The fasting and the furious: reconciling fasting guidelines with glucagon-like peptide-1 receptor agonists, ‘Sip-til-Send’ policies and gastric ultrasound

Abstract

Pre-operative fasting recommendations have been a cornerstone of safe anaesthetic care for nearly 150 years. While many international guidelines for elective surgery stipulate a 6-hour fasting period for solid food and 2 hours for clear fluids, this is often exceeded, with patients inadvertently fasting for much longer [1]. Prolonged fasting can adversely affect patient well-being, causing thirst, nausea and metabolic disturbances which contribute to patient discomfort and anxiety [2]. The classic ‘nil by mouth from midnight’ mantra, rooted in Mendelson's seminal research on chemical pneumonitis in healthy obstetric patients under anaesthesia, is embedded in today's high-throughput surgical settings, and there remains a tendency to continue with conservative fasting practices [3].

However, many clinicians now recognise the benefits of reducing fasting times for clear fluids to under 2 hours. More liberal protocols, such as the ‘Sip-til-Send’ policy, where patients may drink up to 170 ml of clear fluid per hour until called for theatre, are gaining traction [4]. Developed by a team at NHS Tayside and endorsed by the Centre for Peri-operative Care, this approach reduces cognitive load for pre-operative staff through a clear ‘cut-off’ and improves patient comfort. In Australia, similar policies stipulate a 200 ml.h⁻¹ limit [5].

Nevertheless, not everyone is prepared to proceed at full speed due to concerns about the risk of aspiration. In the UK Royal College of Anaesthetists' 7th National Audit Project (NAP7), the aspiration or regurgitation incidence in the non-obstetric population undergoing general anaesthesia or sedation was 1 in 698 cases [6]. In comparison, previously published large cohort studies report a pooled incidence of 1 in 2977 [7-9], though this excludes regurgitation events that did not result in aspiration. The vast majority of patients do not experience aspiration events around the time of surgery. How can we then balance the need to avoid excessive fasting with identifying patients at higher aspiration risk?

Gastric ultrasound has emerged as a sleek, real-time, non-invasive tool for assessment of gastric contents. It is accurate in both adults and children for detecting solid content and estimating clear fluid volumes. A gastric antral volume under 1.5 ml.kg⁻¹ is considered to represent low risk for aspiration in healthy patients [10]. Moreover, studies using gastric ultrasound show clear fluids empty rapidly and may even boost gastric motility [11]. Similar findings have been reported in patients with diabetes, obesity and in pregnancy, reinforcing confidence in liberalised pre-operative clear fluid policies [12-14]. Despite adherence to fasting guidelines, a minority of elective patients will have residual gastric content due to various risk factors, and gastric ultrasound allows clinicians to steer their airway management plans accordingly.

Certain medications are known to delay gastric emptying, including anticholinergics, calcium channel blockers, opioids and tricyclic antidepressants. Recently, there has been increased attention on the effect of glucagon-like peptide-1 receptor agonists (GLP-1 RAs) on gastric emptying. Initially, developed for the management of type 2 diabetes mellitus, GLP-1 RAs are becoming increasingly popular as weight loss drugs. These medications delay gastric emptying, reduce appetite, increase satiety and deliver a range of cardiovascular and neurological benefits [15, 16]. Studies have reported an association between the use of these medications and a reduced risk of periprosthetic joint infection in hip and knee arthroplasties [17], lower mortality in shoulder arthroplasty [18] and reduced risk of postoperative wound dehiscence and readmission [19]; indicating future directions for research on the impact of these medications in the peri-operative period.

The popularity of GLP-1 RAs has surged, fuelled by social media endorsements and expanded indications, leading to global supply shortages. However, their role in increasing residual gastric content, even in adequately fasted patients, has sparked serious concern. Multiple case reports of regurgitation or aspiration have been published, but commentary has often lacked nuance, with little distinction between general anaesthesia and sedation, or endoscopic and surgical procedures.

Two recent meta-analyses merit further discussion. Elkin et al (2025) evaluated 18 studies, grouping endoscopic and surgical patients together, showing a higher incidence of residual gastric contents in fasted patients taking GLP-1 RAs compared to controls (OR 5.96, 95%CI 3.96–8.98) [20]. Another pre-print (not peer-reviewed) meta-analysis separates the presence of residual gastric contents by procedure [21]. In over 9000 upper gastrointestinal (GI) endoscopy patients taking GLP-1 RAs from 25 studies, residual gastric content was found in 9.5%, with a higher rate in upper endoscopy-only patients (12.8%) compared to those also undergoing colonoscopy (4.3%). The turbo-charged combination of a laxative preparation, low-fibre then clear liquid diet and prolonged fasting required for colonoscopy may account for this observation. Notably, in upper GI endoscopy, solid gastric residue is directly visualised while large fluid volumes are often suctioned without being flagged as residual. Deep sedation without definitive airway protection, common in many countries, adds another layer of risk. Ultrasound-based studies in non-endoscopic subjects tell a more cautionary tale. Even after fasting, those taking GLP-1 RAs showed a pooled incidence of 50.5% with residual gastric content (98/194) versus 8.5% (17/200) in controls [21]. The question still remains: does more content translate to more aspiration?

