Meredith Borland, FACEM, Emergency Physician, Deputy Director; Amanda Esson, BHSc., Research Assistant; Franz Babl, FRACP, Paediatric Emergency Physician; David Krieser, FRACP, Paediatric Emergency Physician.
Procedural sedation in children in the emergency department: A PREDICT study
Article first published online: 17 FEB 2009
© 2009 The Authors. Journal compilation © 2009 Australasian College for Emergency Medicine and Australasian Society for Emergency Medicine
Emergency Medicine Australasia
Volume 21, Issue 1, pages 71–79, February 2009
How to Cite
Borland, M., Esson, A., Babl, F. and Krieser, D. (2009), Procedural sedation in children in the emergency department: A PREDICT study. Emergency Medicine Australasia, 21: 71–79. doi: 10.1111/j.1742-6723.2008.01150.x
- Issue published online: 17 FEB 2009
- Article first published online: 17 FEB 2009
- Accepted 13 October 2008
- emergency department;
- procedural sedation
Objective: To investigate current procedural sedation practice and compare clinical practice guidelines (CPG) for procedural sedation at Paediatric Research in Emergency Departments International Collaborative (PREDICT) sites. This will determine areas for improvement and provide baseline data for future multicentre studies.
Methods: A questionnaire of specialist emergency physicians regarding demographics, general procedural sedation practice and specific sedation agents given to children. CPG for general sedation and sedation agents were obtained for each site.
Results: Seventy-five (71%) useable surveys returned from 105 potential respondents. Most commonly used agents were nitrous oxide (N2O) (75, 100%), ketamine (total 72, 96%; i.v. 59, 83% and i.m. 22, 31%) and midazolam (total 68, 91%; i.v. 52, 81%, oral 47, 73%, intranasal 26, 41% and i.m. 6, 9%). Sedation was used for therapeutic and diagnostic procedures. Forty-three (57%) used formal sedation records and sedation checklists and thirty-one (41%) respondents reported auditing sedations. Four sites ran staff education and competency programmes. Nine sites had general sedation CPG, eight for ketamine, nine for N2O, eight for midazolam (four parenteral, five oral and six intranasal) and three for fentanyl. No site had a guideline for propofol administration.
Conclusion: Procedural sedation in this research network commonly uses N2O, ketamine and midazolam for a wide range of procedures. Areas of improvement are the lack of guidelines for certain agents, documentation, staff competency training and auditing processes. Multicentre research could close gaps in terms of age cut-offs, fasting times and optimal indications for various agents.
Procedural sedation is an important component of the day-to-day practice of emergency medicine and has particular relevance in the paediatric population. It has generated a number of scientific studies and management recommendations as summarized in recent reviews.1–3 A review in 2002 by Everitt et al. reported paediatric sedation practice as reported by directors of ED across both adult and paediatric ED in Australia and New Zealand.4 This study sought to quantify paediatric sedation practice, but stopped short of analysing individual physician practice. It did, however, demonstrate a deficiency in the provision of clinical practice guidelines (CPG), particularly in mixed ED.
The Paediatric Research in Emergency Departments International Collaborative (PREDICT) research network (Table 1) has identified procedural sedation as an area for further prospective interventional studies but, before designing these studies, sought to understand the current practices in this area across the network.
|Royal Children's Hospital, Melbourne (Australia)|
|Princess Margaret Hospital, Perth (Australia)|
|Children's Hospital at Westmead, Sydney (Australia)|
|Women's and Children's Hospital, Adelaide (Australia)|
|Mater Children's Hospital, Brisbane (Australia)|
|Starship Hospital, Auckland (New Zealand)|
|Sydney Children's Hospital, Sydney (Australia)|
|Royal Children's Hospital, Brisbane (Australia)|
|†Sunshine Hospital, Melbourne (Australia)|
|†Kidz First Hospital, Auckland (New Zealand)|
|†Monash Medical Centre, Melbourne (Australia)|
The primary objective of the present study was to investigate current procedural sedation practice of senior physicians in the ED of the PREDICT sites. Our secondary objective was to compare CPG for procedural sedation across sites with physician practice. This information will determine areas for improvement and provide a baseline for future multicentre studies.
