BJA Education, 17 (10): 346–350 (2017)

doi: 10.1093/bjaed/mkx021
Advance Access Publication Date: 30 June 2017
Matrix reference
1A01, 2A03, 3J03

Preoperative fasting in children

Ahmed Mesbah MB BCh FCAI FRCA1 and Mark Thomas BSc (Hons) MB BChir
FRCA FFFPM FAoMEd2,*
1
Locum Consultant Paediatric Anaesthetist, Department of Anaesthesia, Great Ormond Street Hospital for
Children NHS Foundation Trust, Great Ormond Street, London WC1N 3JH, UK and 2Consultant Paediatric
Anaesthetist, Department of Anaesthesia, Great Ormond Street Hospital for Children NHS Foundation Trust,
Great Ormond Street, London WC1N 3JH, UK
*To whom correspondence should be addressed. Tel: þ44 20 7829 8865; Fax: þ44 20 7829 8866; E-mail: [email protected]

evidence for liberalizing preoperative fasting regimens in chil-

dren undergoing elective surgery.
Key points
• Prolonged preoperative fasting times for healthy
elective cases have been extrapolated from the Historical perspective
aspiration risk of ‘full-stomach’ emergency cases. For John Snow in 1847, the intention behind preoperative fast-
• Gastric emptying is regulated by hormonal, neu- ing was not to diminish the incidence of aspiration but, rather,
ronal, and metabolic feedback. to reduce the unpleasantness of vomiting.2 A year later, the first
paediatric death due to aspiration was reported, possibly due to
• Residual gastric volume is a poor surrogate for
the anaesthetist pouring brandy into the child’s mouth to re-
the risk of aspiration, and there appears to be no lieve ‘syncope’.3 As more reports followed, fasting recommen-
causal link or critical volume threshold. dations emerged. Sir Joseph Lister was the first to make the
• Perioperative pulmonary aspiration in children is distinction between fasting from food and fluids and in 1883
rare. The incidence is 0.07–0.1% and the conse- recommended no solid matter in the stomach and ‘a cup of tea
quences of clear fluid aspiration are not or beef-tea 2 h previously’. This distinction carried on into the
catastrophic. 1960s with paediatric instructions favouring sweetened liquids
up to 2 h before operation. However, the years following World
• Newer liberal paediatric fasting regimens for elec-
War II saw the widespread adoption of the ‘Nil per os (NPO)
tive cases seem to confer no increase in aspira- from midnight’ regimen, especially in North America, ignoring
tion risk compared with more conservative the previous distinction between solids and liquids. A possible
regimens. reason for this was the work of Curtis Mendelson.

Mendelson’s syndrome
On reviewing 66 cases of aspiration among 44 016 pregnancies,
Mendelson found two immediate deaths due to airway obstruc-
In 1948, Digby Leigh, in his textbook Pediatric Anesthesia, sug- tion following aspiration of solids and no deaths among those
gested that children should fast from clear fluids for 1 h prior to who had aspirated liquids. However, he noted that those
surgery.1 Yet, in the intervening years, fasting times have in- patients that aspirated fluid were critically ill, with an acute
creased in the belief that this may reduce the risk of pulmonary ‘asthma-like’ attack, and mottling on chest X-ray, which cleared
aspiration of gastric contents. This article will review the histor- in 7–10 days. Mendelson then demonstrated that hydrochloric
ical context within which preoperative fasting guidelines have acid, or un-neutralized human vomitus, injected into rabbit
evolved, the physiology of gastric emptying, and the emerging lungs, reproduced this picture, and concluded that pregnant

Editorial decision: April 6, 2017. Accepted: April 12, 2017

C The Author 2017. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved.
V
For Permissions, please email: [email protected]

