Peri‐operative fasting in adults: an international, multidisciplinary consensus statement

Summary

Introduction

Evidence suggests that existing pre‐operative fasting guidelines are associated with prolonged fasting times. Prolonged fasting, particularly from clear liquids, has the potential to harm patients through reduced peri‐operative wellbeing; impaired glucose metabolism and peri‐operative inflammatory response; delayed return of bowel function; and reduced muscle strength. Liberalisation of fasting practices has, therefore, become increasingly common. Such a change in practice dictates the need for updated practice guidance. We aimed to develop recommendations on peri‐operative fasting that reflect increasing global awareness of the adverse effects of prolonged fasting.

Methods

Following a systematic literature review, 13 draft recommendations related to peri‐operative fasting were developed iteratively. These were modified during a three‐round Delphi process by an international, multidisciplinary stakeholder panel, which included: patients; anaesthetists; surgeons; physicians; nurses; and members of relevant international organisations from five continents.

Results

Sixty‐eight stakeholders participated in the Delphi consensus process. The panel subsequently agreed on eight recommendations. We recommend continuing current practices on pre‐operative fasting for solid food and non‐clear liquids. We recommend encouraging clear liquids until 2 h before the start of anaesthesia or sedation, unless institutional protocols allow for more liberal liquid intake. We further recommend implementation of institutional protocols that allow more liberal clear liquid intake < 2 h before the start of anaesthesia or sedation. Salivation stimulants can be used until transfer for the procedure. Oral intake should be resumed as soon as clinically feasible. Preprocedural gastric ultrasound performed by a trained provider may be used to guide clinical decisions when additional information is required.

Discussion

This international, multidisciplinary consensus statement aims to improve the quality of patient care by minimising periprocedural fasting times, within safe margins. To achieve this, liberalised pre‐operative clear liquid intake regimens may be implemented with institutional protocols.

Plain Language Summary

What we did

An international group of doctors, nurses and patients worked together to create new recommendations that make fasting safer and more comfortable for people needing anaesthesia. They reviewed research and reached agreement on eight main recommendations.

Why we did it

Before an operation or procedure, people are told not to eat or drink for several hours so that their stomachs are empty when they receive anaesthesia. This is meant to reduce the small risk of food or liquid going into the lungs during anaesthesia. But many patients fast for much longer than needed, with some finding that this is often more than 14 hours without food and 9 hours without drinks. Long fasting times can make people feel tired, thirsty, anxious or weak, and can affect their recovery after surgery.

What we found

Adults should stop eating solid food and milk products 6 hours before anaesthesia (8 hours after a heavy or fatty meal). Clear drinks like water, tea, coffee with a little milk or juice can usually be taken until 2 hours before surgery. Hospitals may also make protocols allowing clear drinks closer to the time of anaesthesia, such as permitting them to drink clear liquids until they are called for their procedure. Patients should start drinking and eating again as soon as they are able after the operation. Chewing gum or sucking sweets is acceptable until patients are taken for surgery. Overall, these recommendations aim to improve patient comfort, prevent harm from long fasting and support hospitals in adopting safe, modern fasting practices.

Recommendations

1. Patients should fast for at least 6 h from solid food before the start of anaesthesia or sedation. For large, fatty meals, fasting for 8 h or longer may be necessary.

2. Patients should fast for 6 h from non‐clear liquids – including milk (excluding a small volume of milk in tea or coffee, up to one‐fifth of the total volume), milk products, meal replacement drinks and enteral feeding formulas – before the start of anaesthesia or sedation. Clear liquids include water; tea; or coffee with sugar or honey (including a small amount of milk, up to one‐fifth of the total volume); clear juices; lemonade; and clear carbohydrate drinks.

3. Medical facilities should implement institutional protocols to minimise prolonged preprocedural fasting for clear liquids.

4. Patients should consume clear liquids until 2 h before the start of anaesthesia or sedation, unless institutional protocols allow for more liberal liquid intake.

5. Institutions may implement protocols that allow the emerging practice of clear liquid intake < 2 h before the start of anaesthesia or sedation.

6. Preprocedural gastric ultrasound performed by a trained provider may be used to guide clinical decisions in cases where additional information is required.

7. Salivation stimulants such as lollipops and chewing gum may be used until transfer for the procedure and after complete recovery from anaesthesia or sedation.

