A case series of incidents reported to webAIRS relating to glucagon-like peptide 1 agonist use

Abstract

The increasing prescription of glucagon-like peptide 1 (GLP-1) agonists presents a peri-operative challenge for anaesthetists. These drugs delay gastric emptying, potentially increasing pulmonary aspiration risk. Despite recent recommendations from the Australian and New Zealand College of Anaesthetists, Australian Society of Anaesthetists, and American Society of Anesthesiologists, there remain no formalised guidelines regarding preoperative optimisation for patients taking GLP-1 agonists. Given the current lack of evidence, we present a case series of incidents involving patients treated with GLP-1 agonists reported to webAIRS, a web-based anaesthetic incident reporting system in Australia and New Zealand. Among 11,700 reports from July 2009 to April 2024, 13 incidents related to GLP-1 agonist use were identified, including seven cases concerning aspiration. Patient factors potentially contributing to increased aspiration risks in these incidents included a higher body mass index, type 2 diabetes mellitus, gastro-oesophageal reflux disease, emergency case, and time from GLP-1 agonist cessation. Most incidents involved a rapid sequence induction for general anaesthesia with no cases utilising gastric ultrasound, preoperative nasogastric tube insertion, or intravenous erythromycin use. This case series highlights the potential concerns relating to GLP-1 treatment in the perioperative setting. Interdisciplinary collaboration and communication between anaesthetists, surgeons, general practitioners and endocrinologists are required to further investigate and establish preoperative guidelines for safe GLP-1 agonist use. Specifically, consideration in determining actual gastric contents of each patient despite generic fasting guidelines is needed. Early preoperative risk stratification should also improve patient safety and outcomes.

Introduction

With approximately 1.2 million Australians diagnosed with type 2 diabetes mellitus (T2DM) and a rising trend, there is an increasing reliance on pharmacotherapy for glycaemic management.^1,2 Among these medications are glucagon-like peptide 1 (GLP-1) agonists, including dulaglutide, liraglutide, semaglutide, exenatide and tirzepatide, which have gained popularity. Particularly relevant in T2DM, these medications have demonstrated efficacy in lowering glucose levels, inhibiting glucagon secretion and reducing appetite by delaying gastric emptying. ³ In Australia, these medications are given subcutaneously, either daily (liraglutide), twice daily (exenatide) or weekly (dulaglutide, semaglutide, tirzepatide).^4,5 Going forward, new GLP-1 agonists will likely be improved and released across Australia and New Zealand. Most recently, the Australian Therapeutic Goods Administration (TGA) has approved Mounjaro (tirzepatide), another weekly GLP-1 agonist with promising T2DM and weight loss results.^{5 –7}

A summary of currently available GLP-1 agonists in 2024 as approved by the Australian TGA can be found in Table 1. ^{5 –14}

Table 1.

Currently available glucagon-like peptide 1 agonists in 2024 approved by the Australian Therapeutic Goods Administration.^{5,7 –14}

Generic name	Brand/trade name	Common doses	Administration route	Administration frequency	Maximum concentration	Half-life	Approved date
Liraglutide	Victoza/Saxenda	1.2 or 1.8 mg	Subcut.	Daily	8–12 h	13 h	Aug 2010
Exenatide	Byetta	10  $\mathrm{\mu }\mathrm{g}$	Subcut.	Twice daily	4–8 h	2.4 h	Jul 2007
Semaglutide	Ozempic/Wegovy	0.5 or 1.0 mg	Subcut.	Weekly	24–72 s	Seven days	Aug 2019
Dulaglutide	Trulicity	1.5 mg	Subcut.	Weekly	24–72 h	Five days	Jan 2015
Tirzepatide	Mounjaro	2.5 mg	Subcut.	Weekly	24–48 h	Five days	Dec 2022

Subcut.: subcutaneous.