Elkin et al (2025) evaluated nine studies, including over 185,000 patients (six endoscopic and three surgical studies), showing no difference in aspiration rates between patients taking GLP-1 RAs and controls [20]. The preprint pooled data from over 450,000 upper GI endoscopy patients shows an aspiration rate of 17.5 per 10,000 GLP-1 RA users versus 13.6 per 10,000 controls [21]. We performed a simple 2 × 2 table chi-squared analysis for comparison, generating a p-value of 0.0006, while noting that two of the 15 studies used aspiration pneumonitis as their end-point. Data from eight surgical studies [21-23], however, show no increased risk. A problem with research in this area is that in order to adequately power a study for aspiration, fortunately a rare event, the required sample sizes are very large. Aspiration risk is estimated at 1 in 900 to 1 in 10,000 [7], depending on risk factors. Assuming a baseline incidence of 1 in 3000 [7-9], a study would require a sample size of 622,748 patients to detect a 20% increase in aspiration incidence (α value 0.05, power 80%), while a 10% increase would require over 2.4 million patients. As a result, much of our current evidence is from retrospective datasets. There are inherent biases in these data. For example, the inability to complete upper GI endoscopy may be used as a surrogate for retained gastric content, but is operator dependent. Others rely on the diagnosis of aspiration pneumonitis to determine aspiration incidence.

Most analyses do not take indication of GLP-1 RAs into account and do not distinguish between patients using these medications for diabetes, obesity or both. Higher doses are used for weight loss, and diabetes itself affects gastric motility unpredictably. While diabetic gastroparesis affects 40% of patients living with diabetes, 20% have rapid gastric emptying [24]. A recent study showed no difference in baseline gastric volume between fasted patients with diabetes and those without diabetes [25]. Stopping these drugs impairs glycaemic control in patients with diabetes [26], yet we do not know the impact of a short-term pause for the peri-operative timeframe. Withholding at least one dose before a procedure reduces the incidence of residual gastric content, though nowhere near acceptable levels (44% incidence in the surgical population) [20]. Delayed gastric emptying peaks in the first few weeks of therapy with long-acting versions of the drug, though the risk remains elevated compared to non-users even after tachyphylaxis has occurred [27].

Despite strong evidence that GLP-1 RAs delay gastric emptying, pre-procedural recommendations have been variable, with disagreement on decisions to withhold drugs, fasting times, pre-operative diet modification, postponement or cancellation of cases, and type and induction of anaesthesia [28]. A recent multi-specialty consensus statement outlines the different variables and risk factors which should be considered when managing these patients, including drug type and dose, impact of cessation, patient factors (co-morbidities, fasting status), urgency and nature of procedure, risk mitigation interventions, anaesthesia technique and potential outcomes [29]. This new guidance advises the continuation of all GLP1-RAs peri-operatively [29], a view echoed in Australasian guidelines [30]. This strategy avoids glycaemic destabilisation and the inconvenience of retitrating doses, especially given the drugs' long half-lives.

However, recommendations around fasting durations vary. Multidisciplinary American and Australasian consensus statements advocate a 24-hour liquid diet pre-operatively [30, 31], while UK bodies recommend routine fasting [29]. Prokinetic agents, such as erythromycin administered 60–120 min before induction, have been proposed [29, 30], though direct evidence in patients taking GLP-1 RAs is currently limited to a single recent case report [32]. This is unfamiliar terrain, with a developing evidence base and new directions beginning to emerge.

So where do ‘Sip-til-Send’ policies fit in? No studies have yet explored how GLP-1 RAs affect gastric emptying when fluids are consumed continuously in the pre-operative period. If emptying of clear fluids is impaired, is it of clinical significance? Should we revise volume limits? If GLP-1 RA patients are to be considered unfasted until proven otherwise, does it matter if they sip water in the interim? A survey of 38 Australasian hospitals found 69% of sites still included patients taking GLP-1 RAs in their ‘Sip-til-Send’ policies [5].

Consensus guidelines agree on the clinical utility of gastric ultrasound for immediate risk assessment in patients taking GLP-1 RAs [29-31]. It gives clinicians a diagnostic tool to assess the volume of gastric contents, one of the key factors, allowing safer and more individualised care. But like any high-performance tool, peri-operative point-of-care ultrasound requires training and access to mentorship for competence development. The American Society of Regional Anesthesia and Pain Medicine recommends a minimum of 30 supervised scans, with 50 scans targeted to achieve competence [34]. Australasian guidelines recommend 30 supervised scans or 15 scans for those with existing ultrasound competence and a provision for remote supervision [35]. With GLP-1 RA use accelerating, gastric ultrasound training may need to be included on the list of essential anaesthetic competencies.

Pre-operative fasting is shifting gears and guidance must adapt to the conditions of modern clinical practice, including the use of GLP-1 RAs and liberalised fluid intake policies. As ‘Sip-til-Send’ gains traction and the use of GLP-1 RAs continues to grow, clinicians must reconsider rigid strategies and adapt to more flexible patient-centred care. Gastric ultrasound offers the real-time telemetry we need to navigate aspiration risk, outperforming outdated ‘one-size-fits-all’ approaches, to arrive at individualised risk assessment which guide us to appropriate airway techniques. Whether for redefining fluid intake thresholds, assessing GLP-1 RA effects or guiding prokinetic use, it is poised to lead future practice.