2016

Background: Pediatric head injury (HI) clinical decision rules assist in balancing the ability of CTs to detect significant injuries and radiation associated risks. The three highest quality rules, PECARN, CATCH and CHALICE have not been prospectively validated and compared in a large multicenter sample.
Objective: To validate and compare the accuracy of PECARN, CATCH and CHALICE in a multicenter observational study external to their derivation. Methods: We prospectively enrolled patients less than 18 years with HIs of any severity and collected all predictor and outcome variables of the 3 rules. We assessed the accuracy of the rules based on published inclusion and exclusion criteria.
Results: Of 20,255 evaluable patients: 5,402 (26.7%) were < 2 years old; 7,348 (36.3%) were female; 354 (1.7%) had GCS ≤13; CT scans were performed on 2,136 (10.5%); 83 (0.4%) underwent neurosurgery; and 15 (0.07%) died. PECARN 2 years, CATCH and CHALICE were applicable in 4,095 (20.2%), 11,320 (55.9%), 4,974 (24.6%) and 20,147 (99.5%) patients respectively. Validation sensitivities were ranked as follows: PECARN2 years 106/110 (96.4%, 95% CI 91.0% to 99.0%), CATCH 25/26 (92.6%, 95% CI 80.4% to 99.9%) and CHALICE 375/405 (92.6%, 95% CI 89.6% to 94.9%). Negative Predictive Values for all rules were 99% to 100%.
Conclusions: The sensitivities of the PECARN rules were higher than CATCH or CHALICE. Data on the comparative accuracy of the rules to identify patients with clinically important traumatic brain injury will also be presented.

Background: The capacity to generate new evidence is only the first step in improving emergency care outcomes for children.  The second and arguably harder step is to translate new knowledge into practice.  How to most effectively shorten the time period between discovery of new knowledge and its routine wide spread use in the emergency department (ED), is a critically important goal.
Objectives: To systematically review clinical knowledge translation studies carried out in the ‘ED’ setting and focused on paediatric emergency care.  To describe and assess the study design, interventions used and impact on work process and clinical care.
Methods: Electronic databases were searched for knowledge translation studies conducted in the emergency department that included the care of children. Two researchers independently reviewed the studies.
Results: From 1305 publications identified, 15 studies of varied design were included. Four were cluster controlled trials, two patient-level randomized controlled trials; two interrupted time series, one descriptive study and six before and after intervention studies.  Knowledge translation interventions were predominantly aimed at the treating clinician, with some targeting the organisation. Studies assessed effectiveness of interventions over 4-12 months in before and after studies, and 5-28 months in cluster or patient level controlled trials. Changes in clinical practice were variable.
Conclusion: Evidence for effective methods to translate knowledge into practice in paediatric emergency medicine is fairly limited. More optimal study designs with more explicit descriptions of interventions and their rationale are needed to facilitate other groups to effectively apply these procedures in their own setting.

Background:The Paediatric Research in Emergency Departments International Collaborative (PREDICT) performs multicentre paediatric emergency research in Australia and New Zealand (ANZ). Research priorities are difficult to determine, often relying on individual interests or prior work.
Objective:To identify the research priorities of paediatric emergency medicine (PEM) specialists working in ANZ.
Methods:An online database administered surveys in a two-stage, modified Delphi study. Eligible participants were Consultants and senior advanced trainees in PEM from fourteen PREDICT sites. Participants submitted up to three of their most important research questions (Survey I). Responses were collated and refined, a short list of questions was returned to participants for prioritisation (Survey II). A further prioritisation exercise was carried out at a PREDICT members meeting, the Hanlon Process of Prioritisation was used to compare the priorities of active researchers in PEM.
Results:107/254 participants (42.5%) responded to Survey I, 144/245 (58.7%) to Survey II. 246 individual research questions were submitted. Survey II established a list of the top 35 research questions. Priority topics include high flow oxygenation in intubation, fluid volume resuscitation in sepsis, imaging in c-spine injury, intravenous therapy for asthma and vasopressor use in sepsis. There were some differences using the Hanlon process.
Conclusion:This modified Delphi has established a list of research questions which will inform multi-centre PEM research in ANZ. While some research topics were similar to results of North American and UK Delphi studies in PEM, some important differences in focus were identified.