Furthermore, the off-label use of GLP-1 agonists has garnered significant interest and exposure amongst the wider community outside of T2DM therapy. This is in part associated with positive weight loss outcomes endorsed through social media.^{15 –17} There are few side-effects described, with the most common for T2DM patients on GLP-1 agonists including nausea and vomiting. ¹⁸

However, this medication class poses challenges for anaesthetists due to the effect on gastric emptying, thereby possibly increasing the risk of aspiration of gastric contents during anaesthesia.^{4,19 –21} Current fasting guidelines from the Australian and New Zealand College of Anaesthetists (ANZCA) recommend specific durations for preoperative fasting for healthy adults to reduce their perioperative aspiration risk. ANZCA recommends withholding clear fluids and solid food up to 2 and 6 h respectively preoperatively. ²² A recent consensus clinical practice recommendation has been published relating to periprocedural GLP-1 agonist use, supported by ANZCA, the Gastroenterological Society of Australia, Australian Diabetes Society and National Association of Clinical Obesity Services sharing expert opinion on this issue. ²³ Specifically, they recommend that all patients on GLP-1 agonists requiring anaesthesia within four weeks of their last dose should be considered as unfasted patients. ²³ They also acknowledge insufficient data available supporting cessation of GLP-1 agonists prior to anaesthesia for both non-endoscopic and endoscopic procedures. ²³ Overall, current literature remains unclear regarding when gastric emptying normalises after cessation of GLP-1 agonists. ²⁴ Limited literature and individual case reports also acknowledge the risk of unknown gastric contents in patients on GLP-1 agonists undergoing anaesthesia, which could negatively impact intra- and postoperative outcomes.^25,26 More recent literature by the Australian Society of Anaesthetists also highlights the need for a tailored approach that carefully balances the risks and benefits of continued therapy against procedural risk. ²⁷

Since September 2009, webAIRS, a bi-national web-based anaesthesia incident reporting system, has represented the largest voluntary, multicentre de-identified database for anaesthesia incident reporting across Australia and New Zealand. ²⁸ To date, it houses over 11,700 incident reports with the mission to improve patient safety via the widespread collation of self-reported anaesthetic incidents. ²⁹ Through anonymous self-reporting, it provides a secure database that represents the epidemiology of rare and potentially under-reported complications within the anaesthesia field. Data entry is performed voluntarily via non-mandatory text boxes, where the reporter can describe the incident in their own words and include reflection on contributing and alleviating factors. Additionally, non-mandatory drop-down boxes assist with clear categorisation of the incidents and demographic data collection. ²⁸

The aim of this analysis was to identify and review all incidents reported to webAIRS of patients treated with GLP-1 agonists.

Methods

As per PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines for the reporting of a review, ³⁰ a narrative analysis across all text entries was performed on the webAIRS database from its inception in 2009 to 25 April 2024.

A structured criteria search was done using the Boolean operator, ‘OR’ between the generic and brand names for the five GLP-1 agonists currently registered with the Australian TGA.^5,8 These names were dulaglutide (Trulicity), liraglutide (Saxenda), semaglutide (Ozempic), exenatide (Byetta) and tirzepatide (Mounjaro).^5,8 The specific searches conducted in the webAIRS database included ‘dulaglu’ OR ‘liraglu’ OR ‘semaglu’ OR ‘exen’ OR ‘tirze’ OR ‘GLP’ OR ‘glucagon’ OR ‘Truli’ OR ‘Saxen’ OR ‘Ozem’ OR ‘byet’ or ‘mounj’.

An eligibility criterion was also established. All incidents mentioning the patient taking a GLP-1 medication preoperatively were included in the analysis. Incidents were excluded if there was no patient case narrative relating to GLP-1 agonists discussed in the submitted incident report. Two authors, CYXL and YE, independently reviewed all reports to assess fitness for inclusion for this analysis. In the event of disagreement of whether the report met the inclusion criterion, author DYL was allocated as an independent third party arbitrator.

The data collection process involved extracting information based on the webAIRS specific fields, including patient demographics (age, gender, American Society of Anesthesiologists physical status grade (ASA PS), body mass index (BMI), surgical procedure type, anaesthesia type (general anaesthesia versus sedation)), preoperative investigations and any intraoperative complications.