346
 Preoperative fasting in children

women were at especially high risk of aspiration due to delayed stimulate cholecystokinin which inhibits the stimulatory ef-
gastric emptying. He subsequently promoted preoperative fast- fects of gastrin on the antrum and carbohydrates stimulate gas-
ing, alkalization of stomach contents, and the greater use of re- tric inhibitory polypeptide.
gional anaesthesia.4 In addition to these neural and hormonal regulatory mecha-
Following Mendelson’s publication, further work on anaes- nisms, blood glucose also affects emptying. At levels in excess
thesia for similarly high-risk patients emerged. The belief that of 8 mmol l1, the delivery of caloric contents is reduced by a
otherwise healthy patients with no risk factors for aspiration negative feedback mechanism from duodenal receptors to
were also at risk stemmed from the work of Roberts and Shirley maintain a constant rate of delivery of 1–4 kcal min1.7
in 1974 who surmised that 25 ml (0.4 ml kg1) of gastric fluid There is a considerable difference in how liquids and solids
with a pH < 2.5 increased the risk of serious aspiration. Their leave the stomach. For non-caloric liquids, such as water, gas-
work was based on preliminary unpublished data from a single tric emptying begins immediately and exponentially, following
Rhesus monkey which had neither vomited nor regurgitated, first-order kinetics, proportional to the volume present in the
but whose right main bronchus was instilled with 0.4 ml kg1 of stomach and thus the gastro-duodenal pressure gradient. For
acid.5 A relationship between the residual gastric volume and caloric liquids and solids, emptying follows zero-order kinetics,
the volume of fluid instilled into the lungs was never estab- which is linear but biphasic in pattern. The duration of the first
lished but rather extrapolated to the average weight of a preg- phase is related to the caloric content of the food. During this
nant woman. Subsequent studies challenged these findings by phase, solids are digested in the fundus into particles <2 mm to
demonstrating gastric volumes >25 ml and pH < 2.5 in 40–80% facilitate passage through the pylorus during the second phase.
fasted healthy patients.2 Given the confusion over what consti- Motility across the entire gastrointestinal tract peaks at approx-
tutes a significant residual volume or pH in adults, it is unsur- imately 30 min and continues for about 4 h during which liquids
prising that in the heterogeneous paediatric population the and particles with a size of 1–2 mm are emptied from the stom-
situation is even less clear. ach. Indigestible solids, such as cellulose-containing vegetables,
which do not break down to <2 mm in humans, empty via a dif-
ferent mechanism that occurs later when the stomach is
Mechanisms of pulmonary aspiration and risk fasting.7,8
factors
For pulmonary aspiration to occur, gastric contents must over- Fasting gastric motility
come three protective mechanisms: first, they must exceed the Between meals, a cyclic pattern of motor activity occurs in the
lower oesophageal sphincter barrier pressure, then regurgitate stomach every 80–120 min. The first phase is characterized by
up the oesophagus through the upper oesophageal sphincter, quiescence, the second by irregular contractions, and the third
and finally pass down the trachea unimpeded by protective air- by continuous phasic contractions lasting up to 5 min. It is dur-
way reflexes such as laryngospasm or coughing. The risk fac- ing these latter contractions that the stomach empties particles
tors predisposing to this sequence of events can therefore be > 2 mm. This process recurs every 2 h and may take up to 6–12 h
broadly classified into three groups: (i) increased regurgitation, to complete.7,8
as a result of gastroesophageal reflux, strictures, and decreased
lower oesophageal sphincter tone; (ii) loss of protective airway
Measurement of gastric emptying
reflexes, such as those seen in neuromuscular disorders or be-
cause of the effects of general anaesthetics and analgesics; and The earliest attempts to observe gastric emptying were made by
(3) increased gastric volume, as a result of inadequate fasting or William Beaumont, an American military surgeon, who in 1822
delayed gastric emptying.6 Of these, gastric volume is the only treated a fur trapper for a gunshot wound to the stomach. The
risk factor that can be modified before operation. healed wound left a permanent gastric fistula through which
Beaumont made direct observations of gastric emptying times,
noting that for easily digested food, such as meat, potatoes, and
Physiology of gastric emptying
bread, these varied from 1.5 h to 5 h, whereas most fluids passed
The stomach performs both mechanical and chemical break- from the stomach almost immediately.
down of ingested food into a chyme, which is delivered, at a Today, several methods have been developed to evaluate
rate controlled by hormonal, neural, and metabolic mecha- either gastric volume or emptying. Scintigraphy using radiola-
nisms, to the duodenum for enzymatic degradation into molec- belled meals represents the gold standard for the measurement
ular components and absorption through the gut wall. The rate of gastric emptying either directly or via breath tests; however,
of gastric emptying is a function of the pressure gradient be- differing protocols between centres hamper its interpretation.
tween the stomach (mainly antral contraction) and duodenum Ultrasound can be used to evaluate gastric volume and motility
(pyloric resistance). Antral contraction is influenced by gastric with results that correlate well with scintigraphy; however, it re-
volume, the secretion of gut hormones, and the composition of quires an experienced operator. Magnetic resonance imaging
chyme entering the duodenum. However, gastric motility dif- (MRI) has been used to evaluate gastric residual volume, but its
fers between the fed and fasting states. use is limited by expense and availability. Other methods include
the paracetamol absorption test, wireless motility capsules,
Fed gastric motility antropyloroduodenal manometry, and impedance monitoring.8
After a meal, gastric distension stimulates a vago-vagal excit-
atory reflex that enhances antral activity both directly and by
the stimulation of gastrin. The composition of duodenal chyme
Determinants of gastric volume and emptying
further affects the rate of gastric emptying by influencing the We continue to use residual gastric volume as a surrogate for
release of inhibitory gut hormones. For example, acidic, hyper- pulmonary aspiration, although neither a direct link has been
osmolar contents stimulate secretin, which directly inhibits demonstrated nor a critical volume accurately determined.
gastric muscle contraction; fat and protein by-products Furthermore, despite an empty stomach, aspiration can still