Background: The aetiology and clinical course of Bell’s palsy is different in paediatric and adult patients.
There is no placebo controlled, blinded, randomised controlled trial (RCT) to show effectiveness of
prednisolone for Bell’s palsy in children.
Objective: To assess current practice in paediatric Bell’s palsy in Australia and New Zealand Emergency
Departments (ED) and determine the feasibility of conducting a multicentre RCT within the Paediatric
Research in Emergency Departments International Collaborative (PREDICT).
Design/Methods: A retrospective analysis of ED medical records of children less than 18 years diagnosed
with Bell’s palsy during between January 1, 2012 and December 31, 2013 was performed. Potential
participants were identified from ED information systems using Bell’s palsy related search terms. Repeat
presentations during the same illness were excluded but relapses were not. Data on presentation, diagnosis
and management was entered into REDCap.
Results: 323 presentations were included from 14 PREDICT sites. Mean age at presentation was 9.0(SD
5.0) years with 184(57.0%) females. Most (225, 74.0%) presented to ED within 72 hours of symptoms,
163(53.6%) had seen a doctor prior. In ED, 238(74.4%) were treated with steroids, usually prednisolone, for a mean of 8.8days at a mean starting dose of 1.1mg/kg/day, with 35.7% prescribed tapering. 33(10.2%) received antiviral treatment.
Conclusions: Treatment of Bell’s palsy in children presenting to Australasian EDs is varied. Prednisolone is
commonly used in Australasian EDs, despite lack of high level paediatric evidence. The study findings
confirm the need for and feasibility of a planned PREDICT RCT on prednisolone for Bell’s palsy in children.

Background:Pediatric head injury (HI) clinical decision rules assist in balancing the risk of CT use to detect significant injuries and radiation associated risks. The three highest quality rules, PECARN (Paediatric Emergency Care Applied Research Network), CATCH (Canadian Assessment of Tomography for Childhood Head Injury) and CHALICE (Children’s Head Injury Algorithm for the Prediction of Important Clinical Events) have not been prospectively validated and compared in a large multicenter sample.
Objective:To validate and compare the accuracy of PECARN, CATCH and CHALICE rules in a large multicenter observational study outside the derivation settings.
Methods:We prospectively enrolled patients aged less than 18 years with HIs of any severity and collected all predictor and outcome variables of the 3 rules. We assessed the accuracy of the rules as they were designed and analysed, based on published inclusion and exclusion criteria.
Results:Of 20,255 evaluable patients: 5,402 (26.7%) were less than 2 years old; 7,348 (36.3%) were female; 354 (1.7%) had GCS 13; CT scans were performed on 2,136 (10.5%); 83 (0.4%) underwent neurosurgery; and 15 (0.07%) died. PECARN <2 years, PECARN >2 years, CATCH and CHALICE were applicable in 4,095 (20.2%), 11,320 (55.9%), 4,974 (24.6%) and 20,147 (99.5%) patients respectively.Validation sensitivities were ranked as follows: PECARN<2 years 41/41 (100.0%, 95% CI 91.4% to 100.0%), PECARN>2 years 106/110 (96.4%, 95% CI 91.0% to 99.0%), CATCH 25/26 (92.6%, 95% CI 80.4% to 99.9%) and CHALICE 375/405 (92.6%, 95% CI 89.6% to 94.9%). Negative Predictive Values were: PECARN<2 years 2161/2161 (100.0%, 95% CI 99.8% to 100.0%), PECARN>2 years 5096/5100 (99.9%, 95% CI 99.8% to 100.0%), CATCH 2733/2734 (100.0%, 95% CI 99.8% to 100.0%), and CHALICE 9580/9610 (99.7%, 95% CI 99.6% to 99.8%).
Conclusions:The sensitivities of the PECARN rules were higher than CATCH or CHALICE when used strictly as designed. We will also present data on the comparative accuracy of the rules to identify patients with clinically important traumatic brain injury (ciTBI) in a single, cross-over cohort of children with undifferentiated head injuries.