8. Unless specifically contraindicated, postoperative liquids and diet should be resumed as soon as clinically feasible after anaesthesia or sedation.

Why were these guidelines developed?

Millions of patients are routinely exposed to prolonged fasting, with fasting times of 14–17 h for solids and 9–12 h for clear liquids reported [1, 2, 3, 4, 5, 6]. Emerging practices such as liberal liquid consumption before induction of anaesthesia have been proposed to mitigate such prolonged fasting, given the mounting evidence of its harm [7]. To address the global problem of unnecessary prolonged fasting, these recommendations were developed by a broad, international and multidisciplinary panel which included: patients; anaesthetists; surgeons; cardiologists; gastroenterologists; other physicians; nursing staff; and members of relevant international organisations, to improve the quality of care for patients requiring anaesthesia or sedation.

What guidelines currently exist?

There are multiple guidelines available worldwide [8, 9, 10, 11, 12, 13]. In addition, there are guidelines for specific patient groups, such as those in enhanced recovery after surgery (ERAS) protocols [14]. All the recommendations encourage patients to drink clear liquids up to 2 h before the procedure.

How do these guidelines differ from existing guidelines?

These guidelines were developed by an international multidisciplinary panel including patients and clinical stakeholders involved in prescribing and implementing peri‐operative fasting. The recommendations were supported by a literature review and aimed to focus specifically on minimising unnecessarily prolonged fasting within safe margins. These are the first guidelines proposing a liberal liquid intake within institutional protocols before the start of anaesthesia.

Introduction

Patients scheduled for procedures requiring anaesthesia or sedation are instructed to fast from liquids and solids. To avoid the negative effects of prolonged fasting, international guidelines allow patients to consume solids until 6–8 h before surgery and encourage liquids until 2 h before surgery [8, 9, 10, 11, 12]. However, with median fasting times of 14–17 h for solids and 9–12 h for clear liquids reported, these guidelines are implemented inadequately [1, 2, 3, 4, 5, 6]. For example, in an international survey of 971 anaesthetists, 85% stated that their advice to patients followed current society guidelines, although 50% enforced strict fasting and did not allow clear liquids after midnight before surgery [15]. The poor adherence by healthcare professionals highlights the need for improvements in clinical care for patients.

Prolonged fasting, particularly from clear liquids, has the potential to harm patients through reduced peri‐operative wellbeing; impaired glucose metabolism and peri‐operative inflammatory response; delayed return of bowel function; increased likelihood of postoperative delirium; and reduced muscle strength [16, 17, 18, 19, 20, 21, 22, 23, 24, 25]. Improved understanding of these adverse effects has prompted clinicians to intensify their efforts to reduce liquid fasting times. However, despite the implementation of various strategies, including education and training; mobile telephone reminders; surgical planning tools; and multidisciplinary interventions, the targeted pre‐operative clear liquid fasting time of close to 2 h is still not achieved in many centres [6, 26, 27, 28, 29, 30, 31]. This has led clinicians to explore more liberal pre‐operative oral intake and to question the need for pre‐operative clear liquid fasting [32, 33, 34, 35, 36, 37].

Liberal liquid fasting protocols have been implemented in some centres for children for over 20 years and are adopted increasingly for adults [38]. This approach, known as ‘Sip Til Send’, is endorsed by the Centre for Perioperative Care in the UK and the Government of New South Wales in Australia [33, 39, 40]. It is already the standard care in many centres in the UK [33, 35, 39, 41, 42]. The Australasian Sip Til Send Network now encompasses more than 60 hospitals [40, 43]. In Germany, a growing number of hospitals permit unrestricted clear liquid intake until patients are called for their procedure, implemented on a case‐by‐case basis with fasting cards [36, 44, 45, 46]. Similar approaches have been implemented in the Netherlands and Ireland [34, 47, 48].

The primary rationale for requiring patients to fast before the induction of anaesthesia or sedation is to reduce the risk of pulmonary aspiration. In elective procedures, this is a rare event, with an estimated incidence of approximately 2–3 per 10,000 cases, morbidity of 0.8–0.6 per 10,000 cases and mortality of 0.1 per 10,000 cases [49, 50, 51]. As a result, randomised trials to assess the efficacy of pre‐operative fasting to prevent peri‐operative pulmonary aspiration are not available. Previous guidelines on peri‐operative fasting were therefore based on expert opinions and extrapolation of physiological data or non‐outcome‐based research [8, 9]. To improve guideline implementation, their development requires engagement of all key stakeholders [4]. Furthermore, diversity is known to improve guideline development and reduce the risk of bias [52]. Importantly, it is crucial to include patient representatives to encompass patient expectations and perspectives [53].