All paediatric emergency physicians at participating PREDICT sites were eligible to participate in the survey. Data were collected from the 11 PREDICT sites, with a total annual ED census of greater than 366 000.5 The number of emergency physicians at each site was identified by the PREDICT site representatives. The electronic survey was disseminated to all specialist paediatric emergency physicians by the site representatives. Reminders were sent to the site representatives weekly for a period of 6 weeks and then fortnightly for a period of 4 weeks, from August to October 2006.
The questionnaire was anonymous but site-coded. Information was sought regarding demographics, general procedural sedation practice and specific practices relating to the use of ketamine, nitrous oxide (N2O), midazolam, propofol, fentanyl and others (if applicable). These items included procedures, routes used, initial and maximum doses, age limitations, fasting status and monitoring (cardiac rhythm trace, non-invasive blood pressure [NIBP], oxygen saturations and others [if applicable]) for sedation episodes. Physicians were requested to respond from the perspective of personal practice rather than perceived departmental practice. Most items were closed multiple responses, with others free-text. CPG for general sedation as well as all specific sedation agents, if any, were requested from the site representatives at each of the 11 PREDICT sites.
spss (version 14.0) was used to descriptively analyse questionnaires and CPG items. All CPG were submitted by March 2007.
This project was approved by the Ethics in Human Research Committee at Princess Margaret Hospital for Children, Perth, Western Australia, Australia and conforms to the provisions of the Declaration of Helsinki (as revised in Edinburgh, 2000).6
Seventy-seven (73%) surveys were returned from 105 physicians. Two surveys were unusable because of electronic errors, leaving 75 (71%) surveys. Among them, 35 (47%) physicians were Fellows of the Royal Australian College of Physicians only, 26 (35%) were Fellows of the Australasian College for Emergency Medicine only and 13 (17%) were Fellows of both colleges. One physician did not state his/her college affiliation. Of all, 13 (17%) had more than 10 years of experience as a specialist physician, 21 (28%) had 5–10 years of experience, 21 (28%) had 2–5 years of experience and 20 (27%) had up to 2 years of experience. Forty-eight (64%) physicians were employed full-time.
Consent was always obtained from the parent/child before procedural sedation regardless of agent being used by 16 (21%) of respondents whereas 12 (16%) never obtained consent. Of the 46 (61%) who did not always obtain consent, 39 (85%) obtained consent for procedures involving i.v. ketamine, 32 (70%) obtained consent for i.m. ketamine and 31 (67%) obtained consent for other i.v. drugs. Clinical notes were the most commonly used documentation for procedural sedation (61, 81%), followed by medication chart (59, 79%). A formal sedation record sheet and a sedation checklist were used by 43 (57%) of participants.
Four sites indicated that their ED ran a specific staff competency programme for procedural sedation. Competency programmes were identified to be for both doctors and nurses by 26 (59%) respondents, for nurses only by 10 (23%) and for doctors only by 6 (14%). The competency programmes consisted of a multiple-choice questionnaire and oral competency test as reported by 16 (36%) respondents, a lecture by 14 (31%) and an airway training session by 9 (20%).
Discharge information was by pre-printed specific discharge instructions (55, 73%), by verbal instructions (43, 57%) and by handwritten or typed instructions in 11 (15%) physicians.
Thirty-one (41%) respondents and four sites reported auditing sedation practices/procedures, twenty-three (31%) respondents and two sites said that they did not and twenty (27%) respondents and five sites were unsure.
Nitrous oxide (N2O), ketamine and midazolam were reported to be the most commonly used agents for procedural sedation in the ED. This was followed by fentanyl and propofol (Fig. 1). Other agents nominated by the respondents included oral chloral hydrate, anaesthetic gases, etomidate, sucrose and morphine, but further details of these agents' use were not sought in the survey. Etomidate is currently only available under special access scheme in Australia and New Zealand.
Site CPG comparison
All PREDICT sites provided sedation-specific CPG. Nine sites had a general sedation CPG, nine for N2O administration, eight for ketamine administration, eight had a CPG specific for midazolam (four specific for parenteral midazolam, five for oral midazolam and six for intranasal [i.n.] midazolam), three for fentanyl and one had a CPG for drug combinations. Significantly, no site had a CPG for propofol administration.