BJA Education | Volume 17, Number 10, 2017 347

Preoperative fasting in children

Table 1 Factors affecting gastric emptying

Factors increasing emptying Factors decreasing emptying

Physiological factors Large gastric volume Large duodenal volume

Liquid gastric contents High-calorie chyme
Solids <2 mm Acidic chyme
Hypo-/hyper-osmolar chyme
Fatty and amino acid-rich chyme
Hot and cold chyme
Parasympathetic stimulation Sympathetic stimulation
Secretion of motilin and gastrin Secretion of cholecystokinin, secretin, somatostatin,
vasoactive intestinal peptide, and gastric inhibitory peptide
Sitting position (for non-caloric liquids) Left lateral position (for non-caloric liquids)
Pharmacological factors Anticholinergics Antimuscarinics
Metoclopramide Opioids
Domperidone
Erythromycin
Patient factors Hyperthyroidism Pain
Anxiety and stress
Trauma
Pregnancy
Alcohol ingestion
Hypothyroidism
Diabetes
Pyloric stenosis
Intestinal obstruction
Vagotomy

occur due to large volumes of fluid regurgitating from the small Solids and cow milk
and large intestines. Nevertheless, fasting protocols that mini- Cow’s milk separates into liquid and solid (curd) phases on mix-
mize gastric volume are believed to reduce pulmonary ing with gastric acid, and therefore, its emptying is biphasic
aspiration.9 with a rapid initial liquid phase followed by a second-, zero-
Gastric volume is influenced by several factors: order, solid phase. This is why it requires a 6 h fast in common
with other solids. Full meals, with high fat content, may not
(i) Saliva and gastric secretions: During fasting, salivary se-
empty fully even at 8–9 h.
cretions contribute 1 ml kg1 h1 and gastric secretions
0.6 ml kg1 h1 to gastric volume.10
(ii) The rate of gastric emptying: Factors affecting emptying
Quantifying the risk in children
are summarized in Table 1. Premature and full-term neo-
nates are often reported to have slower gastric emptying Paediatric fasting guidelines are intended to reduce the risk of
in comparison with older children and adults due to im- pulmonary aspiration and facilitate the safe and efficient con-
mature neuromodulation of gastric motility. However, a duct of anaesthesia. However, there are numerous benefits
meta-analysis of 1457 patients, from premature neonates when children are fasted before operation as briefly as possible,
to adults, found that age was not a significant determi- including improved patient and parental satisfaction, increased
nant of gastric emptying.11 gastric pH, ingestion of calories, decreased risk of hypoglycae-
(iii) The timing and type of last oral intake. mia, decreased lipolysis, and improved fluid homeostasis.

