Profound hypothermia and no ECMO?

July 11, 2014 by  
Filed under Acute Med, All Updates, ICU, Kids, Resus


Patients in cardiac arrest due to severe hypothermia benefit from extracorporeal rewarming, and it is often recommended that they are treated at centres capable of providing cardiopulmonary bypass or extracorporeal membrane oxygenation (ECMO).

But what if they’re brought to a centre that doesn’t have those facilities?

If you work in such a centre do you have a plan, and are you familiar with what equipment you could use?

One option if you have an ICU is to provide extracorporeal warming using a haemofiltration machine used for renal replacement therapy(1). A double lumen haemofiltration catheter is inserted into a central vein and an ICU nurse can often do the rest, although some variables have to be set by the intensivist, often aided by a standard renal replacement therapy prescription chart. The machines are mobile and can be wheeled into the resus room (I have practiced this set up in resus). It might be worth discussing and practicing this option with your ICU.

Another extracorporeal option is to rig up a rapid infusion device such as a ‘Level 1′ to connect to arterial and venous catheters so that blood from the patient flows through and is warmed by the machine before being returned to the patient(2). Rapid rewarming has been achieved by this method but it requires some modification to the usual set up and so is much less likely to be a realistic option for most teams doing this on very rare occasions.

Less technical options are the traditionally taught warm saline lavage of body cavities such as the thorax and the peritoneal cavity. These can be achieved with readily available catheters and of course should be combined with ventilation with warmed gas and administration of warm intravenous fluid.

Thoracic lavage can be achieved with open thoracotomy or tube thoracostomy. One or two chest tubes can be placed on each side. One technique was described as:

Two 36 French chest tubes were placed in each hemithorax. One tube was placed in the fourth intercostal space in the mid-clavicular line. Another tube was placed into the sixth intercostal space in the mid-axillary line. Sterile saline at 39.0◦C was infused by gravity into each superior chest tube and allowed to drain passively through each inferior tube.(3)

Rapid rewarming at a rate of 6.8◦C per hour was achieved in an arrested hypothermic man using peritoneal lavage. It was done in the operating room with peritoneal lavage (saline 40◦C) with a rapid infusion system (Level 1) through two laparoscopic access sites. It was combined with external forced air rewarming and warm intravenous infusions(4).

Finally some devices manufactured for inducing hypothermia in post-cardiac arrest patients can also be used to rewarm patients, which might be endovascular devices, such as the Cool Line® catheter(5), or external, such as the Arctic Sun® Temperature Management System(6). It’s definitely worth finding out what your critical care services have as far as this equipment goes.

In summary, although the ‘exam answer’ for cardiac arrest due to profound hypothermia is often ECMO/cardiopulmonary bypass, in most centres that’s not an option. It’s helpful to remind ourselves that (1) other extracorporeal rewarming options exist and (2) non-extracorporeal techniques can provide rapid rewarming.


1. Spooner K, Hassani A. Extracorporeal rewarming in a severely hypothermic patient using venovenous haemofiltration in the accident and emergency department. J Accid Emerg Med. 2000 Nov;17(6):422–4. Full text

2. Gentilello LM, Cobean RA, Offner PJ, Soderberg RW, Jurkovich GJ. Continuous arteriovenous rewarming: rapid reversal of hypothermia in critically ill patients. The Journal of Trauma: Injury, Infection, and Critical Care. 1992 Mar;32(3):316–25 PubMed

3. Plaisier BR. Thoracic lavage in accidental hypothermia with cardiac arrest — report of a case and review of the literature. Resuscitation. 2005 Jul;66(1):99–104. PubMed

4. Gruber E, Beikircher W, Pizzinini R, Marsoner H, Pörnbacher M, Brugger H, et al. Non-extracorporeal rewarming at a rate of 6.8°C per hour in a deeply hypothermic arrested patient. Resuscitation. 2014 Aug;85(8):e119–20. PubMed

5. Kiridume K, Hifumi T, Kawakita K, Okazaki T, Hamaya H, Shinohara N, et al. Clinical experience with an active intravascular rewarming technique for near-severe hypothermia associated with traumatic injury. Journal of Intensive Care. BioMed Central Ltd; 2014;2(1):11. link to abstract

6. Cocchi MN, Giberson B, Donnino MW. Rapid rewarming of hypothermic patient using arctic sun device. Journal of Intensive Care Medicine. 2012 Mar;27(2):128–30. PubMed

When to Stop Resuscitation

July 9, 2014 by  
Filed under Acute Med, All Updates, EMS, Guidelines, ICU, Kids, Resus, Trauma

Comments Off

My talk at the SmaccGOLD conference in March 2014

Cliff Reid – When Should Resuscitation Stop from Social Media and Critical Care on Vimeo.