Seventy-five (100%) physicians reported using N2O for procedural sedation, with one physician nominating N2O use failing to complete further details in the survey. It was most frequently administered for wound/burns dressings (65, 87%) and application of plaster/splints (64, 85%) (Table 2). N2O was most commonly administered using premixed (50 : 50) cylinder (42, 57%). A continuous flow device, which delivers variable concentration of N2O from 0 to 70% irrespective of patient respiratory effort, was used by 27 (36%). The median concentration of N2O delivered was 50% and maximum concentration was 70%.
|Procedures||Ketamine n= 72 (96%)||Nitrous oxide n= 75 (100%)||Midazolam n= 68 (90%)||Propofol n= 18 (14%)||Fentanyl n= 48 (64%)|
|Laceration repair||71 (99)||57 (76)||45 (66)||1 (5)||16 (33)|
|Fracture manipulation||69 (96)||58 (77)||30 (44)||7 (39)||29 (60)|
|Fracture splint/plaster||33 (46)||64 (85)||28 (41)||1 (5)||22 (46)|
|Wound/burn management||46 (64)||65 (87)||33 (49)||2 (11)||25 (52)|
|Urine sampling||4 (5)||22 (29)||13 (19)||0||2 (4)|
|Lumbar puncture||15 (21)||39 (52)||38 (56)||0||6 (12)|
|i.v. insertion||5 (7)||58 (77)||24 (35)||0||6 (12)|
|Chest drain insertion||21 (29)||13 (17)||21 (31)||4 (22)||25 (52)|
|Joint relocation||47 (65)||46 (61)||31 (46)||7 (39)||29 (60)|
|Other||11 (15)||12 (16)||15 (22)||1 (5)||1 (2)|
Six CPG specified an upper/lower age limit. Forty-seven (63%) reported using age limits (Table 3), but it is worth noting that fifteen physicians stated that the use of the drug is more dependent on the child's ability to use the device or hold the mouthpiece independently, rather than on the age of the child.
|Initial dose (mg/kg)||Ketamine n= 72 (96%)||Nitrous oxide n= 75 (100%)||Midazolam n= 68 (90%)||Propofol n= 18 (14%)||Fentanyl n= 48 (64%)|
|i.v. n= 59 (83%)||i.m. n= 22 (31%)||‡||i.v. n= 52 (81%)||i.n. n= 26 (41%)||i.m. n= 6 (9%)||Oral n= 47 (73%)||i.v.||i.v.†|
|Median (range)||1.0 (0.5–4.0)||4.0 (1.0–4.0)||50% (50–70%)||0.1 (0.05–0.5)||0.3 (0.1–0.6)||0.1 (0.1–0.2)||0.5 (0.2–1.0)||0.5 (0.5–2.0)||1.0 (0.2–3.0)|
|Total maximum dose (mg/kg)|
|Median (range)||2.2 (1.5–8.0)||5.5 (2.0–8.0)||70% (50–70%)||0.2 (0.1–0.6)||0.4 (0.1–1.0)||0.6 (0.3–1.0)||0.75 (0.1–15§)||2.0 (0.1–5.0)||2.0 (0.5–5.0)|
|Present||59 (82%)||47 (63%)||30 (58%)||16 (62%)||4 (67%)||20 (43%)||7 (39%)||20 (42%)|
|Median (range)||1 year (3 months–2 years)||2 years (6 months–5 years)||1 year (1 month–11 years)||1 year (6 months–2 years)||6 months (1 month–1 year)||1 year (3 months–2 years)||10 years (3–13 years)||1 year (2 months–5 years)|
|Oximetry||71 (99%)||61 (81%)||51 (98%)||22 (85%)||6 (100%)||37 (79%)||11 (61%)||37 (77%)|
|Cardiac monitoring||63 (88%)||10 (13%)||31 (60%)||7 (27%)||2 (33%)||8 (17%)||10 (56%)||29 (60%)|
|NIBP||58 (81%)||11 (15%)||27 (52%)||5 (19%)||1 (17%)||6 (13%)||11 (61%)||26 (54%)|
|Present||54 (75%)||27 (36%)||22 (26%)||16 (34%)||10 (56%)||12 (25%)|
|Solids||Median (range)||4 h (2–6 h)||2 h (0–6 h)||4 h (1–6 h)||2 h (2–6 h)||4 h (3–6 h)||4 h (2–6 h)|
|Liquid||Median (range)||3 h (1–4 h)||2 h (1–3 h)||2 h (1–4 h)||2 h (1–4 h)||2 h (1–4 h)||3 h (2–4 h)|
Only 27 (36%) reported a fasting time for N2O (Table 3). Forty-eight (65%) physicians agreed with their CPG on the use of fasting times for this agent. Monitoring with N2O was mostly by oximetry (61, 81%), but 13 (17%) physicians were not using any monitoring. Thirty-six (49%) agreed with their CPG on the initial dose of N2O. Median difference in initial dose for those who disagreed was 20% N2O (interquartile range [IQR]= 0).
