Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Does the time of fasting affect complication rates during ketamine sedation

Three Part Question

[In children undergoing ketamine sedation] is [prolonged fasting (6 hours or more) better than short term fasting (3 hours)] at [reducing the incidence of vomiting and other complications of sedation]

Clinical Scenario

A 4 year old boy is brought to the emergency department having fallen over at home. He has sustained a 3 cm deep laceration to the forehead. He was never unconcious and you have no concerns of an underlying brain injury. The wound clearly needs closure and cleaning but he is upset and would not be able to cooperate without sedation. You suggest this but his mother states that he ate 3 hours ago. You phone the anaesthetist on call who tells you that you should wait a further 3 hours to ensure that he is fasted. You wonder if this is really necessary.

Search Strategy

1. Ovid MEDLINE(R) 1950 to February Week 3 2008
2. Ovid MEDLINE(R) In-Process & Other Non-Indexed Citations February 29, 2008
(exp Vomiting/ or or exp "Postoperative Nausea and Vomiting"/) AND ( or exp Ketamine/ OR

Search Outcome

Medline 1: 219 papers of which two were relevant to the three part question.
Medline 2: 9 papers none of which were relevant.
One additional paper was known to the author.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Treston G
272 consecutive children undergoing ketamine sedation in the emergency department. Results available on 257. Authors examined the relationship between fasting time and incidence of vomiting. Patients were given IV ketamineProspective cohort studyIncidence of vomiting in children fasted less than one hour2/30 (6%)Patients fasted more than 3 hours are grouped together rather than discriminated at 6 hours. Further information was sought from the original author but data differentiating patients at 6 hours was not available.
Incidence of vomiting in children fasted 1-3 hours14/100 (14%)
Incidence of vomiting in children fasted more than 3 hours20/127 (15.7%)
Statistical trend of vomiting ratesNon significant trend towards increased vomiting with increased fasting time. p=0.08
Effect of age on vomiting ratesMore common in older age groups
Incidence of aspiration pneumonitisThere was no evidence of aspiration in any patients
Roback MG
Database of 2085 children undergoing procedural sedation by emergency physicians. Age range 19 days to 32.1 years in database. Median age was 6.7 years. Fasting time documented in 1555 patients. Analysis of complication rates was divided into 0 to 2, 2 to 4, 4 to 6, 6 to 8 and greater than 8 and not documented. Sedation options were ketamine (IV or IM), midazolam/ketamine, midazolam/fentanyl, midazolam, midazolam/morphine, other.Prospective cohortOverall incidence of serious adverse events172/2085 (8.2%) of all patients had a respiratory adverse event (hypoxia, desaturation, apnea, laryngospasm, aspiration)Incomplete data collection (a quarter of charts were not completed). Specific influence of fasting time on ketamine patients unknown as all agents reported as a group. No differentiation of food vs fluids fasting.
Use of ketamine57% of patients (1199) had ketamine as the sole sedative agent.
Adverse events when ketamine used alone70/1199 (5.8%) had respiratory adverse event. 129/1199 vomited.
Influence of fasting time to adverse events7.3% for 0-2 hours; 7.7% for 2-4 hours; 7.2% for 4-6 hours; 9.6% for 6-8 hours; 6.3% for >8 hours. p=NS
Respiratory adverse events for different agentsAt 5.8% ketamine had the lowest incidence of respiratory complications.
Vomiting with ketamine129/1199 (10.8%) of patients vomited with ketamine (highest incidence between agents range 0.9-10.8%)
IM or IV ketamine?1022 had IV ketamine, 177 had IM ketamine
Agrawal D, et al.
1014 patients undergoing procedural sedation in a paediatric emergency department. A variety of different agents were used but 474 (47%) had ketamine. Fasting time against established guidelines was recorded (<6 months 4- 6 hours, 6-36 months 6 hours, >36 months 6-8 hours). Or 2 hours for clear liquids in all age groups.Prospective case seriesData collection standards905 (98%) had fasting status recorded.Less than 14 patients aged less than 6 months in the study. 11% of patients did not have fasting time documented. Numerous agents used so difficult to determine if there are drug specific differences.
Number of minor adverse events (e.g. desaturation, emesis, aspiration, hypotension, laryngospasm)75 events in 66 patients. All succesfully treated with minor intervention (e.g. airway repositioning, suction etc.)
No of patients with evidence of pulmonary aspirationNo events (1 sided 97.5% confidence interval 0-0.4%)
No of patients requiring admission2 patients admitted following sedation. One had multiple medical problems and received pentobarbital. One had a generalised fit with ketamine 2 hours post injection.
Fasting status396/905 (44%) of patients met fasting status. 509/905 (56%) did not
Adverse events in relation to fasting status32/396 (8.1%, 95CI 5.6% to 11.2%) of fasted patients vs. 35/509 (6.9%, 95% CI 4.8% to 9.4%) for those not fasted. p=0.49
Adverse events in relation to time of fastingMedian fasting duration in patients with adverse events was 7.3 hours for solids and 6.6 hours for fluids vs. 6.8 hours and 6.0 hours for those without. p=0.13
Emesis in relation to time of fastingMedian duration fasting in patients with emesis was 6.8 hours for solids and 5.8 hours for fluids vs. 6.8 hours and 6.0 hours p=0.70


The issue of ketamine use in the emergency department is controversial despite overwhelming evidence regarding its efficacy and safety. Recent guidelines from the Scottish Intercollegiate Guidelines Network (SIGN) have suggested that children should be starved for more than 6 hours. However, British Association of Emergency Medicine (BAEM) and American College of Emergency Physicians (ACEP) guidelines suggest that 3 hours is adequate. The papers listed above support the view that prolonged fasting is unlikely to significantly reduce vomiting or affect other complications of sedation. Fasting is intended to reduce the incidence of pulmonary aspiration (hence the focus on vomiting rates). None of these papers had any patients with this complication and we are unaware of any reports of pulmonary aspiration following ED procedural sedation with ketamine in children. It must be noted that the incidence of vomiting is a proxy measure for aspiration of gastric contents. If vomiting occurs after the period of sedation (as is typically the case with ketamine) then it is unlikely to result in aspiration. However, what is clear, from these and other studies, is that the incidence of aspiration is low. The first paper examines the use of IV ketamine only. The applicability of the results to IM ketamine must be inferred.

Clinical Bottom Line

Prolonged fasting for ketamine sedation in children is unnecessary.

Level of Evidence

Level 2 - Studies considered were neither 1 or 3.


  1. Treston G Prolonged pre-procedure fasting time is unnecessary when using titrated intravenous ketamine for paediatric procedural sedation. Emergency Medicine Australia 2004;16:145-150
  2. Roback MG, Bajaj L, Wathen JE, Bothner J. Preprocedural fasting and adverse events in procedural sedation and analgesia in a pediatric emergency department: Are they related. Ann Emerg Med 2004:44:454-459
  3. Agrawal D, Manzi SF, Gupta R, Krauss B. Preprocedural fasting state and adverse events in children undergoing procedural sedation and analgesia in a pediatric emergency department. Ann Emerg Med 2003;42:636-646