New ESAIC fasting guidelines for clear fluids in children

Much ado about nothing or is it?

This Invited Commentary accompanies the following original article: Frykholm P, Disma N, Anders-son H, et al. Preoperative fasting in children: guideline from the European Society of Anaesthe-siology and Intensive Care. Eur J Anaesthesiol 2022: 39; 4–25.

Preoperative fasting, one of the pillars of anaesthesia since the late 19th century, was introduced into clinical practice to reduce the frequency of vomiting and aspiration of gastric contents under anaesthesia.¹ Although proof that fasting prevents vomiting and aspiration has remained elusive beyond anecdotal evidence,² it nonetheless has become a sanctified prerequisite for every patient undergoing elective surgery. Recently Frykholm and colleagues identified a long list of failings attributed to parents who are non-compliant with the current 2 h fast after clear fluids, thus leading to prolonged fasting intervals in children and, based on their determination, concluded that a 1 h fast after clear fluids is the ‘holy grail’ solution to these failings.³ Although I strongly support challenging established dogma periodically, I also believe that amending long-standing and safe guidelines such as our fasting guidelines should take place only in the face of substantive and compelling evidence. After reviewing the evidence presented by the authors and the balance of the literature on this subject, I disagree with the authors that aiming for a 1 h fast after clear fluids prevents prolonged fasting and the sequelae that may follow.

The authors advocate a 1 h fast after clear fluids to address the poor compliance with the current 2 h fasting interval and the associated large number of children who fast for prolonged periods. Investigators studied several approaches to improve compliance including allowing children to drink clear fluids until the time of surgery. In one study using this approach, fasting compliance improved compared with a 2 h fast after clear fluids, although >30% of the children in the ‘paediatric suites’ fasted >6 h, similar to that in the control (2 h) fasting group.⁴ In a larger multicentred cohort study in which children were instructed to drink up to 1 h before surgery, they drank until 2.5 h not 1 h before surgery and 25% fasted up to 27.5 h.⁵ Even within the context of research studies, liberal drinking of clear fluids failed to eliminate prolonged fasting. A second approach that is known to be very labour intensive, expensive, and site-specific was to implement a quality improvement strategy. Streamlining communication and clarifying the fasting instructions with parents dramatically reduced the fasting interval.⁶ Although effective, it is unrealistic to expect that a quality improvement strategy could be implemented in every paediatric anaesthesia location globally. A third approach, which proved to be the simplest, least expensive, and most effective strategy was to text message the parents on the afternoon before surgery in addition to the standard written and verbal instructions, a strategy that resulted in 92% of the children fasting appropriately after clear fluids compared with 33% who did not receive text messages (odds ratio 29.2).⁷ Sending a text message to the parents the afternoon before surgery is the most affordable and effective approach to ensure a very high compliance with clear fluid fasting guidelines.

In the ‘80 s and ‘90 s with elective surgery conducted primarily on inpatients, children were offered clear fluids until 2 h hours preoperatively to dissuade them from taking drink or food from their roommate's trays! However, as surgery moved to the outpatient setting, enforcing the fasting guidelines became the responsibility of the parents. Unfortunately, the result was that prolonged fasting became all too common.⁸ When asked why they did not offer clear fluids before surgery, parents cited a litany of reasons including being unwilling to awaken their child to drink, difficulty understanding fasting instructions, and believing that withholding preoperative fluids reduced vomiting after surgery.^9–12 Although the quality improvement approach addressed some of the reasons why parents are noncompliant with the fasting guidelines, by educating them about the need to adhere to fasting guidelines and to give clear fluids preopera-tively, a 1 h fast addressed none of the reasons.

The authors cite a panoply of sequelae that have been attributed to prolonged fasting including complaints of thirst, hypotension, difficulty in establishing intravenous access, and metabolic derangements including ketoaci-dosis and hypoglycaemia. However, a comprehensive review of the evidence failed to substantiate these as clinical problems, and none have harmed children even after prolonged fasts.^13–20 In terms of thirst, children generally tolerated prolonged fasts with ∼ 25%, complaining of being very thirsty although it is noteworthy that the perception of thirst did not correlate with the duration of fasting.^13,14 The authors examined children's thirst after fasting using a dichotomous outcome, but a yes/no response gives a less than meaningful interpretation ofthirst.¹⁵The title ofTable 4,³ ‘Ketoacidosis,’ might suggest to the casual reader that a ‘prolonged fast’ predisposes to ketoacidosis, but that is misleading. In a study of the metabolic changes in children <3 years of age during fasting, the blood concentration of ketones correlated poorly with the duration of fasting,¹⁶ consistent with published evidence.¹⁷ Furthermore, in children <3 years old who fasted for up to 14 h, blood ketone concentrations ≥ 1.0 mmol l⁻¹ (defined as ketosis) were present in only 10% and ≥3.0 mmol l⁻¹ (defined as ketoacidosis) in 1%.¹⁸ Physiological and metabolic sequelae after even prolonged fasting are trivial and inconsequential in infants and children.