The narratives of all incidents were subsequently evaluated by all authors. Additional information involving patient factors that potentially affect one’s baseline aspiration risk and possible event factors, including anaesthesia/operation type and techniques that influence gastric volume and/or motility, was analysed. Other factors including incident type, incident management and outcome were also reviewed in detail. Descriptive data analysis was then performed with figures described in counts and percentages.

The data collected by webAIRS is owned and funded by ANZCA, the Australian Society of Anaesthetists, and the New Zealand Society of Anaesthetists. ²⁸ Data was collected in compliance with the National Health and Medical Research Council recommendations for quality assurance data. The methods of de-identified data collection have been reported previously. ³¹

Multicentre ethics approval is maintained at the Royal Brisbane and Women’s Hospital Human Research Ethics Committee (HREC/11/QRBW/311), the Nepean Blue Mountains Local Health District (HREC/12/NEPEAN/18) and the Health and Disability Ethics Committee (MEC/09/17/EXP) in New Zealand.

Results

Thirteen incidents out of 11,700 reports to webAIRS were eligible to be included into the case series for review and analysis. One report was excluded as the narrative was only a comment, not an incident, in which the anaesthetist questioned the lack of guidelines regarding GLP-1 agonist use preoperatively. All incidents were reported between August 2023 and April 2024, representing more than one case per month.

Table 2.

Distribution of patient demographics for critical incidents from webAIRS related to glucagon-like peptide 1 agonist use.

ID	Age (years)	Gender	BMI	ASA PS	T2DM	GORD
1	60–69	F	>40	3	Y	NS
2	70–79	F	25–30	3	Y	NS
3	30–39	M	NS	2	Y	NS
4	60–69	F	>35	3	Y	Y
5	30–39	M	>40	2	Y	NS
6	60–69	M	>35	3	Y	NS
7	40–49	F	>40	3	NS	Y
8	40–49	F	25–30	3	NS	NS
9	30–39	F	NS	1	NS	N
10	70–79	F	25–30	3	Y	NS
11	50–59	M	>30	2	Y	Y
12	60–69	F	>40	3	Y	Y
13	50–59	F	>30	2	Y	NS

ASA PS: American Society of Anesthesiologists Physical Status; BMI: body mass index; F: female; GORD: gastro-oesophageal reflux disease; M: male; NS: not stated; T2DM: type 2 diabetes mellitus; Y: yes.

Table 3.

Glucagon-like peptide 1 agonist use, fasting status preoperatively, gastric ultrasonography and nasogastric tube insertion.

ID	Drug	GLP-1 ceased	Fasted >6 h	GI symptoms on day	Preoperative gastric ultrasound performed	Preoperative nasogastric tube inserted
1	Semaglutide	One week	NS	NS	NS	NS
2	Semaglutide	NS	Y	NS	NS	NS
3	Semaglutide	NS	Y	NS	NS	NS
4	Semaglutide	NS	Y	Y	NS	NS
5	Semaglutide	One week	NS	NS	NS	NS
6	Semaglutide	NS	Y	NS	NS	NS
7	Semaglutide	NS	NS	NS	NS	NS
8	Semaglutide	Two weeks	Y	NS	NS	NS
9	Semaglutide	One week	NS	Y	NS	NS
10	Semaglutide	24 h	NS	NS	NS	NS
11	Semaglutide	NS	NS	NS	NS	NS
12	Semaglutide	Not ceased	NS	N/A	N/A	N/A
13	Liraglutide	NS	NS	NS	NS	NS

GI: gastrointestinal; GLP-1: glucagon-like peptide 1; N/A: not applicable; NS: not stated; Y: yes.

Table 4.

Case summary of critical incidents from webAIRS related to glucagon-like peptide 1 agonist use.