Here are the slides:

Down with “down” time!

May 11, 2014 by  
Filed under Acute Med, All Updates, EMS, Resus

CPR-icon2A man in his 40s has a witnessed collapse and CPR is immediately started. Paramedics are on scene within 5 minutes and initiate advanced cardiac life support. He has refractory ventricular fibrillation which degenerates to asystole. He arrives in an emergency department where, with good ongoing CPR, he appears reasonably well perfused and even demonstrates some spontaneous movements and reactive pupils. He is placed on a mechanical CPR device and activation of the cardiac cath lab is requested. The patient has been in cardiac arrest now for 32 minutes. The cardiology fellow appears and asks: ‘what’s the down time?’

What’s the right answer? Would you say ‘half an hour’? ’32 minutes’?
And does it matter? Why is the cardiology fellow asking? Does she have an arbitrary cut off in mind, over which emergency coronary reperfusion will be denied?

I think there are several problems with conversations like these.
The first, is what does ‘down time’ even mean?
The second, is how relevant is a cardiac arrest time interval to prognosis in an individual patient?
The third, is what is the significance of any time interval in a patient who at the time of assessment has some signs that CPR is providing some perfusion and there is some evidence of brain function?

Let’s take the first. The definition of ‘down time’ does not appear to be standardised:

In this publication it appears to refer to the time before resuscitation is commenced, where it is demonstrated to be prognostically important.

Similarly, in this medical dictionary, it is defined as the ‘temporal duration from cardiac arrest until beginning cardiopulmonary resuscitation or advanced cardiac life support.

However, a post in Life in the Fast Lane defines it as ‘time to return of spontaneous circulation

This appears to agree with The New South Wales Government’s Intensive Care Monitoring and Coordination Unit who define it as ‘the time from when a person’s heart stops beating to the time it starts beating again

Yet another definition is used in King County, Washington, where it is defined as ‘the time interval from collapse to call 911‘.

So the first thing is to clarify what we’re talking about: “This patient received immediate bystander CPR. He has had resuscitation for 32 minutes”. My friend in the UK, nurse resuscitationist Fernando Candal Carballido, coined the term ‘Time of Supported Circulation‘, or TOSC. I quite like this and think it could catch on.

The next question is so what? What if it was 90 minutes? At what point do we declare futility? This is where I believe the game has changed. Multiple survivors of prolonged resuscitation are springing up in the news and in the literature. Particularly in the subgroup of patients with minimal comorbidity, early CPR, and who receive circulatory support via ECMO or mechanical CPR while they undergo coronary reperfusion.

For a great example of a prolonged CPR survivor, check out paramedic Wayne Schneider’s story,

…or listen to Steven Bernard describe amazing results from ECMO used in Melbourne in the CHEER study, which includes survivors of over two hours of CPR.

So, in summary:

  • Be clear on your definitions when communicating with colleagues. ‘Down time’ does not appear to have a standard definition, so I would avoid its use.
  • Some patients without comorbidities who have had early bystander CPR may survive despite long periods of CPR (or ‘TOSC’), provided the underlying cause can be treated or is reversible.
  • ECMO and even more widely available mechanical CPR devices are extending the period in which these causes can be addressed.

Emergency Medicine – A Great Job

May 10, 2014 by  
Filed under All Updates

Comments Off

I was asked to speak at the Australasian Conference for Emergency Medicine‘s Annual Scientific Conference in Adelaide in November 2013. The title they gave me was ‘What a great job’. It was a great opportunity for me to explore some of the literature around what makes people happy, and whether emergency medicine has the ingredients to do that. It does. But not if you do too much.

The College has generously made available many of the conference talks as FOAM here.

Here’s my talk. The slideset is below.

2013 ASM: Dr Cliff Reid – What a great job from ACEM Digital Media on Vimeo.

Time to change thinking on ‘cricoid pressure’

April 22, 2014 by  
Filed under All Updates

I don’t like cricoid pressure. Some people do. There is insufficient evidence that it is of any benefit. There is some consistent evidence that it worsens laryngoscopic view.