Although the risk of pulmonary aspiration has never been correlated with the residual gastric fluid volume (GFV), a pH<2.5 for gastric fluid pH and volume >0.8 ml kg⁻¹ remain the threshold triggers for pneumonitis after aspi-ration.²¹ When children were permitted to drink up to 1 h before surgery, the frequency of GFVs >1 ml kg⁻¹ was three-fold greater than after a 2 h fast,²² and was four-fold greater when the threshold was 0.8 ml kg⁻¹.^21,23 However, not a single child in that study assigned to fast 1 h after clear fluids attained that target, leading me to question whether such a fasting interval is even achievable in clinical practice. If all children fasted 1 h preoperatively, bringing the time of surgery forward due to a cancellation could result in a child fasting for as little as 30 min after clear fluids, the effect of which could reach a 15-fold increase in the number of children with a residual GFV >2 ml kg⁻¹ compared with a 2 h fast.¹⁵ Or the clinician might delay the case, wasting valuable operating room time. Fasting <2 h after clear fluids holds a tight, inverse, and exponential relationship with the residual GFV (r² = 0.90), which moves in the diametrically opposite direction to the improvement in safety achieved by fast-ing.²⁴ There is no margin to treading the fine line of inducing anaesthesia by inhalational induction in the presence of potentially exponentially large gastric fluid volumes in children.

Regurgitation is rare in children during elective surgery, although these authors reported a shocking 1% incidence in 10 019 children.²⁵ This incidence could also be the incidence of aspiration if the child had ingested a large volume of clear hyperosmolar fluids after fasting ≤1 h. Large residual GFVs are compounded by the wide variability in gastric emptying half-times after clear fluids, and the wide variability between measurement techniques, the latter rarely factored into the risks.^23,26–29 If a 1 h fast is adopted without qualification, we can guarantee that children with gastric dysmotility, diabetes and other bowel disorders will present with even larger GFVs than those highlighted above. When summarised, a model of gastric emptying data revealed enormous interindividual variability that is unaccounted for in a 1 h fast after clear fluids and emphasises the large residual GFV that may be present in as many as 50% of children.²⁸ If 1% of these children regurgitated and aspirated during anaesthesia, then aspiration pneumonitis will become the next major focus of morbidity and mortality in paediatric anaesthesia! The variabilities in gastric emptying after various clear fluids and in measurement techniques should cause every paedi-atric anaesthesiologist to pause and question the safety of pursuing a 1 h fast after clear fluids in children.

The rate of pulmonary aspiration in children undergoing elective surgery in the past decade has varied between 9.3³⁰ and 0.6: 10000.³¹ The vast majority of the aspirations after clear fluids resulted in full recovery, although in one study, 3% of children required prolonged ventilation postoperatively.³⁰ The rarity of pulmonary aspiration poses enormous challenges for investigators studying the factors that predispose to aspiration and designing clinical studies to further reduce the rate. The authors claim that the rates of aspiration do not differ between 1 and 2 h fasts after clear fluids in their study,⁵ although the design was powered to prove noninferiority in the overall complication rate (∼1%), not in the aspiration rate. When children drank clear fluids ad libitum, the frequency of documented aspiration was 3:10000,^5,25 although the frequency of documented and suspected aspiration combined exceeded that incidence by three- to seven-fold, 10.6 to 20: 10000.^5,25 Knowing how difficult it is to confirm aspiration, grouping documented and suspected aspirations is an appropriately cautious approach. Regardless of the lack of long-term sequelae, any shift in practice guidelines that potentially increasetheaspirationrateshouldbestronglydiscouraged.

What is the physiological benefit of revising the fasting guidelines to allow children to ingest a small volume of clear fluids preoperatively? Most children arise in the morning after an overnight fast with a fluid deficit that is regarded as mild dehydration. Segar and Holiday recognised this level of dehydration and recommended 10 ml kg⁻¹ h⁻¹ intravenous balanced salt solution to restore euvolaemia and downregulate antidiuretic hor-mone.³² This intravenous fluid regimen in children dwarfs the fractional contribution of preoperative oral fluids, restores euvolaemia, and is more salutary to children under anaesthesia than a preoperative drink.

Until compelling evidence is forthcoming,³³ I am resolute in my conviction that attempting to achieve a 1 h fast after clear fluids in children is perilous, increasing the potential risk for peri-operative regurgitation and aspiration on the one hand while offering negligible benefits on the other. Knowing that the residual GFV after ingesting water immediately before induction of anaesthesia is small is very reassuring but not knowing the magnitude of the residual GFV after ingesting a large volume of clear fluids with an increased electrolyte concentration immediately before induction is very disquieting, and should raise concerns about aspiration. Several strategies have been highlighted to obviate prolonged fasting in children, but a 1 h fast is not one of them. Anaesthesiologists remain a conservative breed, having taken almost 10 years to implement the previous fasting guidelines of 2 h after clear fluids.³⁴ Change for the sake of change without unambiguous evidence that the risk/benefit ratio overwhelmingly favours drinking clear fluids 1 h before surgery should cause us all to pause and reflect on our oath of ‘primum non nocere’.