ID	Surgery type	Initial plan	Actual procedure	RSI performed	Specific GLP-1 agonist	Complication	Time when event occurred	Muscle relaxant used	NGT preop.	Elective/emergency case
1	Urology	GA w/ETT	GA w/ETT	Y	Semaglutide	Acute pulmonary oedema (non-cardiac)	Emergence	Rocuronium	NS	Elective
2	Gastroenterology	Sedation	GA w/ETT	Y	Semaglutide	Aspiration risk	Induction	Suxamethonium	NS	Elective
3	ENT	GA w/ETT	GA w/ETT	NS	Semaglutide	Aspiration event	Induction	Rocuronium	NS	Elective
4	Gastroenterology	Sedation	Sedation	N/A	Semaglutide	Aspiration risk	Induction	Nil	NS	Elective
5	Neurology	GA w/ETT	GA w/ETT	Y	Semaglutide	Respiratory de-recruitment	Emergence	Rocuronium	NS	Elective
6	Gastroenterology	GA w/ETT	GA w/ETT	Y	Semaglutide	Aspiration risk	Induction	Suxamethonium	NS	Elective
7	Gynaecological	GA w/ETT	GA w/ETT	NS	Semaglutide	Awareness	Induction	Rocuronium	NS	Elective
8	Laparoscopic surgery	GA w/ETT	GA w/ETT	Y	Semaglutide	Aspiration risk	Emergence	Rocuronium	NS	Elective
9	Laparoscopic surgery	GA w/ETT	GA w/ETT	Y	Semaglutide	Anaphylaxis	Induction	Rocuronium	NS	Elective
10	Gastroenterology	GA w/ETT	GA w/ETT	Y	Semaglutide	Aspiration risk	Induction	Suxamethonium	NS	Emergency
11	Laparoscopic surgery	GA w/ETT	GA w/ETT	Y	Semaglutide	Aspiration risk	Induction	Suxamethonium	NS	Elective
12	Laparoscopic surgery	N/A	NS	N/A	Semaglutide	Case cancelled	N/A	N/A	N/A	N/A
13	General surgery	GA w/ETT	GA w/ETT	Y	Liraglutide	Anaphylaxis	Induction	Suxamethonium	NS	Elective

ENT: ear, nose and throat; GA: general anaesthesia; GLP-1: glucagon-like peptide 1; N/A: not applicable; NGT: nasogastric tube; NS: not stated; preop.: preoperatively; RSI: rapid sequence induction; w/ETT: with endotracheal tube; Y: yes

A summary of patient demographics relating to critical incidents from webAIRS for patients on GLP-1 agonists is outlined in Table 2. Most of the patients were female (9/13, 69.2%), with the overall patient cohort between 30 and 79 years of age. Patients with an ASA PS of 3 were most common (8/13, 61.5%), with six patients (46.2%) having a BMI of 35 kg/m² and above. Ten reports noted an established T2DM diagnosis for which GLP-1 agonist treatment was a part of its management. A diagnosis of gastro-oesophageal reflux disease (GORD) was noted for four cases (30.8%). All but one patient presented for elective procedures.

Table 3 outlines the specific GLP-1 agonist type used alongside its withholding period and patients’ preoperative fasting duration. Almost all patients were treated with semaglutide (12/13, 92.3%). Documented cessation of semaglutide preoperatively ranged from not being ceased at all to up to two weeks. Cessation time was not provided for the single patient on liraglutide, nor for six of the patients on semaglutide. The fasting status was mentioned as being more than 6 h in five (38.5%) patients. Two (15.4%) patients were documented to have gastrointestinal symptoms preoperatively, with no patients having a gastric ultrasound performed or nasogastric tube (NGT) inserted prior to surgery.

A detailed case summary of the intraoperative anaesthesia event and possible factors that influenced the patient’s gastric volume and/or motility are listed in Table 4. Of the incidents, there was an equal number of gastroenterological procedures and laparoscopic procedures performed at four (30.8%) cases each respectively.

Ten (76.9%) cases had planned a general anaesthesia with endotracheal tube (ETT) and two (15.4%) cases were planned sedations. Both planned sedation cases were gastroenterological procedures. One of these was converted to a general anaesthesia with ETT, due to presence of gastric contents on direct endoscope view and to minimise aspiration risk. Overall, there was one emergency general anaesthesia, 10 elective general anaesthesias and one elective sedation performed for cases on GLP-1 agonists. One elective procedure was cancelled by the anaesthetist owing to the patient having not ceased their sodium-glucose co-transporter-1 and semaglutide medication preoperatively. Rocuronium and suxamethonium were used in six (46.2%) and five (38.5%) cases respectively.