Clear fluids, breast, and formula milk

How long do children actually fast?
Clear fluids rapidly empty from the stomach within 30 min. Almost all national guidelines advocate the 6, 4, and 2 h regi-
Sweetened drinks are slower depending upon the type of sug- men for clear fluids, breast milk, and solids, respectively; how-
ar—fructose, sucrose, and galactose empty faster than glucose. ever, children continue to be fasted for significantly longer.
Drinks higher in calories and osmolality delay emptying; how- A UK study of 1350 children presenting for dental procedures re-
ever, these differences do not seem clinically relevant. vealed median fasting times of 12.08 h and 7.95 h for solids and
Breast milk is emptied faster than formula milk due to a liquids, respectively, with the majority of children reporting
higher whey to casein ratio and therefore behaves more like feeling ‘extremely hungry’ or thirsty on admission.12 Similarly,
water than cow’s milk. However, its high lipid content slows a US study of 219 children found that mean fasting from liquids
gastric emptying in comparison with clear fluids. The emptying was 10.44 h, breast milk 8.3 h, and solids 10.62 h.13 Possible rea-
of formula milk is variable and depends upon its composition. sons for these extended fasts may have been the children being
Whey-predominant formulae empty faster than casein-rich for- fed earlier in the evening on the night before surgery or alter-
mulae and both are significantly faster than cow’s milk. As with ations in the operating schedule. It may, therefore, be better to
clear fluids, an increase in acidity, osmolality, and fatty acid advise parents to give their children food and drink at specific
concentration will slow emptying.9,10 times rather than instructions for fasting at 2 h and 6 h before

348 BJA Education | Volume 17, Number 10, 2017

 Preoperative fasting in children

operation, accepting that this may reduce list flexibility. Parents were not found to have higher gastric volumes or lower gastric
also seem understandably reluctant to wake children for drinks pH than those fasted longer. These children were also less
and are anxious about delaying the start of the anaesthetic so thirsty, less hungry, and less irritable than those who fasted for
err on the side of caution and avoid any intake. more than 6 h.21

What are the metabolic effects of fasting and surgery?

Should children fast at all?
During fasting, metabolism slows and the primary source of glu-
From a practical perspective, the 6-4-2 fasting regimen is only
cose gradually becomes hepatic glycogenolysis. As hepatic gly-
possible for the first child on the list as the subsequent children
cogen stores are depleted, hepatic gluconeogenesis and lipolysis
are given only approximate fasting times. With emerging evi-
ensue with subsequent fatty acid beta-oxidation and ketogene-
dence of the safety of liberalizing clear fluid ingestion to either
sis becoming the main energy sources. Indeed, ketoacidosis has
1 h or up to the time of anaesthesia, fasting times can be re-
been demonstrated in children less than 3 years of age fasted
duced even further. In a review of 10 015 children allowed un-
more than 7 h with a subsequent reduction in ketone bodies and
limited intake of clear fluids up to the time of general
hypotension on induction of anaesthesia when an optimized
anaesthesia, Andersson et al.19 found the incidence of pulmo-
fasting schedule was introduced.14
nary aspiration to be 0.03% with an average fasting time of 1.7 h.
In contrast to the reduction in metabolic rate seen with fast-
This compares very favourably with the aspiration incidence
ing, the trauma of surgery triggers a neuroendocrine stress re-
for the less liberal regimens quoted above and allows greater
sponse that increases metabolism, which further depletes
flexibility in the operating list. Also, 1 h clear fluid fasting may
hepatic and skeletal muscle glycogen stores and releases free
not significantly influence gastric volume or pH compared with
fatty acids and amino acids from the adipose tissue and skeletal
2 h,22 with gastric emptying occurring with a median half-life of
muscle, respectively. A key feature of this perioperative shift of
<30 min when measured by MRI.23 Furthermore, fasting status
metabolism has been found to be the development of insulin
may not be an independent predictor of aspiration or related
resistance, the severity of which is proportional to the degree of
adverse events such as unplanned admission, cardiac arrest, or
surgical trauma.15 The development of insulin resistance im-
death according to a prospective analysis of 139 142 children
pairs its anabolic effects and is a known risk factor for the de-
anaesthetized or sedated outside the main operating complex
velopment of postoperative complications and increasing
by different practitioners, including anaesthetists, emergency
length of hospital stay.15,16
physicians and paediatric intensivists.20
Finally, compliance with preoperative fasting by children
Preoperative carbohydrate drinks and parents is incomplete due to a variety of reasons such as
Providing non-caloric clear fluids, such as water, 2 h prior to sur- misunderstanding what fasting entails, the reason for fasting,
gery does not provide the substrates required to change the met- inadequate supervision, or deliberately misleading to avoid de-
abolic effects of fasting and surgery. Therefore, the main lays. Liberalizing preoperative oral intake may reduce these
objective of preoperative carbohydrate drinks is to stimulate an breaches.24
insulin response similar to that of a regular meal and therefore
switch the preoperative fasted state described above into a fed
state with normal postprandial insulin levels and minimal glyco-
Summary
gen store depletion. While there are physiological data to support
preoperative carbohydrate drinks, studies investigating the type There is increasing recognition that a prolonged preoperative
of drink and its clinical impact in children are limited.16 fast is not desirable let alone advantageous. Gastric physiology
is under a complex set of control factors that combine to ensure
a steady release of nutrients to the small bowel and beyond. For
How common is aspiration in children? clear fluid, there is good evidence in children that emptying can
Perioperative pulmonary aspiration in children remains infre- occur well within the advocated 2 h guidelines. Even when a
quent, is more likely in emergency rather than elective surgery, rare clear fluid aspiration occurs, the consequences do not ap-
and serious respiratory complications are rare (Table 2).3,6,17–20 pear to be severe or long term. These same guidelines, when
An updated Cochrane review of 25 trials involving 2543 children strictly adhered to, often result in much longer clear fluid fast-
found only one reported incidence of perioperative aspiration. ing times than are desirable. More liberal clear fluid regimens
Furthermore, children allowed to drink 2 h before operation for elective cases seem to confer no increased risk of aspiration