Seventy-two (96%) physicians reported the use of ketamine for procedural sedation, with one physician nominating ketamine use but failing to complete further details in the survey. It was used by all respondents for laceration repair (99%) and fracture manipulation (96%) (Table 2). The median initial and total doses prescribed, fasting times and monitoring methods are presented in Table 3. Eighteen (25%) physicians indicated that they routinely used atropine and ketamine concurrently. Of 70, 7 (10%) routinely used ketamine and midazolam concurrently.
Thirty (42%) agreed with their CPG on the initial dose of i.v. ketamine, whereas nine (13%) agreed with the initial i.m. dose. Median difference in initial dose for those who disagreed was 0.5 mg/kg (IQR = 0.5 mg/kg) for i.v. and 1.0 mg/kg (IQR = 1.0 mg/kg) for i.m. Fifty-six (79%) physicians agreed with their CPG on the use of fasting times for this agent.
Eight CPG recommended the use of an age limit or specified an upper or lower age limit (eight for i.v. and i.m., and three for oral); however, there were nine (13%) physicians using ketamine who did not use an age limit.
Sixty-eight (90%) reported that they used midazolam for procedural sedation episodes, but four physicians did not provide details of their use. It was most frequently used for laceration repair (66%) and lumbar punctures (56%) (Table 2). Eighteen (26%) physicians specified routinely using midazolam and fentanyl concurrently.
Only three (5%) agreed with their CPG on the initial dose of oral midazolam (median difference 0.5 µg/kg, IQR 0.06 µg/kg) and two (4%) agreed with the initial i.v. dose (median difference 0.45 µg/kg, IQR 0.45 µg/kg). For i.m. midazolam no physicians giving an initial dose had a corresponding CPG.
Two CPG specified an upper and/or lower age limit (two for i.v. and oral, and one for i.m. and i.n.). With respect to physician use of age limits, there were 17 (36 %), 16 (31%), 4 (15%) and 2 (33%) for oral, i.v., i.n. and i.m., respectively, who did not use an age limit. Sixteen (34%) physicians agreed with their CPG on the use of fasting times for oral midazolam and 18 (35%) for i.v. midazolam.
The median initial dose was 0.1 mg/kg with a median total dose of 0.2 mg/kg (Table 3). Thirty physicians (58%) used a lower age limit for i.v. midazolam, with 26% recommending fasting times for parenteral midazolam. Monitoring methods for i.v. midazolam included oximetry, NIBP and cardiac monitoring (Table 3).
Initial and total doses were higher than for i.v. midazolam. Children were less monitored using i.n. midazolam as compared with i.v. midazolam. Of 26 physicians, 4 reported no monitoring with this method.
Median initial dosing was 0.5 mg/kg (range 0.2–1.0), with a number of physicians reporting a maximum total dose of 15 mg (10, 24%) (Table 3). Twenty (43%) reported age limitations for oral midazolam, with the majority (90%) limiting use to over 1 year old. Oximetry was used for monitoring by 37 respondents, whereas eight reported no monitoring.
Forty-eight (64%) physicians reported using fentanyl, but eleven did not complete details of their use of this agent. The route of administration used was i.v. in 32 (87%), i.n. in 15 (41%) and i.m. in 2 (5%). Fentanyl was used more commonly for therapeutic orthopaedic procedures (Table 2). The dose, fasting times and age limitations are presented in Table 3. Most physicians used oximetry, NIBP and cardiac monitoring for fentanyl sedation.