We sought to develop updated recommendations for peri‐operative fasting through consensus among an international, multidisciplinary group of stakeholders.

Methods

This international multidisciplinary consensus was led by a steering committee of nine clinicians (all authors of this guideline). To inform the development of the draft recommendations by the steering committee, we conducted a systematic literature review (online Supporting Information Appendices S1 and S2). The literature search followed a population/intervention/comparison/outcome framework and focused on three major topics: the impact of pre‐operative fasting on peri‐operative outcomes; the utility of ultrasound in assessing gastric content and the associated risk of aspiration; and the effects of postoperative fasting on patient outcomes. The systematic literature review included assessments of risk of bias and the overall level of evidence using GRADE methodology [54]. Draft recommendations were modified during a three‐round Delphi approach in accordance with the conducting and reporting of Delphi studies (CREDES) guidance to develop final recommendations (online Supporting Information Appendix S3) [55].

We convened a Delphi panel, inviting experts on peri‐operative fasting selected via a focused literature search and non‐experts selected via relevant organisations (online Supporting Information Appendix S4). The panel comprised anaesthetists; surgeons; nurses; cardiologists; gastroenterologists; other physicians; patient representatives; and members of international organisations related to the topic, including patient safety organisations, ERAS societies and others (Table 1).

Table 1.

Stakeholder panel (n = 68). Data are number (proportion).

Total	n = 68
Sex; female	29 (43%)
Continent
Europe	40 (59%)
North America	14 (21%)
Australia	9 (13%)
Asia	4 (6%)
South America	1 (2%)
Stakeholder group
Steering committee	9 (13%)
Corresponding author of fasting guidelines	5 (7%)
Anaesthetists; research on fasting	6 (9%)
Anaesthetists; comments on fasting	2 (3%)
Anaesthetists; clinicians; no research on fasting	4 (6%)
Surgeons; research on fasting	3 (4%)
Surgeons; international organisations	3 (4%)
Sedation; cardiologists; research	4 (6%)
Sedation; cardiologists; international organisations	1 (2%)
Sedation; gastroenterologists; international organisations	2 (3%)
Nursing; research on fasting	5 (7%)
Nursing; international organisations	4 (6%)
Other clinicians – research on fasting	4 (6%)
Patient representatives	3 (4%)
Members of relevant initiatives and organisations	13 (19%)
Patient safety organisations	5 (7%)
Enhanced recovery after surgery societies	4 (6%)
Others	4 (6%)

The first and second Delphi round were completed electronically and anonymously. Draft recommendations from the first round and revised recommendations from the second round were distributed to all stakeholders, who rated each recommendation anonymously as ‘include’, ‘exclude’ or ‘revise’, and provided comments via a Microsoft Excel spreadsheet (Microsoft, Inc., Redmond, WA, USA; online Supporting Information Appendix S5). These responses were then tabulated and analysed. Recommendations with ≥ 75% agreement proceeded unchanged to the next round or underwent revision based on the Delphi input, if deemed necessary by the steering committee. Recommendations with 51–74% agreement were revised and those with ≤ 50% agreement were not included. Stakeholders were allowed to propose additional recommendations.

The third round involved a virtual round‐table discussion, where all stakeholders were invited to a Zoom online meeting (Zoom Communications, Inc., San Jose, CA, USA). Results of the first two rounds were presented, and all stakeholders were given equal opportunity to briefly voice their views through a structured discussion facilitated by a moderator with experience in round‐table moderation. Recommendations were rephrased if requested, followed by anonymous live voting on agreement with each recommendation. A final list of recommendations was then prepared.

Results

We invited 110 stakeholders, of whom 68 agreed to participate and joined in the first round, with 67 and 40 participating in the second and third rounds, respectively (Table 1 and online Supporting Information Appendices S4 and S6).