In my clinical practice of critical care in and out of hospital, I can’t afford to risk delaying the securing of my patients’ airways with a procedure in which in my view the risks of harm outweigh any unproven chance of benefit.

I had erroneously thought after many online ‘debates’ that the critical care community had settled on a compromise – if you want to use it great, just take it off if it’s causing a problem. If you don’t want to use it, then that’s equally fine.

However a Google Plus conversation last week ignited a storm! There was a suggestion that cricoid pressure represented a ‘standard of care’, and that not to use it in a critical care intubation would potentially invite legal proceedings, catalysed by colleagues prepared to testify against those of us who have carefully weighed the balance of evidence and selected what we feel is the best approach for our patients.

I wrote a post to challenge the very thinking that what might be considered a ‘standard of care’ in elective anaesthesia in some guidelines should ever be applied to a critical care airway. I proposed a tongue in cheek change of terminology to emphasis what we know about cricoid pressure in the critically ill: that it can delay intubation, distort and compress the airway, and move rather than compress the oesophagus (although I concede the latter point may be irrelevant in terms of CP’s proposed mechanism).

Some people got upset. I reworded the post and added a big fat disclaimer to avoid any perception of ad hominen attack. I wanted to attack and ridicule the procedure, not its proponents. I still got attacked using some bizarrely offensive comparisons by people you would expect to know better. It got ugly.

The combination of support by some people I hold in very high regard and a currently crazy schedule (I’ve been in the UK for three hours having just travelled from Australia) meant the post stayed up for a while until I could consider the feedback. I still haven’t read it all. But I’ve read enough.

I respect the people I disagree with. I respect absolutely their right to hold different views from my own. But I don’t respect all their views, and I don’t necessarily think people have a right not to be offended by my views. However if the WAY I EXPRESS those views causes UNNECESSARY offence I have to reconsider my message.

The science around cricoid pressure is there in the literature. The arguments that it can acceptably be discarded in critical care are powerful. If we need a new acronym it doesn’t need to be one that can be pronounced and construed in a way different to that which I’d envisaged. As Dr Brent May so insightfully put: ‘You can’t emphasise a syllable on Twitter‘.

I want to thank EVERYONE who provided constructive feedback on and off social media. I apologise unreservedly to anyone offended by the post. It’s gone. The battle against unthinking enforcement of a potentially harmful technique goes on, but the unwitting offence of innocent parties is not an acceptable consequence. I will try to be more intelligent in subsequent debate.


Breaking with tradition in paediatric RSI

April 8, 2014 by  
Filed under All Updates, EMS, ICU, Kids, Resus

‘Traditional’ rapid sequence induction of anaesthesia is often described with inclusion of cricoid pressure and the strict omission of any artifical ventilation between paralytic drug administration and insertion of the tracheal tube. These measures are aimed at preventing pulmonary aspiration of gastric contents although there is no convincing evidence base to support that. However it is known that cricoid pressure can worsen laryngoscopic view and can occlude the paediatric airway. We also know that children desaturate quickly after the onset of apnoea, and although apnoeic diffusion oxygenation via nasal cannula can prevent or delay that, in some cases it may be preferable to bag-mask ventilate the patient while awaiting full muscle relaxation for laryngoscopy.

A Swiss study looked at 1001 children undergoing RSI for non-cardiac surgery. They used a ‘controlled rapid sequence induction and intubation (cRSII)’ approach for children assumed to have full stomachs. This procedure resembled RSI the way it is currently done in many modern critical care settings, including the retrieval service I work for:

  • No cricoid pressure
  • Ketamine for induction if haemodynamically unstable
  • A non-depolarising neuromuscular blocker rather than succinylcholine
  • No cricoid pressure
  • Gentle facemask ventilation to maintain oxygenation until intubation conditions achieved
  • Intubation with a cuffed tracheal tube
  • Still no cricoid pressure

The authors comment:

The main finding was that cRSII demonstrated a considerably lower incidence of oxygen desaturation and consecutive hemodynamic adverse events during anesthesia induction than shown by a previous study on classic RSII in children. Furthermore, there was no incidence of pulmonary aspiration during induction, laryngoscopy, and further course of anesthesia.

Looks like more dogma has been lysed, and this study supports the current trajectory away from traditional teaching towards an approach more suitable for critically ill patients.