Table 2 Studies examining the incidence of paediatric aspiration and its complications. All studies considered aspiration to have occurred if
non-respiratory material was visualized in the larynx or tracheobronchial tree on laryngoscopy, bronchoscopy, or suctioning or, alternatively,
if clinical and/or radiological signs of aspiration developed. No deaths were reported in any study

Study Time period Study design Study size Incidence of Incidence of postoperative
aspiration events (%) ventilation among children
who aspirated (%)

Borland et al.17 1988–93 Retrospective 50 880 52 (0.1%) 4 (7.7%)

Warner et al.18 1985–97 Prospective 56 138 24 (0.04%) 6 (25%)
Walker3 2010–11 Prospective 118 371 24 (0.02%) 5 (20.8%)
Andersson et al.19 2008–13 Retrospective 10 015 3 (0.03%) 0 (0%)
Tan and Lee6 2000–13 Retrospective 102 425 22 (0.02%) 2 (9.1%)
Beach et al.20 2007–11 Prospective 139 142 10 (0.007%) Not reported

BJA Education | Volume 17, Number 10, 2017 349

Preoperative fasting in children

without subjecting large numbers of children and carers to the elective outpatient pediatric patients. Pediatr Anesth 2011;
distress of a prolonged fast. 21: 964–8
13. Williams C, Johnson PA, Guzzetta CE et al. Pediatric fasting
Declaration of interest times before surgical and radiologic procedures: bench-
marking institutional practices against national standards. J
None declared.
Pediatr Nurs 2014; 29: 258–67
14. Dennhardt N, Beck C, Huber D et al. Optimized preoperative
MCQs fasting times decrease ketone body concentration and stabi-
The associated MCQs (to support CME/CPD activity) can lize mean arterial blood pressure during induction of anes-
be accessed at http://www.oxforde-learning.com/journals/ by thesia in children younger than 36 months: a prospective
subscribers to BJA Education. observational cohort study. Pediatr Anesth 2016; 26: 838–43
15. Nygren J, Thorell A, Ljungqvist O. Preoperative oral carbohy-
Podcasts drate therapy. Curr Opin Anaesthesiol 2015; 28: 364–9
16. Ljungqvist O. Preoperative fasting and carbohydrate treat-
This article has an associated podcast which can be accessed at
ment. In: Feldman LS, Delaney CP, Ljungqvist O and Carli F,
http://dx.doi.org/10.1093/bjaed/mkx021. ed. SAGES/ERASV R Soc Man Enhanc Recover Programs Gastrointest