In the first Delphi round, 13 recommendations developed following the systematic review were voted on (online Supporting Information Appendix S5). Following steering committee revisions and the addition of one recommendation proposed by stakeholders, nine recommendations were presented in the second Delphi round (online Supporting Information Appendix S5). After revisions, all nine recommendations were presented for the third Delphi round (online Supporting Information Appendix S5), with eight being agreed on as final recommendations with a level of agreement of 89–100%, with one omitted due to poor agreement of 50% (Table 2).

Table 2.

Recommendations for peri‐operative fasting in adults.

Solid food and non‐clear liquids		Level of agreement
1.	Patients should fast for at least 6 h from solid food before the start of anaesthesia or sedation. For large, fatty meals, fasting for 8 h or longer may be necessary.	89%
2.	Patients should fast for 6 h from non‐clear liquids – including milk, milk products, meal replacement drinks and enteral feeding formulas* – before the start of anaesthesia or sedation.	95%
Clear liquids
Definition	Clear liquids include water, tea, coffee with sugar or honey (including a small amount of milk, up to a fifth of the total volume), clear juices, lemonade, clear carbohydrate drinks.
3.	Medical facilities should implement institutional protocols to minimise prolonged preprocedural fasting for clear liquids.	100%
4.	Patients should consume clear liquids until 2 h before the start of anaesthesia or sedation, unless institutional protocols allow for more liberal liquid intake.	100%
5.	Institutions may implement protocols that allow the emerging practice of clear liquid intake < 2 h before the start of anaesthesia or sedation.	100%
Gastric ultrasound
6.	Preprocedural gastric ultrasound performed by a trained provider may be used to guide clinical decisions in cases where additional information is required.	97%
Salivation stimulants
7.	Salivation stimulants, such as lollipops and chewing gum, may be used until transfer for the procedure and after complete recovery from anaesthesia or sedation.	92%
Postoperative oral intake
8.	Unless specifically contraindicated, postoperative liquids and diet should be resumed as soon as clinically feasible after anaesthesia or sedation.	100%

^*Preprocedural discontinuation of tube feeding is not required for most patients with a secure airway, such as a cuffed tracheostomy tube or a tracheal tube already in situ.

These recommendations are intended for all healthcare professionals as guidance for peri‐operative fasting in adults undergoing sedation or anaesthesia. They cannot address every individual condition or circumstance. In some areas, the guidance is intentionally broad, reflecting uncertainty in clinical practice as well as existing (or upcoming) guidelines for specified patient groups. Clinical judgement remains paramount [53].

These recommendations apply to adult patients receiving anaesthesia or sedation for elective procedures, with two notable exceptions. First, prolonged fasting to optimise procedural conditions, such as for bowel preparation. Second, clinical conditions that substantially increase the risk of pulmonary aspiration, such as severe gastroesophageal reflux, resulting in regurgitation of gastric contents in the supine position. In these scenarios, decision‐making should be made on a case‐by‐case basis.

Solids and non‐clear liquids

This consensus statement supports current pre‐operative fasting practices for solid food and non‐clear liquids, reflecting the lack of meaningful new evidence (online Supporting Information Appendix S1) [8, 9].

Gastric emptying for solids varies considerably. It is typically rapid in young, healthy adults but may be delayed in people with pre‐existing conditions such as diabetes mellitus; gastroparesis; end‐stage kidney disease; or those taking medications such as glucagon‐like peptide‐1 receptor agonists [56, 57, 58, 59, 60, 61]. Therefore, the fasting time for solid food may need to be extended for the categories of patients mentioned above, particularly after high‐calorie meals. Any potential harm from prolonged fasting for solid food is likely mitigated by the pre‐operative intake of non‐clear or clear liquids, as recommended by this consensus statement.

Five stakeholders (7%) noted that gastric emptying of milk occurs in < 6 h, a finding supported by physiological studies in healthy volunteers [62, 63]. Nevertheless, the majority of stakeholders supported the 6‐h interval.

Five stakeholders (7%) advocated for a standard 6‐h fasting period for solid food, regardless of quantity and calorie content, noting that this is standard UK guidance [8]. The Australian and New Zealand College of Anaesthesiologists and the American Society of Anesthesiologists (ASA) recommend that solid food with low calorific content (light meal) may be allowed up to 6 h before induction of anaesthesia [9, 12]. For fatty foods or meat, the ASA recommends fasting for 8 h or longer [9]. The Indian Society of Anaesthesiologists states that “if the patient has consumed heavy meals, it may be prudent to wait for at least 10 h prior to administration of sedation or anaesthesia” [11]. Accordingly, several stakeholders noted that the current statement on solid food is appropriately cautious.