Of the seven total cases involving concerns surrounding aspiration, six (46.2%) of these cases occurred during induction of anaesthesia. Aspiration risk was defined as the presence of a full stomach on direct endoscope view. An aspiration event was defined when liquid or solid material enters the trachea and lungs due to insufficient laryngeal protective reflexes, resulting in inhalation of oropharyngeal or gastric contents.

There were six (46.2%) incidents that involved an aspiration risk due to a full stomach and one (7.7%) aspiration event was reported. The one aspiration event happened in a patient presenting for an ear, nose and throat surgery who had been fasting for greater than 6 h whilst on semaglutide for blood glucose level control for T2DM. During induction the patient had immediate regurgitation of a large volume of solid gastric contents prior to intubation with cricoid pressure. Presence of gastric contents when suctioning from the endotracheal tube after induction was suggestive of an aspiration event. The patient also experienced a prolonged desaturation requiring high airway recruitment pressures and additional administration of oxygen. An NGT was inserted and a large volume of gastric content was removed prior to emergence. The patient was extubated awake and required an unplanned admission to the intensive care unit.

Discussion

There were 13 incidents reported to webAIRS over a period of eight months, with the first incident reported in August 2023. Various patient and possible event factors may have contributed to the development of the event and these should be considered early for patients on GLP-1 agonists to prevent adverse events. Given that all aspiration cases involved the use of semaglutide, the authors’ conclusions drawn have to be confined to this drug only.

Patient factors

BMI

Numerous studies have identified the association between obesity and aspiration risk during anaesthesia. Specifically, a BMI >30 kg/m² has been associated with increased postoperative morbidity and mortality compared with patients with a BMI between 18 kg/m² and 25 kg/m². ³² In this analysis, there were eight patients with a BMI >30 kg/m² and who thereby had an increased baseline risk of aspiration. The pathophysiological mechanisms relating to an increased aspiration risk are multifaceted, including elevated intra-abdominal and intragastric pressures. ³³

Although only four cases reported established GORD, this is potentially under-reported in this voluntary database given the statistically significant association of increased risk of GORD and erosive oesophagitis in obese patients. ³⁴ It is assumed that the risk of potential aspiration is increased multifactorially for this patient population as it is not only due to their body habitus, but is also secondary to GLP-1 use.

T2DM control

The majority of patients in this dataset reported being on GLP-1 agonist for T2DM control. Numerous studies have highlighted the relationship between delayed gastric emptying time in T2DM compared with healthy volunteers.^35,36 Of direct concern to anaesthesia providers is the impact on gastric emptying and aspiration risk in patients with T2DM treated with GLP-1 agonists. Whilst glycated haemoglobin (HbA1c) levels or time since T2DM diagnosis were not reported in the 13 reports, gastric emptying has been reported to be delayed in up to 30–50% of patients with poorly controlled T2DM (HbA1c >8.5%, long-standing T2DM (>10 years) and patients with a high prevalence of microvascular complications). ³⁷

Perioperative management of patients on GLP-1 agonists

Perioperative withholding of GLP-1 agonists and alterations to preoperative fasting duration remain points of discussion amongst anaesthesia providers. Recently, the literature has begun to recognise the potential aspiration risk of GLP-1 continuation and delayed gastric emptying.^19,38 Of the seven patients identified in the reported incidents as a perceived or actual aspiration risk, all were treated with semaglutide (Ozempic).

In 2023, the American Society of Anesthesiologists (ASA USA) provided a consensus-based guidance regarding this issue in both adults and children, based upon days prior and day of procedure. ³⁹ Importantly, this is regardless of the clinical indication the GLP-1 agonist has been prescribed for, often for weight loss if not T2DM. The ASA USA suggest that for daily dosed GLP-1 agonists (e.g. liraglutide), they should be withheld on the day of surgery. ³⁹ This was agreed by clinical practice recommendations by ANZCA. ²³ GLP-1 agonists requiring weekly dosing, for example semaglutide, should be withheld one week prior to surgery. ³⁹ The ANZCA clinical recommendations did not comment on the time frame for withholding semaglutide, but acknowledged the lack of data at present currently supporting specific time periods for omission preoperatively. ²³