Background: The Paediatric Research in Emergency Departments International Collaborative (PREDICT)
performs multicenter pediatric emergency research in Australia and New Zealand. Research topics often
rely on individual interests or prior work, rather than a systematic assessment of priorities from endusers
of new knowledge.
Objective: To identify the research priorities of paediatric emergency medicine (PEM) specialists
working in ANZ.
Design/Methods: An online database was used to administer surveys in a two-stage, modified Delphi
study. Eligible participants were attendings or senior advanced trainees in PEM from fourteen PREDICT
sites in Australia and New Zealand. Participants submitted up to three of their most important research
questions (Survey I). Responses were collated and refined, with a short list of questions returned to
participants for prioritization (Survey II). A further prioritization exercise was carried out using the
Hanlon Process of Prioritization at a PREDICT members meeting where members were asked to assess
prevalence of the condition (A), seriousness of the condition (B) and feasibility of the question to
research (C), a priority score was then calculated ((A+2B) X C) to establish the priorities of the active
researchers in PEM.
Results: 107/254 participants (42.5%) responded to Survey I, 144/245 (58.7%) to Survey II. 246
individual research questions were submitted. The areas generating most questions were respiratory
diseases 19.5%, infectious diseases 10.9%, trauma 7.3%, analgesia and sedation 6.9% and critical care
5.7%. Following Survey II a list of the top 35 research questions was established. Priority topics include
high flow oxygenation prior to intubation, fluid volume for resuscitation in sepsis, imaging in cervicalspine
injury, intravenous therapy for asthma and vasopressor use in sepsis. Using the Hanlon process
top ranked topics were intravenous agents in asthma, treatment of urinary tract infections, respiratory
support in asthma, cervical-spine decision rules and pre-intubation check lists.
Conclusions: This modified Delphi Method has established a clear list of research questions and
priorities which will inform multi-centre PEM research in Australia and New Zealand.

Background: The Paediatric Research in Emergency Departments International Collaborative (PREDICT) network was established in 2004 to improve emergency care for children by undertaking multicentre research in areas where there is a lack of evidence, incomplete translation of evidence to practice, and limited policy development/dissemination. The 20-site network has substantially improved the evidence-base for a wide range of conditions in paediatric emergency medicine. Despite this, evidence-based care is inconsistently provided. Translating research evidence into practice in the emergency department setting is known to present particular challenges. PREDICT through its NHMRC funded Centre for Research Excellence (CRE) is undertaking research to identify effective strategies to address these challenges and improve the health outcomes of children and families.

Objectives: To describe PREDICT’s research including studies to improve the uptake of research evidence in the acute care of children with mild head injuries (MHIs) and infants with bronchiolitis.

Method: The research will be described in relation to components of the Knowledge to Action Framework¹.

Results: The most appropriate clinical decision rule (CDR) to guide the management of paediatric MHIs in Australasia has been determined in a NHMRC funded study which assessed the performance accuracy of international head injury CDRs in 20,000+ children. An evidence-based Australasian acute care bronchiolitis guideline has also been developed. For these two projects we will explore potential barriers/enablers to evidence-based care and develop tailored interventions to improve uptake. The bronchiolitis intervention will be evaluated in a CRE-funded cluster randomized controlled trial followed by a study to measure sustainability of change.

Background: Computed Tomography (CT) is an excellent tool for understanding the nature and severity of head injury. In paediatric head injury judicious use is urged however as the risk of later malignancy has been reported between 1 in 1000-5000. Clinical Decision Rules (CDRs) are often used to help determine when a CT is likely to be of most use, thus reducing unnecessary exposure to radiation. The Townsville Emergency Department has largely endorsed the use of CHALICE to date.

Method: Townsville participated as one of 10 sites across Australia in this prospective, observational study. Sufficient information was collected to make clinical decisions according to three commonly used CDRs for CT use in paediatric head injuries – CHALICE, PECARN and CATCH. Also collected was outcome information through follow up phone calls to parents.