Controlled rapid sequence induction and intubation – an analysis of 1001 children
Paediatr Anaesth. 2013 Aug;23(8):734-40

BACKGROUND: Classic rapid sequence induction puts pediatric patients at risk of cardiorespiratory deterioration and traumatic intubation due to their reduced apnea tolerance and related shortened intubation time. A ‘controlled’ rapid sequence induction and intubation technique (cRSII) with gentle facemask ventilation prior to intubation may be a safer and more appropriate approach in pediatric patients. The aim of this study was to analyze the benefits and complications of cRSII in a large cohort.

METHODS: Retrospective cohort analysis of all patients undergoing cRSII according to a standardized institutional protocol between 2007 and 2011 in a tertiary pediatric hospital. By means of an electronic patient data management system, vital sign data were reviewed for cardiorespiratory parameters, intubation conditions, general adverse respiratory events, and general anesthesia parameters.

RESULTS: A total of 1001 patients with cRSII were analyzed. Moderate hypoxemia (SpO2 80-89%) during cRSII occurred in 0.5% (n = 5) and severe hypoxemia (SpO2 <80%) in 0.3% of patients (n = 3). None of these patients developed bradycardia or hypotension. Overall, one single gastric regurgitation was observed (0.1%), but no pulmonary aspiration could be detected. Intubation was documented as ‘difficult’ in two patients with expected (0.2%) and in three patients with unexpected difficult intubation (0.3%). The further course of anesthesia as well as respiratory conditions after extubation did not reveal evidence of ‘silent aspiration’ during cRSII.

CONCLUSION: Controlled RSII with gentle facemask ventilation prior to intubation supports stable cardiorespiratory conditions for securing the airway in children with an expected or suspected full stomach. Pulmonary aspiration does not seem to be significantly increased.

Palpating neonatal tracheal tubes

April 6, 2014 by  
Filed under All Updates, EMS, ICU, Kids, Resus

Comments Off

infant-intubate-iconAfter neonatal intubation, the incidence of malposition of the tip of the tracheal tube is fairly high.

A technique was evaluated involving palpation of the tube tip in the suprasternal notch, which in this small study was superior to insertion length based on a weight-based nomogram.

The suprasternal notch was chosen because it anatomically corresponds to vertebral level T2, close to the optimal position at the mid-tracheal point. Correct position on the chest radiograph was defined as any position <0.5 cm above the interclavicular midpoint and more than 1 cm above the carina.

During tracheal tube placement, the tip was gently palpated in the suprasternal notch with the index or little finger of the left hand while holding the body of the tube with the fingers of the right hand. The tube tip was adjusted until the bevelled edge was just palpable in the the suprasternal notch.

Digital palpation of endotracheal tube tip as a method of confirming endotracheal tube position in neonates: an open-label, three-armed randomized controlled trial.
Paediatr Anaesth. 2013 Oct;23(10):934-9

OBJECTIVE: To compare the malposition rates of endotracheal tubes (ETTs) when the insertional length (IL) is determined by a weight-based nomogram versus when IL is determined by palpation of the ETT tip.

DESIGN: Open-label, randomized controlled trial (RCT).

SETTING: Level III neonatal intensive care unit (NICU).

SUBJECTS: All newborn babies admitted in NICU requiring intubation.

INTERVENTIONS: Subjects were randomly allocated to one of three groups, wherein IL was determined by (i) weight-based nomogram alone, (ii) weight-based nomogram combined with suprasternal palpation of ETT tip performed by specially trained neonatology fellows, or (iii) combination of weight-based and suprasternal methods by personnel not specially trained.

PRIMARY OUTCOME: Rate of malposition of ETT as judged on chest X-ray (CXR).

RESULTS: Fifty seven babies were randomized into group 1(n = 15), group 2 (n = 20), and group 3 (n = 22). The proportion of correct ETT placement was highest in group 2, being 66.7%, 83.3%, and 66.7% in groups 1 through 3, respectively (P value = 0.58). No complication was attributable to palpation technique.

CONCLUSION: Suprasternal palpation shows promise as a simple, safe, and teachable method of confirming ETT position in neonates.

Atropine for Paediatric RSI?

April 5, 2014 by  
Filed under All Updates, EMS, ICU, Kids, Resus

Comments Off

paedRSIdrugiconIn some areas it has been traditional to pre-medicate or co-medicate with atropine when intubating infants and children, despite a lack of any evidence showing benefit. It is apparently still in the American Pediatric Advanced Life Support (PALS) Provider Manual when age is less than 1 year or age is 1–5 years and receiving succinylcholine. However it is not recommended with rapid sequence intubation in the British and Australasian Advanced Paediatric Life Support manual and course.