Fasting

Fasting times were poorly documented, with only five patients documented to have fasted for more than 6 h. Current guidelines by ANZCA recommend elective surgery patients over 16 years of age to fast from solid food at least 6 h prior to anaesthesia, with clear fluids up to 2 h prior to anaesthesia. ⁴⁰ In the absence of other risk factors, adhering to these recommendations is reasonably expected to result in an empty stomach preoperatively and thereby a reduced aspiration risk. Of note, ‘sip-til-send’, which is an emerging trend involving access to clear fluids until the patient is sent to the operating theatre, has not currently demonstrated a more significant aspiration risk when compared with traditional fasting fluid guidelines. ²² However, there have been numerous single-centre retrospective case reports published regarding increased aspiration risk of patients treated with GLP-1 agonists, despite prolonged fasting time preoperatively. ⁴¹

Gastrointestinal symptoms on day of surgery

Two patients in this cohort were reported to have gastrointestinal symptoms on the day. Both were treated with semaglutide and both adverse events occurred during induction.

The first incident involved an aspiration risk of a patient between 60 and 69 years of age and a BMI >35 kg/m² scheduled for an elective upper endoscopy. The patient complained of bloating and discomfort preoperatively despite fasting for 8–12 h. Sedation was administered via total intravenous infusion, followed by significant undigested gastric contents visualised directly by the endoscope. Given that the patient tolerated the procedure well, the upper endoscopy was completed with a low threshold for withdrawing the endoscope to facilitate suctioning if gagging/retching occurred. The second incident was a suspected anaphylaxis, involving a patient between 30 and 39 years of age with an unreported BMI undergoing an elective diagnostic laparoscopy. The patient preoperatively complained of right iliac fossa pain along with nausea. There was no gastric ultrasound, pre-emptive NGT insertion, or regurgitated gastric contents documented. A rapid sequence induction (RSI) using rocuronium with cricoid pressure was performed.

The ASA USA recognises that should gastrointestinal symptoms be present, including nausea, vomiting, retching, bloating or abdominal pain, consideration into delaying the elective procedure and specific discussion of aspiration risk should be communicated between the patient and proceduralist. ³⁹ Whilst there was no documentation for either of these cases in the database regarding withholding time of GLP-1 preoperatively, both cases took precautionary measures to minimise aspiration risk intraoperatively. These related to a low threshold for endoscope withdrawal for the upper endoscopy for the first case and undergoing an RSI with cricoid pressure with reverse Trendelenberg for the second case. Given concerns for retained gastric contents prior to endoscopy, utilising a topical anaesthesia approach to facilitate a minimally sedated gastroscopy (with a 5 mm ultrathin gastroscope) could help for direct stomach inspection and presence of intragastric contents. ²³

Event factors

RSI

Performing an RSI is a specific technique well known by anaesthesia providers to minimise the risk of pulmonary aspiration in patients who have a higher than normal risk. ⁴² Of the 12 cases that involved an incident in the operating theatres, nine were documented to have undergone an RSI. This is in keeping with current recommendations in the literature whereby when a GLP-1 receptor agonist is unable to be withheld for at least three half-lives, an RSI should be considered in order to minimise aspiration risk.^25,41

Of these nine incidents undergoing a general anaesthesia and RSI, five clinicians used suxamethonium and four used rocuronium for their RSI. Using suxamethonium, whilst providing a rapid onset of action and potentially improving intubating conditions, has been associated with a more rapid oxygen desaturation during RSI in obese patients.^43,44

Preoperative gastric ultrasonography

None of the reporters mentioned the performance of preoperative gastric ultrasounds. Gastric ultrasonography has been suggested by ANZCA in their fasting guidelines as a potential tool preoperatively to determine consistency and volume of gastric contents to guide further management. ²³

However, there are various limitations when performing gastric ultrasonography. Specifically, there is potentially a general lack of training and experience amongst anaesthesia providers in performing a gastric ultrasound by the bedside. Literature acknowledges that a median of 24 cases are required by anaesthetists to achieve a 90% success rate in visualisation and interpretation of sonography. ⁴⁵ Additionally, in 2–3% of patients, the result may be inconclusive despite adequate technique and experience. ⁴⁵ In patients with a large hiatus hernia or previous gastric surgery, accuracy of sonographic findings may be limited given that ingested gastric contents may be localised in the herniated part of the stomach that cannot be easily visualised. ⁴⁶ Availability of sourcing an ultrasound in the operating theatre environment may also represent a logistical challenge in some institutions.

