastric volume occurred in 1 of 32 patients assigned to hold and 7 of 28 assigned to continue: 3.1% versus 25.0%, an absolute difference of 21.9 percentage points (90% CI, 7.0–36.7). The trial stopped early after crossing its stopping boundary. Its composite included contents that prevented examination, required termination or intubation, or produced an aspiration event requiring additional care. Participants were on stable doses; planned general anesthesia and documented gastroparesis were among the exclusions. This small endoscopy trial does not establish a surgical aspiration-prevention rule. The article's visual abstract and Figure 2 labels were corrected in May 2026; the corrected text supplies these group assignments. OCULUS, corrected report.

Another 2026 prospective study, RESIDUAL, found increased ultrasound content in 42% of patients who had held weekly medication for at least seven days and 24% of nonexposed controls. Its adjusted prevalence ratio was 1.65 (95% CI, 0.7–3.7). The uncertainty spans no difference, and the study does not compare randomized withholding durations. It adds caution against treating a seven-day hold as proof of an empty stomach. RESIDUAL, original abstract.

Professional recommendations differ in meaningful ways

The 2024 multisociety guidance supports shared decisions among the procedural, anesthesia, and prescribing teams. Patients without elevated risk may continue treatment. Escalation, dose, gastrointestinal symptoms, and other causes of delayed emptying inform assessment; a liquid diet for at least 24 hours, gastric ultrasound, and changes to the anesthetic plan are options when concern exists. The document identifies itself as clinical guidance developed amid limited evidence. Multisociety guidance.

SPAQI's 2025 consensus also recommends continuation in patients without significant gastrointestinal symptoms, but pairs it with broader fasting changes: 24 hours without solids, eight hours for clear liquids containing at least 10% glucose, and four hours for lower-carbohydrate clear liquids. These are SPAQI recommendations for its defined population, with differing evidence grades, rather than a universal fasting prescription adopted by every society. SPAQI, Box 1.

The ADA's 2026 hospital standards recommend individualized perioperative management considering indication, glycemia, drug characteristics, symptoms, urgency, and anesthesia. They also address an alternative glycemic strategy if withholding is chosen and hyperglycemia is anticipated. This diabetes-focused context matters when balancing medication interruption against metabolic consequences. ADA, perioperative GLP-1 section. ASA's official patient information, medically reviewed in October 2025, likewise says most patients can continue and emphasizes coordinated instructions and additional precautions for higher-risk situations. ASA patient guidance.

Sen reported no conflicts and NIH support for one investigator. OCULUS received Cleveland Clinic and ASGE funding; an investigator disclosed outside Novo Nordisk consulting, among other relationships. These disclosures warrant transparency without determining the results' validity. Sen disclosures, OCULUS disclosures.

For perioperative teams, the useful endpoint is a coherent plan that identifies who reconciles medication, diet, glucose management, and procedural decisions. A research headline cannot supply that ownership. The evidence now includes randomized endoscopy data as well as imaging and claims studies, but still requires clinicians to match each finding to its population and outcome before applying it.