A French non-randomised observational study compares intubations with and without atropine in the neonatal and paediatric critical care setting. Atropine use was associated with significant acceleration of heart rate, and no atropine use was associated with a higher incidence of new dysrhythmia, the most common being junctional rhythm, but with none appearing to be clinically significant.

The incidence of the most important peri-intubation cause of bradycardia – hypoxia – is not reported. It is also not clear how many intubation attempts were required. The authors admit:

it is not possible using our methodology to deduce whether bradycardia was due to hypoxia, laryngoscopy, or sedation drugs.

Actual rapid sequence was rarely employed – their use of muscle relaxants was low – making this difficult to extrapolate to modern emergency medicine / critical care practice.

My take home message here is that this study provides no argument whatsoever for the addition of atropine in routine RSI in the critically ill child. Why complicate a procedure with an unnecessary tachycardia-causing drug when the focus should be on no desat / no hypotension / first look laryngoscopy?

The Effect of Atropine on Rhythm and Conduction Disturbances During 322 Critical Care Intubations
Pediatr Crit Care Med. 2013 Jul;14(6):e289-97

OBJECTIVES: Our objectives were to describe the prevalence of arrhythmia and conduction abnormalities before critical care intubation and to test the hypothesis that atropine had no effect on their prevalence during intubation.

DESIGN: Prospective, observational study.

SETTING: PICU and pediatric/neonatal intensive care transport.

SUBJECTS: All children of age less than 8 years intubated September 2007-2009. Subgroups of intubations with and without atropine were analyzed.


MEASUREMENT AND MAIN RESULTS: A total of 414 intubations were performed in the study period of which 327 were available for analysis (79%). Five children (1.5%) had arrhythmias prior to intubation and were excluded from the atropine analysis. Atropine was used in 47% (152/322) of intubations and resulted in significant acceleration of heart rate without provoking ventricular arrhythmias. New arrhythmias during intubation were related to bradycardia and were less common with atropine use (odds ratio, 0.14 [95% CI, 0.06-0.35], p < 0.001). The most common new arrhythmia was junctional rhythm. Acute bundle branch block was observed during three intubations; one Mobitz type 2 rhythm and five ventricular escape rhythms occurred in the no-atropine group (n = 170). Only one ventricular escape rhythm occurred in the atropine group (n = 152) in a child with an abnormal heart. One child died during intubation who had not received atropine.

CONCLUSIONS: Atropine significantly reduced the prevalence of new arrhythmias during intubation particularly for children over 1 month of age, did not convert sinus tachycardia to ventricular tachycardia or fibrillation, and may contribute to the safety of intubation.

Resus Team Size and Productivity

April 3, 2014 by  
Filed under All Updates, Kids, Resus, Trauma

paedsimiconA paediatric trauma centre study showed that in their system, seven people at the bedside was the optimum number to get tasks done in a paediatric resuscitation. As numbers increased beyond this, there were ‘diminishing marginal returns’, ie. the output (tasks completed) generated from an additional unit of input (extra people) decreases as the quantity of the input rises.

The authors comment that after a saturation point is reached, “additional team members contribute negative returns, resulting in fewer tasks completed by teams with the most members. This pattern has been demonstrated in other medical groups, with larger surgical teams having prolonged operative times and larger paramedic crews delaying the performance of cardiopulmonary resuscitation.

There are several possible explanations:

  • crowding limits access to the patient or equipment;
  • “social loafing”- staff may feel less accountable for the overall group performance and less pressure to accomplish individual tasks;
  • seven is the number recommended in that institution’s trauma activation protocol, with optimal role allocation described for a team of that size;
  • teams with redundant personnel may experience role confusion and fragmentation, resulting in both repetition and omission of tasks.

In my view, excessive team size results in there being more individuals to supervise & monitor, and hence a greater cognitive load for the team leader (cue the resus safety officer). More crowding and obstruction threatens situational awareness. There is more difficulty in providing clear uninterrupted closed loop communication. And general resuscitation room entropy increases, requiring more energy to contain or reverse it.

However, for paediatric resuscitations requiring optimal concurrent activity to progress the resuscitation, I do struggle with less than five. Unless of course I’m in my HEMS role, when the paramedic and I just crack on.

More on Making Things Happen in resus.