Preoperative NGT insertion

While no reported incidents noted insertion of an NGT preoperatively, there has been discussion in the literature regarding pre-emptive NGT placement, specifically related to reducing aspiration risk at the time of induction. ⁴⁷ However, whilst pre-emptive NGT insertion may allow the clinician to ‘empty’ the stomach contents to reduce aspiration risk intraoperatively, careful consideration of the benefits and risks should be done given the potential for inducing vomiting and thereby potential aspiration in some patients. ⁴⁷ It is also unlikely that solid contents will be successfully removed via NGT.

Intravenous erythromycin preoperatively

The literature has recently recognised erythromycin’s ability to accelerate gastric emptying by GLP-1. ⁴⁸ For emergency or urgent endoscopic procedures with patients on GLP-1 agonists, ANZCA recommends discussion with the anaesthetist regarding the role of a single intravenous dose of 3 mg/kg (maximum dose of 250 mg) 1–2 h preoperatively. ²³ They also note its potential for use in elective non-endoscopic procedures. In this analysis, no patients were given erythromycin preoperatively to facilitate gastric emptying.

Recommendations

Cessation time of GLP-1 agonists remains a point of discussion with currently only low-quality evidence in case reports and case series present in the literature. Patient care should be individualised and best evidence based where possible, with careful assessment of each patient’s history and reason for GLP-1 prescription. It appears important that patients remain well fasted, there is appropriate cessation time of GLP-1 agonists preoperatively and specific anaesthesia precautions are taken during surgery. There should always be a focus on clear interdisciplinary communication between anaesthetists, surgeons, general practitioners (GPs) and endocrinologists.

Anaesthesia considerations

A thorough preoperative anaesthetic assessment is crucial for all patients, including whether the procedure is elective or urgent. In addition to the routine anaesthesia assessment, there should be specific consideration to patient, surgical and anaesthetic factors that increase aspiration risks. ⁴⁹ Patient factors relate to having a ‘full stomach’ (gastric volume >1.5 ml/kg), delayed gastric emptying, insufficient lower oesophageal sphincter tone, or oesophageal diseases such as previous gastrointestinal surgery or morbid obesity. ⁴⁹

Specifically relating to patient medications, the assessment should include a detailed review to identify patients on GLP-1 agonists, including the type, dose, duration since being prescribed and when the last dose was taken. The indication of whether it is for weight loss or diabetic control is important, as this determines risk secondary to poor glycaemic control. Co-prescription of medications such as opioids and proton pump inhibitors, or recent use of recreational drugs such as cannabis and alcohol also increase one’s baseline aspiration risk.^{50 –53} The presence of gastrointestinal symptoms on the day of procedure should be screened as this may be an indicator for a full stomach.

Surgical type should also be considered, whereby upper gastrointestinal surgery, procedures requiring lithotomy or head-down position, laparoscopic procedures and abdominal surgeries all increase one’s aspiration risk. ⁴⁹

There are numerous factors that may influence the timing of when GLP-1 agonists should be withheld, specifically if it is for weight loss or diabetic control. For patients taking the drug for weight loss, the authors recommend withholding the medication for three half-lives preoperatively to allow for 88% of the drug to clear, following ASA USA guidelines. This equates for patients on daily dosing (e.g. liraglutide), withholding the medications three days prior and for those on weekly dosing (e.g. semaglutide) withholding for three weeks. For diabetic patients, it is advisable to organise a referral to an endocrinologist by their GP. They can assist with having an early discussion on the risk–benefit of withholding the medication for three half-lives preoperatively and whether bridging therapy for glycaemic control, such as insulin, preoperatively is required.