• Patients should fast for at least 6 h from solid food before the start of anaesthesia or sedation. For large, fatty meals, fasting for 8 h or longer may be necessary.

• Patients should fast for 6 h from non‐clear liquids – including milk, milk products, meal replacement drinks and enteral feeding formulas – before the start of anaesthesia or sedation.

• Preprocedural discontinuation of tube feeding is not required for most patients with a secure airway, such as a cuffed tracheostomy tube or a tracheal tube already in situ.

Clear liquids

In these recommendations, clear liquids include water; tea or coffee with sugar or honey (including a small amount of milk up to one‐fifth of the total volume); clear juices; lemonade; and clear carbohydrate drinks [8, 64, 65, 66].

This consensus statement recommends encouraging clear liquids until 2 h pre‐operatively, whilst also allowing for institutional protocols which permit clear liquids < 2 h before the start of anaesthesia or sedation [8, 9, 10, 39, 40]. The primary goal of these recommendations is to prevent prolonged pre‐operative fasting for clear liquids.

Guidelines have recommended 2‐h fasting for clear liquids since the end of the last century [67]. Compared with the practice of nil by mouth after midnight, the 2‐h recommendation was historically revolutionary. However, the 2‐h limit is not evidence‐based in a strict sense, since it has not been shown that this reduces risk of pulmonary aspiration. The practice of 2‐h fasting was recommended based on the fact that consuming clear liquids 2–4 h before induction of anaesthesia does not affect stomach volume or even reduce it compared with longer fasting periods [9]. However, allowing clear liquids within 2 h of anaesthesia or sedation does not appear to result in either a clinically relevant increase in sonographically‐ or gastroscopically‐assessed residual gastric volume or an increase in the incidence of pulmonary aspiration [33, 34, 35, 36, 41, 44, 48, 68, 69].

A significant concern with a 2‐h (or any) fasting requirement for clear liquids is that it is difficult to implement and promotes prolonged fasting [15, 70]. The currently recommended 2‐h clear liquid fasting results in a median fasting duration of 9–12 h [1, 2, 3, 4, 5, 6]. Such prolonged fasting can impair the peri‐operative wellbeing of a patient and glucose metabolism; deteriorate peri‐operative inflammatory response; delay the return of bowel function; and reduce muscle strength (online Supporting Information Appendix S1) [16, 17, 18, 19, 20, 21, 22, 23]. Therefore, many healthcare facilities have established institutional protocols to minimise prolonged clear liquid fasting; however, neither of these interventions achieved the targeted pre‐operative fasting time of close to 2 h [6, 26, 27, 28, 29, 30, 31].

Only institutional protocols allowing clear liquid intake < 2 h before induction of anaesthesia have achieved fasting times close to the currently recommended 2 h. Such protocols involved reducing liquid fasting times to 1.5 h; 1 h; allowing clear liquids until departure from home or arrival at the clinic; deciding on a case‐by‐case basis; or introducing the ‘Sip Til Send’ approach [32, 33, 34, 35, 36, 37, 42, 44, 45, 71, 72, 73]. Allowing clear liquids, including carbohydrate‐containing liquids, until the time a patient is called for their procedure could improve wellbeing and reduce caffeine withdrawal symptoms, without increasing the risk of aspiration [32, 33, 34, 35, 41, 45, 74, 75, 76, 77]. Liberal clear liquid intake is already standard practice in some UK, Australian, Dutch and German institutions [33, 34, 35, 36, 41, 43].

The first Delphi round included a statement on carbohydrate drinks. Some stakeholders expressed concern that a separate recommendation for carbohydrate drinks would overemphasise the role of this category of clear liquids, given that carbohydrate drinks are not necessary or useful for all patients or procedures. Although there is reliable evidence that carbohydrate drinks 2 h before a procedure are beneficial compared with prolonged fasting, there is little evidence that carbohydrate drinks are beneficial compared with other clear liquids (online Supporting Information Appendix S1). A more patient‐centred approach was proposed, which involved informing patients about the benefits of clear liquids with calorific content while respecting their personal preferences. Practices reported by stakeholders varied; one stakeholder introduced lemonade ice lollies, some clinicians prescribe apple juice before induction of anaesthesia or encourage coffee with sugar and some milk.