Own The Resus talk

Resus Room Management site

Factors Affecting Team Size and Task Performance in Pediatric Trauma Resuscitation.
Pediatr Emerg Care. 2014 Mar 19. [Epub ahead of print]

OBJECTIVES: Varying team size based on anticipated injury acuity is a common method for limiting personnel during trauma resuscitation. While missing personnel may delay treatment, large teams may worsen care through role confusion and interference. This study investigates factors associated with varying team size and task completion during trauma resuscitation.

METHODS: Video-recorded resuscitations of pediatric trauma patients (n = 201) were reviewed for team size (bedside and total) and completion of 24 resuscitation tasks. Additional patient characteristics were abstracted from our trauma registry. Linear regression was used to assess which characteristics were associated with varying team size and task completion. Task completion was then analyzed in relation to team size using best-fit curves.

RESULTS: The average bedside team ranged from 2.7 to 10.0 members (mean, 6.5 [SD, 1.7]), with 4.3 to 17.7 (mean, 11.0 [SD, 2.8]) people total. More people were present during high-acuity activations (+4.9, P < 0.001) and for patients with a penetrating injury (+2.3, P = 0.002). Fewer people were present during activations without prearrival notification (-4.77, P < 0.001) and at night (-1.25, P = 0.002). Task completion in the first 2 minutes ranged from 4 to 19 (mean, 11.7 [SD, 3.8]). The maximum number of tasks was performed at our hospital by teams with 7 people at the bedside (13 total).

CONCLUSIONS: Resuscitation task completion varies by team size, with a nonlinear association between number of team members and completed tasks. Management of team size during high-acuity activations, those without prior notification, and those in which the patient has a penetrating injury may help optimize performance.

High Flow Nasal Cannulae In Paediatric Retrieval

April 2, 2014 by  
Filed under All Updates, EMS, ICU, Kids

Comments Off

High Flow Nasal Cannulae (HFNC) oxygen therapy was introduced in paediatric interfacility retrievals undertaken by the Mater Children’s PICU Retrieval Team in Queensland, Australia. In 793 under 2 year olds, HFNC was associated with a reduction in infants receiving invasive or non-invasive ventilation. 77% of the patients had bronchiolitis.

The rationale for this treatment is explained as:

Owing to the inherent properties of the infant respiratory system with small airways and high chest compliance, the risk of developing atelectasis is high in bronchiolitis. HFNC therapy applied early in the disease process may prevent progression of the disease and maintain normal lung volumes, thereby preventing atelectasis. As a result, the functional residual capacity can be maintained and work of breathing reduced, which may stabilize the patient sufficiently to avoid the need for intubation. For this purpose we used flow rates of 2 L/kg/min which have been shown to result in a positive end-expiratory pressure of 4–5 cmH2O

Read more on high-flow nasal cannula oxygen therapy.

High-flow nasal cannula (HFNC) support in interhospital transport of critically ill children
Intensive Care Med. 2014 Feb 15. [Epub ahead of print]

BACKGROUND: Optimal respiratory support for interhospital transport of critically ill children is challenging and has been scarcely investigated. High-flow nasal cannula (HFNC) therapy has emerged as a promising support mode in the paediatric intensive care unit (PICU), but no data are available on HFNC used during interhospital transport. We aimed to assess the safety of HFNC during retrievals of critically ill children and its impact on the need for invasive ventilation (IV).

METHODS: This was a retrospective, single-centre study of children under 2 years old transported by a specialized paediatric retrieval team to PICU. We compared IV rates before (2005-2008) and after introduction of HFNC therapy (2009-2012).

RESULTS: A total of 793 infants were transported. The mean transport duration was 1.4 h (range 0.25-8), with a mean distance of 205 km (2-2,856). Before introduction of HFNC, 7 % (n = 23) were retrieved on non-invasive ventilation (NIV) and 49 % (n = 163) on IV. After introduction of HFNC, 33 % (n = 150) were retrieved on HFNC, 2 % (n = 10) on NIV, whereas IV decreased to 35 % (n = 162, p < 0.001). No patients retrieved on HFNC required intubation during retrieval, or developed pneumothorax or cardiac arrest. Using HFNC was associated with a significant reduction in IV initiated by the retrieval team (multivariate OR 0.51; 95 % CI 0.27-0.95; p = 0.032).

CONCLUSIONS: We report on a major change of practice in transport of critically ill children in our retrieval system. HFNC therapy was increasingly used and was not inferior to low-flow oxygen or NIV. Randomized trials are needed to assess whether HFNC can reduce the need for IV in interhospital transport of critically ill children.

Next Page »