Preoperative fasting for patients who take GLP-1 receptor agonists: should we Sip Til Send?

Background: GLP-1RA medications and their clinical use

This article relates to patients taking glucagon-like peptide-1 receptor agonists (GLP-1RAs) and dual GLP-1/glucose-dependent insulinotrophic peptide receptor co-agonists (GLP-1/GIPRAs), hereafter collectively referred to as GLP-1RA medications. GLP-1RA medications have been prescribed in clinical use for more than 15 years [2], initially for the treatment of type 2 diabetes, then subsequently approved for the management of obesity in 2014 [3]. Their use has since expanded rapidly, even more so following evidence of cardiac and renal protective effects [3, 4].

Given these benefits, GLP-1RA medications play an evolving perioperative role in optimising comorbidities, with evidence also suggesting a reduction in postoperative complications [2, 5]. Wu and colleagues [6] found an associated reduction in relative risk (RR) of all-cause mortality (RR: 0.38) and major adverse cardiovascular events (RR: 0.58) favouring patients taking GLP-1RA medications. Another retrospective study found reduced risk of readmission, wound dehiscence and haematoma [7].

Mechanism: GLP-1RA effect on gastric emptying

GLP-1RA medications work via multiple mechanisms. They promote insulin secretion and glucagon inhibition, slow gastric emptying, reduce postprandial glycaemic increases and suppress appetite [8]. They affect gastric emptying through activation of the GLP-1 receptor that inhibits vagal activity in the stomach [9]. Conversely, glucose-dependent insulinotropic peptide (GIP) has little, if any, effect on gastric emptying [9].

What the gastric emptying evidence shows

Patients who take GLP-1RA medications have delayed gastric emptying [9]. Data from multiple trials show increased incidence of residual gastric contents (RGC) in these patients after fasting before anaesthesia [10, 11, 12]. However, the exact nature of the relationship between RGC and pulmonary aspiration risk remains unquantified [13]. In particular, it is uncertain what gastric volume (GV) threshold should be considered a marker of significantly increased pulmonary aspiration risk. A GV of >1.5 ml/kg body weight is currently considered ‘at risk’, although the actual GV threshold that increases risk remains unknown and might even be larger [13]. Multiple factors contribute to the occurrence of pulmonary aspiration, making a specific GV threshold harder to define. Furthermore, ultrasound findings of a ‘full stomach’ occur in ~3–5% of fasted patients even though pulmonary aspiration is rare [13].

It is important to acknowledge that the digestive phases of solids and liquids in the stomach are different. Liquid intake does not significantly affect the rate of solid emptying [14, 15]. Evidence regarding gastric emptying of liquids in patients who take GLP-1RA medications is conflicting and limited [9]. A systematic review found minimal effect of GLP-1RAs on liquid-phase emptying, with acetaminophen-based absorption testing (AAT) showing a small delay at 1 h but not at 4 or 5 h [16]. Authors considered this delay of questionable clinical significance, with no substantial differences in liquid emptying overall. A major limitation was that the AAT indirectly measures GV and is considered unreliable [9]. A subsequent study measured liquid emptying using ‘gold standard’ scintigraphy [17]. This showed delayed emptying of a 300-ml glucose drink in patients taking lixisenatide (a GLP-1RA) vs placebo. In the lixisenatide group, the median volume of drink remaining was 276 ml (92% of the initial) at 120 min after intake. The median time to 10% gastric retention was 307 min (range: 204–880). Authors concluded a fast of 2–4 h before a procedure is almost certainly insufficient. Another small study used gastric ultrasound to assess GV after an overnight fast. Participants then drank 350 ml water and fasted another 2 h before GV was reassessed [18]. More than half of the patients in the GLP-1RA group had ‘unsafe’ RGC at both timepoints, with solids being found more frequently than clear fluids only. Overall, this limited available evidence suggests there is some delay in clear fluid emptying in patients who take GLP-1RA medications.

From gastric volume to pulmonary aspiration risk

The evidence base on the interplay between increased RGC from GLP-1RA medications and perioperative pulmonary aspiration risk remains limited and conflicting. In 2023, case reports and other studies raised concerns of a link between GLP-1RA medications and pulmonary aspiration under anaesthesia [19, 20, 21, 22]. In 2024, a large retrospective health insurance database review found a small, statistically significant increase in the incidence of aspiration pneumonia in patients undergoing endoscopies who take GLP-1RAs, compared with those not taking these medications [23]. In the general theatre setting, a single-centre report also showed an increase in aspiration rate [24]. Other studies, including multiple systematic reviews, have not found such a link [25, 26, 27, 28, 29, 30, 31, 32]. Conversely, a recent large retrospective study found a reduction in pulmonary aspiration risk [6]. This may reflect an increased use of aspiration mitigation measures amongst practitioners.

Sip Til Send: protocol and the premise

Sip Til Send is a clear fluid fasting protocol that allows patients to drink specific volumes of clear preoperative fluids until call to theatre [1]. ‘Think Drink’ is a similar drink-until-call protocol [33]. The basis of these protocols is the rapid gastric emptying of liquids in healthy adults. Clear liquids such as water or clear carbohydrate drinks empty from the stomach in an exponential fashion, with a half-time of 10–20 min [14, 33]. Because of this, if clear preoperative fluids are consumed judiciously until call, a patient’s GV is unlikely to exceed 1.5 ml/kg body weight at anaesthetic induction.