Allowing some milk in tea or coffee is important for many patients. In the UK, 100 million cups of tea are consumed daily, 67% of which are drunk with milk [78]. Tea and coffee with a small amount of milk up to one‐fifth of the total volume empties the stomach as fast as black tea or water [64, 65]. The diluted milk forms a suspension that can pass through the pylorus unhindered. In contrast, pure milk has a substantially higher calorie content, which delays gastric emptying significantly [62]. Whereas 500 ml of water (or tea with some milk) empties from the stomach in < 20 min, a significant amount of milk is still in the stomach 2 h after drinking 500 ml of milk in its pure form [62]. Therefore, a small amount of milk in tea or coffee is considered a clear liquid, whereas pure milk is considered a non‐clear liquid, requiring the 6‐h fasting interval for non‐clear liquids.

By allowing institutional protocols that include clear liquid intake < 2 h before induction of anaesthesia, this consensus statement offers medical facilities the flexibility to decide how best to avoid prolonged fasting. Options include adhering to the 2‐h minimum fast or accepting shorter liquid fasting times, such as 1 h or until called for the procedure, according to local preferences and acceptability. Institutions can either restrict the quantity and the types of clear liquids allowed or allow unrestricted quantity and all types of clear liquids. This approach acknowledges that a generalised definition of liberal fasting times, permitted liquids and quantities that accommodates all patients – from young and healthy to frail and comorbid – and all medical facilities – from small ambulatory units to tertiary hospitals – is unlikely to be achievable.

• Medical facilities should implement institutional protocols to minimise prolonged preprocedural fasting for clear liquids.

• Patients should consume clear liquids until 2 h before the start of anaesthesia or sedation, unless institutional protocols allow for more liberal liquid intake.

• Institutions may implement protocols that allow the emerging practice of clear liquid intake < 2 h before the start of anaesthesia or sedation.

Point‐of‐care gastric ultrasound

An increasing number of studies have proposed the use of gastric ultrasound, particularly in patients at an elevated risk of pulmonary aspiration [79]. However, the relationship between gastric volume/content and pulmonary aspiration has not been quantified. A recent meta‐analysis concluded that patients using glucagon‐like peptide‐1 receptor agonists are at a greater risk of residual gastric contents, but the available evidence did not indicate that this increases the risk of pulmonary aspiration [57]. Additionally, the widely used surrogate measure of aspiration risk, referred to as a ‘full stomach’ and defined as 1.5 ml.kg^‐1, has also been questioned [80].

Since inaccurate findings may lead to patient harm, point‐of‐care gastric ultrasound requires adequate training and skill maintenance and may be more challenging in some situations, such as patients with previous gastric surgery [81, 82, 83]. Furthermore, up to one in 10 patients who adhere to routine fasting guidelines do not have an ‘empty stomach’ [79]. But the incidence of aspiration is only 2–3 per 10,000 cases [49, 50, 51]. Therefore, routine gastric sonography without a clear definition of a ‘stomach at risk’ may increase the use of rapid sequence induction and tracheal intubation techniques inadvertently and unnecessarily, which carry their own risks.

However, when equipment and a trained provider are available, gastric ultrasound may be useful in scenarios of diagnostic uncertainty, and when used to support clinical decision‐making, might improve the safety of patient care [79, 82]. This consensus statement, therefore, concludes that ultrasound may be used to guide clinical decision‐making when additional information is required.

• Preprocedural gastric ultrasound performed by a trained provider may be used to guide clinical decisions in cases where additional information is required.

Salivation stimulants

Small clinical trials have shown that salivary stimulants, such as chewing gum and lollipops, did not increase gastric volume or decrease gastric pH, but did reduce pre‐operative anxiety, thirst and postoperative ileus (online Supporting Information Appendix S1) [84, 85, 86]. Their use is therefore appropriate before induction of anaesthesia; however, chewing gum and salivary stimulants should be removed before the induction of anaesthesia or sedation. In addition, small‐ and medium‐sized clinical trials have shown that chewing gum after recovery from anaesthesia is well‐tolerated and might facilitate postoperative recovery, including reduced postoperative gastrointestinal complications and decreased duration of hospital stay [87, 88, 89].