Eighteen (49%) physicians using fentanyl had a CPG for their sites. Twelve (32%) agreed with their CPG on the initial dose of fentanyl (ten with 1.0 µg/kg and two with 2.0 µg/kg). Median difference in initial dose for those who disagreed was 1.0 µg/kg (range 0.5–2.0 µg/kg, IQR 0.8 µg/kg). None of these CPG recommended the use of an age limit or specified an upper or lower age limit.
There were nine (24%) physicians using fentanyl who did not use an age limit. Thirty (81%) physicians agreed with their CPG on the use of fasting times for this agent.
Eighteen physicians (14%) reported the use of propofol for procedural sedation, but seven did not specify any details of their use of this agent. Seven used it for fracture/joint reductions (Table 2). The median dose, age limits and fasting times are presented in Table 3. No site had a CPG for propofol.
Oral chloral hydrate (18 respondents), 25% oral sucrose solution (four respondents), anaesthetic gases (three respondents) and etomidate (one respondent) were all agents indicated as potential other agents considered for use by the ED physicians. Details of their use were not explored in the survey and no site provided CPG for these agents.
The present study was undertaken to determine the practice of paediatric emergency physicians when administering procedural sedation to children in the ED across a number of institutions. Our study has shown that procedural sedation is now being used in a variety of conditions within the ED, both for diagnostic purposes, such as urine sampling and lumbar punctures, and therapeutic purposes, such as i.v. insertion, wound care and orthopaedic trauma.
When studying the individual agents being used for sedation, there have been some significant changes in the usage patterns of ketamine since the report by Everitt.4 Almost all physicians (83%) used ketamine i.v. and 31% used ketamine i.m. in our survey, whereas in Everitt's study only 33% used ketamine i.v. and 66% i.m. Parenteral ketamine has been shown to be a safe and efficacious agent in the ED setting.7,8 There was a wide range in the dosing of ketamine i.v. or i.m. across the PREDICT sites, suggesting some confusion regarding the appropriate dose for each route of administration.
For procedures using midazolam, the i.v. route is used more frequently than the oral or i.n. route. When used orally, many physicians set a maximum dose of 15 mg. The initial oral dose range was wide, from 0.2 mg/kg to 1.0 mg/kg with a median of 0.5 mg/kg. There is some evidence that lower doses of oral midazolam might result in paradoxical hyperactivity.9,10 Oral midazolam is more difficult to titrate if inadequate initial doses are given. A higher initial dose might give a more reliable clinical response and ensure better patient cooperation during the procedure.9
In order to safely provide sedation in the ED, it is important to have staff educated and trained in the appropriate procedures, although there is controversy in the literature on how this should be achieved.1,11–14 Only four PREDICT sites run specific sedation competency programmes for their staff with lectures and competency packages for nursing and medical staff, although all sites have access to courses, such as Advanced Paediatric Life Support. Credentialing is also a requirement of these sites before staff undertaking procedural sedation. Recent sedation credentialing programmes have been developed in Australia to promote this and have been shown to improve proxy markers of sedation safety.5,15,16 These proxy markers included improvement in documentation of fasting status, consent, formal risk assessment, written drug orders, appropriate vital signs and depth of sedation.
We noted inconsistency in obtaining formal consent. Consent might be difficult to obtain in the setting of paediatric emergency medicine because of the acuteness of presenting problems and the age of the patients.13 Failure to adequately inform the child's guardian has significant medico-legal implications for the treating physicians in the event of an adverse event.
Fasting for ED procedural sedation has been controversially discussed in the literature, with some recent papers questioning the need for fasting in ketamine and N2O sedation episodes in the ED.11,17–21 Recent anaesthetic literature suggests some liberalization of fasting times, even for general anaesthesia.22,23 This liberalization particularly applies to the paediatric age groups. Unified fasting times are difficult to proscribe considering the variety of agents used and different sedation depth required.24 This is reflected in our study's findings indicating a wide variety of fasting times recommended and practised for the various agents used. The most consistent fasting time was for i.v. propofol, but this agent was only infrequently used by 11 physicians in our study.