Results: In total 20,255 patients were included in the study with 1064 contributed from Townsville. Sensitivities (95% CI) were ranked as follows: PECARN <2 years 100.0% (91.4% to 100.0%), PECARN >2 years 99.1% (95.0% to 100.0%), CATCH 96.0% (79.6% to 99.9%) and CHALICE 93.8% (91.0% to 95.9%). PECARN <2 years did not miss any patients, PECARN >2 years missed one patient who did not require neurosurgery. CATCH missed one patient with a bleeding disorder who required neurosurgery. CHALICE missed 28 patients, two of whom required neurosurgery.

Conclusions: The sensitivity of the PECARN CDRs was higher than CATCH or CHALICE when the CDRs were used as designed as well as in an undifferentiated cohort of children with mild head injury.

Background: Paediatric endotracheal intubation is a rarely performed, high risk procedure in the emergency department (ED). Recent studies estimate that it is performed in approximately 1:1000 paediatric ED attendances. This rarity raises questions about ongoing airway procedural exposure for paediatric emergency clinicians.

Objectives: To determine the recent performance or supervision, and confidence for various paediatric airway procedures by senior paediatric emergency clinicians.

Method: Web-based survey of senior paediatric emergency clinicians regarding performance, supervision, and confidence relating to airway procedures in children aged 0-18 years. The survey was distributed through paediatric emergency research networks (PERN) in the UK, USA, Canada, Europe, South America, Australia and New Zealand.

Results: 1602 clinicians responded to the survey, with an overall response rate of 65%. 1522 (95%) respondents reported their most recent procedural experience. In the last 12 months, 1297 (85%) had performed bag-valve mask ventilation, 900 (59%) had performed intubation, 248 (17%) had placed a laryngeal mask airway, 18 (1%) had performed a surgical airway, and 348 (23%) had changed a tracheostomy.  1316 (87%) of clinicians had never supervised or performed a surgical airway, 600 (40%) had never supervised or performed a tracheostomy change, and 467 (31%) had never supervised or performed placement of a laryngeal mask airway. Confidence (at least 3 on a 5-point Likert scale) was low for surgical airways, and depended somewhat on age for other airway procedures.

Conclusions: Intubation and bag-valve-mask ventilation are the most common paediatric airway procedures. Supervision and experience in other emergency airways is rare.

Background: Non-airway paediatric critical procedures are rarely performed in the ED. Recent studies suggest that approximately one per thousand ED attendances may require some sort of critical procedure, with intubation being by far the most common. It is unknown how often critical chest procedures such as chest decompression, CPR, ED thoracotomy, defibrillation and pacing are performed by paediatric emergency clinicians.

Objectives: To determine the recent performance or supervision, and confidence for various paediatric critical chest procedures by senior paediatric emergency clinicians

Method: Web-based survey of senior paediatric emergency clinicians regarding performance, supervision, and confidence relating to critical chest procedures in children aged 0-18 years. The survey was distributed through paediatric emergency research networks (PERN) in the UK, USA, Canada, Europe, South America, Australia and New Zealand.

Results: 1602 clinicians responded to the survey, with an overall response rate of 65%. 1508 (94%) respondents reported their most recent chest procedural experience.

In the last 12 months, 962 (64%) had performed CPR, 190 (13%) had performed needle thoracostomy, 245 (16%) had performed tube thoracostomy, 380 (25%) had performed DC cardioversion or defibrillation, and 57 (4%) had performed transcutaneous pacing. 18 (1%) had performed pericardiocentesis, and 21 (1%) ED thoracotomy.

More than 70% of respondents had never supervised or performed pacing, pericardiocentesis or ED thoracotomy. ED thoracotomy and pericardiocentesis had the lowest frequency of respondents reporting confidence in performing the procedure.

Conclusions: Invasive critical chest procedures are very rarely performed in children. Clinical practice is unlikely to ensure adequate procedural exposure.

Background: It is unknown how often invasive vascular access procedures such as venous cutdown, and insertion of intraosseous, central venous and arterial lines are performed by paediatric emergency clinicians.

Objectives: To determine the recent performance or supervision, and confidence for various invasive paediatric vascular access procedures by senior paediatric emergency clinicians.