• Salivation stimulants such as lollipops and chewing gum may be used until transfer for a procedure and after complete recovery from anaesthesia or sedation.

Postoperative oral intake

There is considerable variation in the definition of early oral feeding, in terms of timing, quantity, and type of oral intake [54]. Liberal postoperative oral intake is generally safe, well‐tolerated and beneficial for patients and can be started in the post‐anaesthesia care unit for the majority of patients [90]. Some patients may have an increased risk associated with early postoperative oral intake. Therefore, the decision to resume the diet postoperatively should be made jointly with the surgical team.

• Unless specifically contraindicated, postoperative liquids and diet should be resumed as soon as clinically feasible after anaesthesia or sedation.

Discussion

This international, multidisciplinary consensus statement makes eight recommendations on peri‐operative fasting, aimed at improving the quality of care for patients requiring anaesthesia or sedation (Fig. 1). We largely reaffirm existing recommendations for non‐clear liquids, solids, salivation stimulants and postoperative oral intake (available in online Supporting Information Table S1) [8, 9, 10]. However, we also empower institutions to consider more liberal clear liquid policies, to improve patient experience. Our recommendations support the resumption of liquids and diet as soon as clinically feasible after anaesthesia or sedation. Finally, we acknowledge the emerging role of point‐of‐care gastric ultrasound when used appropriately.

Figure 1.

Recommendations for solid food and liquids.

For most patients, a prolonged fasting period for solids is inevitable unless they eat in the middle of the night; most patients prefer not to do this [91]. It is therefore imperative to keep fasting times for clear liquids short and introduce calorie intake within those clear liquids to minimise calorie deficit.

Allowing liberal clear liquid regimes pre‐operatively is more likely to bring fasting times closer to the recommended 2 h, rather than eliminating fasting altogether [34, 35, 36, 45]. This supports the hypothesis that the strict 2‐h rule may paradoxically be a barrier to achieving optimal fasting durations [41, 70]. This practice has already been implemented in many institutions [32, 33, 34, 35, 36, 41, 43]. High‐quality evidence for the risk of pulmonary aspiration when liberal clear liquid regimens are used is still lacking, because aspiration in elective procedures is very rare [49, 50, 51]. Of note, prolonged fasting does not reliably result in an empty stomach, and even if regurgitation and pulmonary aspiration of clear fluid occur, it is unlikely to result in morbidity [68, 70, 79].

This consensus statement recommends liberal clear liquid fasting regimens only within institutional protocols. These protocols, approved by local authorities, are customised to local circumstances, follow clear rules, and can be adopted by other healthcare facilities. Importantly, the onus should be with the institution rather than individual clinicians. This is because implementation requires a systematic approach involving patient education, staff training and institutional policies and procedures. Moreover, this approach allows data collection on an institutional level to allow safety monitoring and reporting.

These consensus recommendations have some limitations. The evidence base which underpins them is extremely limited. The Delphi panel was convened based on the selection of experts via a literature search, with non‐experts recruited from international organisations. These selection criteria might be viewed as somewhat arbitrary. Due to the online format of the third Delphi round and the global distribution of stakeholders across five continents, not all could attend the final online meeting. As all final recommendations were agreed on in at least two Delphi rounds, this had little impact on the results. Finally, this consensus statement is not supported by any organisation or society. However, most professional societies are national or regional and would not fully represent the highly diverse group of our Delphi stakeholders from five continents, who represent all those involved in peri‐operative fasting.

In summary, this international, multidisciplinary consensus statement aims to improve quality of patient care by minimising periprocedural fasting times within safe margins, including the implementation of liberal liquid fasting regimens in institutional protocols.

Supporting information

Appendix S1. Pre‐operative studies and grading.

Appendix S2. Literature search details for gastric ultrasound.

Appendix S3. Draft recommendations.

Appendix S4. Details of the stakeholder selection and participation.

Appendix S5. Delphi results.

Appendix S6. Stakeholders participating in the Delphi process.

Table S1. Comparison of current consensus statement with the guidelines from the European Society of Anaesthesiology and Intensive Care (ESAIC) and the American Society of Anesthesiologists.