Surg. Cham, Switzerland: Springer International Publishing,

References 2015; 41–9
1. Leigh MD, Belton MK. Pediatric Anesthesia. New York: 17. Borland LM, Sereika SM, Woelfel SK et al. Pulmonary aspira-
Macmillan, 1948 tion in pediatric patients during general anesthesia: inci-
2. Maltby JR. Fasting from midnight—the history behind the dence and outcome. J Clin Anesth 1998; 10: 95–102
dogma. Best Pract Res Clin Anaesthesiol 2006; 20: 363–78 18. Warner MA, Warner ME, Warner DO et al. Perioperative pul-
3. Walker RWM. Pulmonary aspiration in pediatric anesthetic monary aspiration in infants and children. Anesthesiology
practice in the UK: a prospective survey of specialist pediatric 1999; 90: 66–71
centers over a one-year period. Pediatr Anesth 2013; 23: 702–11 19. Andersson H, Zarén B, Frykholm P. Low incidence of pul-
4. Mendelson C. The aspiration of stomach contents into the monary aspiration in children allowed intake of clear fluids
lungs during obstetric anesthesia. Am J Obstet Gynecol 1946; until called to the operating suite. Pediatr Anesth 2015; 25:
52: 191–205 770–7
5. Roberts RB, Shirley MA. Reducing the risk of acid aspiration 20. Beach ML, Cohen DM, Gallagher SM, Cravero JP. Major
during cesarean section. Anesth Analg 1974; 53: 859–68 adverse events and relationship to nil per os status in pedi-
6. Tan Z, Lee SY. Pulmonary aspiration under GA: a 13-year au- atric sedation/anesthesia outside the operating room: a
dit in a tertiary pediatric unit. Pediatr Anesth 2016; 26: 547–52 report of the pediatric sedation research consortium.
7. Hellström PM, Gryba € ck P, Jacobsson H. The physiology of Anesthesiology 2016; 124: 80–8
gastric emptying. Best Pract Res Clin Anaesthesiol 2006; 20: 21. Brady MC, Kinn S, Ness V, O’Rourke K, Randhawa N, Stuart
397–407 P. Preoperative fasting for preventing perioperative com-
8. Rayner CK, Hebbard GS, Horowitz M. Physiology of the astral plications in children. In: Brady MC, ed. Cochrane Database
pump and gastric emptying. In: Johnson LR, ed. Physiology of Syst Rev. 2009. DOI: 10.1002/14651858.CD005285.pub2. http://
the Gastrointestinal Tract, 5th Edn. London: Elsevier, 2012; 1: www.cochranelibrary.com (accessed November 1, 2016)
959–76 22. Schmidt AR, Buehler P, Seglias L et al. Gastric pH and resid-
9. Cook-Sather SD, Litman RS. Modern fasting guidelines in ual volume after 1 and 2 h fasting time for clear fluids in
children. Best Pract Res Clin Anaesthesiol 2006; 20: 471–81 children. Br J Anaesth 2015; 114: 477–82
10. Splinter WM, Schreiner MS. Preoperative fasting in children. 23. Schmitz A, Kellenberger CJ, Liamlahi R, Studhalter M, Weiss M.
Anesth Analg 1999; 89: 80–9 Gastric emptying after overnight fasting and clear fluid intake:
11. Bonner JJ, Vajjah P, Abduljalil K et al. Does age affect gastric a prospective investigation using serial magnetic resonance
emptying time? A model-based meta-analysis of data from imaging in healthy children. Br J Anaesth 2011; 107: 425–9
premature neonates through to adults. Biopharm Drug Dispos 24. Cantellow S, Lightfoot J, Bould H, Beringer R. Parents’
2015; 36: 245–57 understanding of and compliance with fasting instruction
12. Engelhardt T, Wilson G, Horne L, Weiss M, Schmitz A. for pediatric day case surgery. Paediatr Anaesth 2012; 22:
Are you hungry? Are you thirsty?—Fasting times in 897–900

350 BJA Education | Volume 17, Number 10, 2017