While there was no use of preoperative ultrasonography from the webAIRS data, the authors recognise its utility in patients who have not withheld the GLP-1 agonist for three half-lives.^41,46 The aim of the assessment is to determine gastric volume and whether it is safe to proceed with elective surgery. The authors appreciate the potential need for further education to perform reliable bedside gastric ultrasound in determining gastric volume and guiding their anaesthesia plan. By focusing on further education with upskilling or teaching of gastric ultrasound, this would likely lower the threshold to utilise this adjunct when approaching patients on GLP-1 agonists. If the gastric volume is >1.5 ml/kg or if there is uncertainty relating to determining the volume, then potential consideration to delay elective procedures, or adjustments in the anaesthetic plan, including considerations of regional anaesthesia without sedation or RSI are recommended.

Administering intravenous erythromycin 1–2 h pre-induction as a single dose of 3 mg/kg (maximum dose 250 mg) can be considered for emergency or urgent endoscopies, as well as for elective non-endoscopic procedures.

The authors do not recommend the blanket extension of fasting times due to a lack of evidence regarding a reliable time of gastric emptying currently.

Role of the GP

It is important to stress the crucial role GPs play in community management of T2DM in Australia. With the increasing prevalence of GLP-1 agonists, such as liraglutide and semaglutide, included in the Pharmaceutical Benefits Scheme, the literature has continued to reinforce its cardiovascular and metabolic benefits.^54,55 As such, there is a need to ensure that GPs are well-educated regarding the comprehensive and specific implications of these medications prior to prescribing but also of the possible perioperative implications.

Limitations

The authors acknowledge several limitations to this study. Because the study involves evaluation of a voluntary, de-identified incident reporting system, being webAIRS, the reports are subjective in nature and limited to the experiences reported. By using tickboxes and free text manual entries of the narrative regarding the event helps enable assessors to attain a fairly complete picture of the incident and extrapolate this information. However, whilst the entry fields provide valuable information, they are not mandatory and some reports were incomplete, which affects the reliability of the conclusions drawn. It is not possible to retrieve this data retrospectively. Another limitation is that incidents reported in patients taking GLP-1 agonists were not compared with similar incidents reported in patients not taking these medications. Hence, relative risks cannot be determined. The number of incidents identified that were related to GLP-1 agonist was also small, which may affect generalisability to the wider Australian and New Zealand population.

As webAIRS does not collect denominator data and relies on voluntary reporting of de-identified incidents, it is not possible to calculate or determine the actual risk of aspiration in patients treated with GLP-1 agonists. As reporting to webAIRS is not mandatory, it is possible that the true number of patients presenting with increased aspiration risk or aspiration as a side-effect of GLP-1 agonists treatment is under-reported.

Confounding factors, including variability of patient characteristics and reporting bias relating to GLP-1 agonist use and aspiration risk, may also exist within the dataset. Whilst it was documented whether the aspiration risk/event occurred during induction or emergence, the exact time frame in relation to the anaesthetic and initial symptom presentation was not recorded for the majority of cases.

In summary, with the rising prescription of GLP-1 agonists across Australia and New Zealand, anaesthetists will likely encounter patients in the perioperative period who are treated with GLP-1 agonists for T2DM management or weight loss goals. Currently, clear guidance on risk management for these patients remains in development. Aspiration risk continues to be an issue, with the potential of causing significant patient harm. Further research and comprehensive guidelines are necessary to aid anaesthesia providers in risk stratification for patients on GLP-1 agonists and determining the optimal duration for withholding medication. The role of ultrasound has been described as a valuable adjunct in determining volume of gastric contents preoperatively and may be of benefit in these patients and more broadly. Careful consideration in balancing the risks of aspiration and glycaemic control associated with withholding GLP-1 agonist in the context of T2DM is required.

The authors welcome both the recent ANZCA and Australian Society of Anaesthetists recommendations and recognise the interdisciplinary approach required for distribution of a clear information pathway from the time of first prescription of GLP-1 agonists to the perioperative management. Effective communication and collaboration with anaesthetists, surgeons, gastroenterologists, GPs and endocrinologists are required to best optimise patients on GLP-1 agonists preoperatively.