Method: Web-based survey of senior paediatric emergency clinicians regarding performance, supervision, and confidence relating to critical chest procedures in children aged 0-18 years. The survey was distributed through paediatric emergency research networks (PERN) in the UK, USA, Canada, Europe, South America, Australia and New Zealand.

Results: 1602 clinicians responded to the survey, with an overall response rate of 65%. 1499 (94%) respondents reported their most recent invasive vascular access experience.  In the last 12 months, 979 (64%) had personally inserted an intraosseous line, 283 (19%) a central venous line, and  265 (18%) an arterial line. 19 (1%) respondents had performed a venous cutdown. 1231 (82%) had never performed or supervised a venous cutdown, while 332 (22%) and 348 (23%) had never performed or supervised insertion of a central venous line or arterial line respectively. Procedural confidence for intraosseous lines was high, while confidence increased with increasing patient age for central venous access and arterial lines. Confidence was low for venous cutdown.

Conclusions: More than half of the paediatric emergency clinicians surveyed had inserted an intraosseous needle within the last 12 months. Other invasive vascular access performance was less common, and associated with less procedural confidence.

Background: Exacerbation of asthma in children is a common reason for presentation to emergency departments. Severe acute asthma is uncommon, however, is subject to significant variation in practice.

Objective: To collate and discuss the primary and secondary outcomes used in comparative intravenous treatment studies for acute asthma.

Method: We systematically searched MEDLINE, EMBASE, COCHRANE, PUBMED databases to identify published studies in children (aged 1-18 years) with acute severe asthma. Studies were included if they compared intravenous therapy to another treatment. Primary and secondary outcomes were determined and collated prior to analysis.

Results: 30 papers were identified. The most commonly used primary outcome was a clinical outcome score (13/30), with similar distributions between pulmonary index, clinical asthma score and asthma score not otherwise specified. Other identified primary outcomes included tests of respiratory function (7/30), length of hospital stay (7/30) and timing of salbutamol administration (3/30). Timing of data collection for primary outcomes varied between studies. For secondary outcomes, most studies comprised more than one measure with a total of 60 outcomes identified across the 30 papers. The most commonly identified secondary outcome was hospital length of stay (17/60).

Conclusions: For studies comparing intravenous treatment modalities for children with acute asthma, the selection of type, number and timing of outcome measures varies greatly.  In order to establish the effectiveness of individual therapies, there is a need to develop consensus on outcome measures  for acute severe paediatric asthma.

Background: Children with head injuries (HIs) frequently present to acute care settings. Most injuries are mild, but a small portion of patients have clinically significant intracranial injuries. An attempt to estimate the cost of paediatric HIs using prospectively collected data has never been attempted in Australia.

Objectives: To estimate the total cost of head injury in Australia and provide an understanding of the costs associated with different causative and presentation factors.

Method: The study was embedded in a prospective observational study of 20,255 children <18 years designed to validate three published clinical decision rules. Each patient’s hospital presentation was micro-costed and then extrapolated to an annualised Australia-wide figure using Australian Independent Hospital Pricing Authority data.

Results: There were 109,729 cases of head injuries in children aged 0-18 years across Australia in the fiscal year 2012-2013 (1.67% of total emergency presentations). The average cost of each head injury was $1348.21 [95%CI: 1173.02 to 1523.40]. The total cost of head injury in Australia was $147.94 million [95% CI: 128.71 to 167.17].

Regression and generalized linear models both demonstrate that motor vehicle accidents (MVAs) and non-accidental injury (NAI) are the most significant drivers of cost, as are injuries presenting with vomiting, seizure and loss of consciousness greater than 5 minutes.

Conclusion: Paediatric head injury represents a significant cost to the Australian healthcare system. Efforts to reduce the main drivers of cost – MVAs and NAI – may reduce this.

Background: Children with head injuries (HIs) frequently present to acute care settings. Most injuries are mild, but a small portion of patients have clinically significant intracranial injuries. An attempt to estimate the cost of paediatric HIs using prospectively collected data has never been attempted in Australia.

Objectives: To estimate the total cost of head injury in Australia and provide an understanding of the costs associated with different causative and presentation factors.