Sip Til Send and other drink-until-call protocols are associated with a significant reduction in fasting time compared with ≥2 h clear fluid fasting protocols [33, 34]. This is because protocols that are fixed to the timing of surgery often result in prolonged fasting due to uncertainty of theatre scheduling. Benefits of Sip Til Send over ≥2 h fasting protocols include reduced rates of postoperative nausea [35, 36, 37], reduced preoperative thirst [35, 37] and improved patient experience [36, 37, 38]. Reduced fasting time also reduces the risk of adverse complications from dehydration [39, 40].

Does the Sip Til Send premise hold for patients taking GLP-1RA medications?

As already outlined, available evidence suggests that GLP-1RA medications could prolong gastric emptying of clear fluids. Therefore, patients who take these medications should not be considered fully fasted if they have undergone recent fluid intake (<2 h). However, clinicians must be aware of the pitfalls of using more conservative ≥2 h clear fluid fasting protocols and mitigate the risk of dehydration. This could be addressed with careful fasting timing, or administration of intravenous fluids. Patients who are more susceptible to the effects of dehydration or derangements in blood glucose must not be left nil by mouth (NBM) indefinitely.

How do major guidelines navigate preoperative fasting for patients taking GLP-1RA medications?

Several guidelines exist worldwide that advise on the perioperative management of patients taking GLP-1RA medications [41]. Most do not recommend routinely withholding these medications preoperatively [41]. This reflects the deleterious effect of interrupting these drugs on comorbidity management, regardless of the drug’s indication. Other aspects of perioperative advice vary significantly between the guidelines, likely reflecting the limited evidence base.

The current UK consensus guidelines on perioperative management for adults taking GLP-1RA medications [5] recommend using standard fasting guidelines, with risk assessment and other safety strategies before, during and after anaesthesia. Point-of-care gastric ultrasound is endorsed as an optional tool for risk stratification; however, current variability in training may limit its broader application. The guidelines also recommend discussing risks and mitigation strategies with patients. Anaesthetists may be less familiar with the use of shared decision-making in the context of fasting. However, it is pertinent to ensure the patient is aware of – and involved in – the decisions that shape their care, particularly when equipoise exists.

The UK Guidelines do not directly address Sip Til Send, but by suggesting ‘standard fasting guidance’ they are likely to refer to the recognised Sip Til Send fasting guideline from the CPOC [1]. The CPOC guidelines endorse a Sip Til Send approach but list gastric immobility or class III obesity as a reason for anaesthetists to potentially use a ≥2 h clear fluid fasting protocol instead.

Contrary to the UK guidance, some other guidelines recommend a 24 h preoperative clear fluid diet (24 h CFD) for patients taking GLP-1RA medications [41]. This is based on derived evidence suggesting pulmonary aspiration risk is comparable with the general population when this approach is used [42]. In Australia and New Zealand, the 24 h CFD includes a final 6 h fasting period with clear fluids permitted until 2 h preoperatively, then NBM thereafter [42]. In the absence of additional concerns, standard airway management may then be considered appropriate, as would be chosen for a patient without other risks. This approach aims to avoid ‘unnecessary’ rapid sequence intubations; however, it has been criticised for the risks and wellbeing impacts associated with prolonged fasting [43]. Sip Til Send is less feasible using the 24 h CFD, as fluid intake within 2 h of anaesthesia would prompt additional risk mitigation, potentially negating the intended reduction in additional airway precautions required.

Where does this leave clinicians?

Using the UK recommendations, the first step for our patient is to assess their risk factors including, but not limited to, feelings of gastric fullness, other comorbidities that affect gastric emptying and how recently the GLP-1RA medication was commenced or uptitrated. The evidence and recommendations would then be discussed with the patient. If the patient is not deemed to be at high aspiration risk, Sip Til Send may be reasonable, particularly when combined with preoperative gastric ultrasound and/or prokinetic medication, in addition to airway protection measures such as a rapid sequence intubation, unless gastric ultrasound clearly demonstrates a low-risk GV.

An appropriate alternative is to keep the patient NBM and consider commencing intravenous fluids. This approach may be prioritised in patients with multiple additional risk factors for gastroparesis, again utilising further airway risk mitigation measures tailored to the clinical context. Keeping the patient NBM may also be preferred for those who have already undergone a 24 h CFD in order to allow usual airway protection measures.

In summary, the evidence that guides fasting approaches for patients taking GLP-1RA medications remains insufficient, with equipoise on the ideal approach. All patients require a comprehensive risk assessment and fasting plan, ideally well in advance of their surgery date. These plans should be made using a shared decision-making approach, prioritising safety and wellbeing. Sip Til Send may be appropriate in select patients who take GLP-1RA medicines; however, given the limited evidence, these patients should not be considered fully fasted if they have had a clear drink within 2 h.

Research ethics statement

Not applicable to this article.

Consent for publication statement

This narrative article is based on a fictional clinical vignette and does not describe a real patient or clinical encounter so consent for publication is not necessary. Images are used solely for illustrative purposes and are not related to the vignette. Where real patient images are included, the author confirms that written informed consent for publication has been obtained for this purpose in accordance with the journal’s patient consent requirements.

Conflicts of interest statement

The author declares no conflicts on interest with this article.