Monitoring was more intensive for i.v. sedation episodes, with most physicians using cardiac monitoring, oximetry and NIBP. For nitrous sedation, no monitoring was reported to be used by many physicians. No respondents reported using capnography to monitor ventilation despite recent evidence suggesting that this is a useful tool to detect ventilatory deterioration at an earlier time than oximetry.25,26 This highlights a potential area for improvement in practice that could be explored in future research. More recently, bispectral monitoring has been suggested as a method of measuring depth of sedation. The evidence for this monitoring method in the ED suggests that it is unreliable at detecting depth of sedation.11,27–29 This method of monitoring was not reported by any respondents.
The age of the patient, particularly the child less than 12 months, was quoted by a number of respondents as a limiting factor for the provision of procedural sedation. Inhaled N2O, in particular, requires a certain tidal volume and cooperation of the patient. Propofol, although used infrequently, tended to be limited to use in older children. Some respondents also limited the use of ketamine to children less than 12 years. This might be related to concerns of emergence phenomenon, which is more frequent in older age groups.30 The setting of age limits for various agents is an area requiring more detailed research.
Documentation was also variable, with only 57% of physicians using a formal sedation record. These standardized sedation records encourage safe practice with respect to consent, monitoring, documentation and written discharge information.11 Formal auditing of sedation episodes was only undertaken by 41% of respondents, with the majority of these only auditing i.v. sedations. Auditing is a valuable tool of quality improvement and can provide important data for sedation research. This points to an area for potential future research to establish safety and consistency of sedation events across the PREDICT sites.
All departments have CPG either for general sedation or for specific agents; however, there was considerable variability in the depth of information presented. The majority of site respondents did appear to follow their CPG when available, although the survey design limited the capacity to fully assess this variable. The development of evidence-based, joint CPG for all PREDICT sites, especially if developed with coordinated education and credentialing packages, would reduce duplication and would be anticipated to be well received. Ideally, prospective auditing of all sedations should be a routinely used tool for quality assurance.
The survey response rate was only 73% despite enrolling the PREDICT site representative to encourage return of the surveys and with repeated reminders to the respondents via email. There might have been a perception that the survey responses could be linked to individuals despite it being coded only for site. The individual site representatives nominated the number of expected surveys per site, but might not have taken into account staff being unavailable to complete surveys. In addition, there were some incomplete surveys that might have reflected insufficient time to complete the survey.
Although nominally anonymous, the site coding of the survey responses and collection of the survey bythe site representatives might have placed increased pressure on physicians to follow their local CPG recommendations. It is difficult to assess whether survey responses correlate to clinical practice. A retrospective or prospective data review might gauge physician practice patterns more accurately. This survey was representative of senior physicians at each site who might not be the most active or representative of everyday practice at the site. This has prevented an understanding of the sedation practice by junior medical staff. In addition, physicians were asked only for their reported practice at one point in time, whereas practice evolves over time and guidelines might lag behind the development of newer practices.
We only surveyed specialist physicians at PREDICT sites and are unable to determine if the procedural sedation practices are reproducible at non-PREDICT sites, particularly in smaller mixed ED.
Procedural sedation by physicians in this research network is broadly similar with a focus on N2O and ketamine for a wide range of procedures. Areas for improvement include developing CPG for certain agents, improving completion of written sedation records and discharge information, improved auditing of sedation episodes and the development of staff credentialing programmes in some sites. Multicentre research could close evidence gaps in terms of age cut-offs, fasting times and optimal indications for various agents.
Jocelyn Neutze Kidzfirst, Middlemore Hospital (Auckland, New Zealand), Peter Ngo, Children's Hospital Westmead (New South Wales), Jason Acworth Royal Children's Hospital Brisbane (Queensland), Jacquie Schutz Women's and Children's Hospital (South Australia), Fiona Thomson Mater Children's Hospital (Queensland), Elizabeth Cotterell Sydney Children's Hospital (New South Wales), Sarah Jamison Starship Children's Hospital (Auckland, New Zealand), Peter Francis Monash Medical Centre (Victoria).
- 6World Medical Association. World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects 2004 9.10.2004 [Cited 6 Jun 2007.] Available from URL: http://www.wma.net/e/policy/b3.htm
- 13Scottish Intercollegiate Guidelines Network. Safe Sedation of Children Undergoing Diagnostic and Theraputic Procedures. 2004 April 2004 [Cited 5 Jun 2007.] Available from URL: http://www.sign.ac.uk