Method: The study was embedded in a prospective observational study of 20,255 children <18 years designed to validate three published clinical decision rules. Each patient’s hospital presentation was micro-costed and then extrapolated to an annualised Australia-wide figure using Australian Independent Hospital Pricing Authority data.

Results: There were 109,729 cases of head injuries in children aged 0-18 years across Australia in the fiscal year 2012-2013 (1.67% of total emergency presentations). The average cost of each head injury was $1348.21 [95%CI: 1173.02 to 1523.40]. The total cost of head injury in Australia was $147.94 million [95% CI: 128.71 to 167.17].

Regression and generalized linear models both demonstrate that motor vehicle accidents (MVAs) and non-accidental injury (NAI) are the most significant drivers of cost, as are injuries presenting with vomiting, seizure and loss of consciousness greater than 5 minutes.

Conclusion: Paediatric head injury represents a significant cost to the Australian healthcare system. Efforts to reduce the main drivers of cost – MVAs and NAI – may reduce this.

Background: Informed prospective consent into research is undertaken to preserve and respect patient’s autonomy. In time critical emergency research this may not always be possible and retrospective consent processes are recognised as appropriate by ethics committees. The use of retrospective consent processes in children has further ethical considerations due to children being considered a population at risk. Recently the use of retrospective consent has gained considerable media interest, until recently this process had not been utilised in Australasian emergency departments (EDs) in paediatric research, and acceptability to families is unknown.

Objectives: To report on the experience and acceptability of retrospective consent in two large paediatric multi-centre RCTs in Australasian EDs; ConSEPT (12 EDs), levetiracetam vs. phenytoin in status epilepticus and PARIS (17 EDs), high flow vs. nasal prong oxygen in bronchiolitis.

Method: Prospective data collection of randomisation and enrolment into ConSEPT and PARIS studies. All ConSEPT patients are consented retrospectively. All PARIS patients in Australia are consented retrospectively. PARIS patients in New Zealand (NZ) are consented retrospectively if they present with an oxygen requirement, but are consented prospectively if admitted without an oxygen requirement.

Results: ConSEPT: 88 patients (20 NZ) randomised and approached retrospectively for consent, 0 refused. PARIS: In Australia 927 randomised and approached retrospectively for consent, 37 (4%) refused. In NZ 337 randomised and approached retrospectively for consent, 15 (4%) refused; 154 approached prospectively for consent, 19 (12%) refused.

Conclusion: Use of retrospective consent in paediatric trials is feasible locally and appears to be acceptable to families.

Background: Bronchiolitis is the commonest lower respiratory tract infection in children less than 12 months and is the most frequent cause of hospitalisation in infants under 6 months of age in Australasia.  Research conducted by the Paediatric Research in Emergency Department International Collaborative (PREDICT) identified substantial variation in practice patterns in the management of bronchiolitis, despite high levels of evidence supporting some areas of treatment.

Objective: To improve consistency of Australasian practice, an evidence based consensus led guideline for the management of bronchiolitis in infants presenting to, and admitted into hospitals was developed.

Method: The nine step process recommended by the National Health and Medical Research Council (NHMRC) and the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology were adapted.  Formulation of the guideline included identification of population, intervention, comparator, outcomes and time (PICOt) questions, a systematic literature search, grading of evidence using the NHMRC process and GRADE methodology.  Consensus with the Development Committee was sought using nominal group technique principles to formulate the clinical recommendations and practice points.  The final step was consultation and review by key paediatric health professional bodies seeking endorsement to ensure the guideline was relevant to the management of bronchiolitis in the Australasian emergency and ward settings.

Results: Developing evidence-based clinical guidelines is a complex process.  Involvement of stakeholders, who would ultimately be end-users as members of the Development Committee, provided valuable clinical knowledge and ultimately lead to greater likelihood of implementation and utilisation.  However, this did present challenges with committee members located over 2 countries and 5 time zones, with variable baseline knowledge of evidence grading.

Conclusion: The GRADE and NHMRC processes provided a systematic and transparent approach, ensuring a final structure that includes a useable clinical interface for bedside use and a descriptive summary of the evidence base and evidence tables for each key statement.