Grace Leo. The 8th Bubble Wrap, Don't Forget the Bubbles, 2017. Available at:
With millions upon millions of journal articles being published every year it is impossible to keep up. Every month we ask some of our friends from the world of paediatrics to point out something that has caught their eye.
Article 1: Can technology help us recognize sepsis sooner?
Balamuth, F., Alpern, E.R., Abaddessa, M.K., Hayes, K., Schast, A., Lavelle, J., Fitzgerald, J.C., Weiss, S.L. and Zorc, J.J., 2017. Improving recognition of pediatric severe sepsis in the emergency department: contributions of a vital sign–based electronic alert and bedside clinician identification. Annals of Emergency Medicine.
What’s it about?
Mortality from sepsis is a worldwide health concern and in the last 5 years there have been huge efforts to raise awareness and improve survival from this devastating disease. In paediatric practice, the major challenge in the developed world has been improving treatment without increasing negative balancing measures. The chief of these is overtreatment resulting in unnecessary hospital stay and contributing to antimicrobial resistance.
The search for the valid and specific model of sepsis recognition continues at pace and this study explore the combination of an electronic alert with clinician assessment.
It was essentially a quality improvement initiative with a before and after design. The intervention was an electronic alert (defined as elevated pulse rate or hypotension, concern for infection, and at least one of the following: abnormal capillary refill, abnormal mental status, or high-risk condition). The alert would then prompt a clinical review to determine the delivery of a sepsis protocol.
The outcome measure was defined as activation of this protocol or development of severe sepsis requiring ICU admission (within 24 hours)
Why does it matter?
The study found in over 95000 Emergency Department visits after the intervention (approx 85000 included before implementation) that 326 (0.3%) patients were treated as severe sepsis within 24 hours. There were 1112 (1.2%) positive electronic alerts of which 847 were not treated by the sepsis protocol following clinical review. Of these 847, within 24 hours 16 (1.8%) of them ended up being treated for sepsis. Of those not identified by electronic alert 43 patients were identified in the ED and subsequently treated. 2 patients who were neither electronically identified, nor clinician identified, subsequently ended up being treated for severe sepsis.
Mortality and disposition did not change pre- and post- the intervention and other potential outcomes i.e. length of stay, blood/culture results were not reported.
The authors stated the electronic alert alone detected severe sepsis with a sensitivity of 86.2% (i.e. 13.8% of patients were falsely negative) and a specificity of 99.1% (i.e. 0.9% were falsely positive). These figures improved with the addition of the clinical review.
The challenge in interpreting the results is that the outcome measure was one of two very different things. PICU admission is a very reasonable proxy, even if the underlying illness had been a viral infection it would seem reasonable to conclude that the child was unwell enough to require intensive support. However the use of the sepsis bundle as a marker of outcome in itself leads to the curiously high specificity results i.e. the intervention was looking to improve the use of the intervention.
Clinical Bottom Line:
Electronic alerts may help prioritise at risk patients to senior decision makers but other factors/cultures/interventions are likely to be needed to impact on patient centered outcome measures.
Reviewed by: Damian Roland
Article 2: What dose of paracetamol for preterm neonates?
Flint RB, Roofthooft DW, van Rongen A, van Lingen RA, van den Anker JN, van Dijk M, Allegaert K, Tibboel D, Knibbe CAJ, Simons SHP. Exposure to acetaminophen and all its metabolites upon 10, 15 and 20 mg/kg intravenous acetaminophen in very preterm infants. Pediatr Res. 2017 May 29.
What’s it about?
Paracetamol/acetaminophen dosing and pharmacokinetics in neonates might seem like a pretty hard/dry subject for something in Bubble Wrap. This Dutch-based trial investigated paracetamol pharmacokinetics in 59 premature neonates from 24-32/40, at doses of 10, 15 and 20mg/kg.
Levels of acetaminophen, acetaminophen–glucuronide, acetaminophen–sulfate, acetaminophen–glutathione, acetaminophen–cysteine, and acetaminophen–mercapturate were measured and AUCs plotted. The key clinical/safety question being to establish if there was saturation of either the nontoxic sulfation or glucuronidation pathways, which would lead to overflow into the potentially hepatotoxic CYP2E1 pathway (with the formation of NAPQI).
In short, none of the doses lead to saturation, on single infusions of paracetamol in the ranges above.
Why does it matter?
I’m always relatively sceptical about our knowledge of pharmacotherapeutics in paediatrics, and in no group more than in the premature neonates. The specific toxicity for paracetamol in neonates relates their low glucuronidation capacity and potential hepatotoxicity of the oxidative CYP2E1 pathway.
This study goes some way to answering part of the “effective dose vs safe dose” equation for paracetamol, a particularly pertinent question as our practice aims to more effectively identify and treat pain in neonates. Having used several differing protocols for dosing neonates with paracetamol, this study provides a range that I’m happier to work within, although the authors note that they did not investigate the effect of repeated dosing on the levels of paracetamol nor its metabolites.
The Bottom Line
Paracetamol is widely used for analgesia and increasingly so in the neonatal population. The authors are cautious in their recommendations, urging that their study be read as part of a broader literature base to establish a safe dosage range for paracetamol. Interestingly, in the discussion, for repeated dosing they suggest an initial load of 12mg/kg with subsequent dosing 6mg/kg q6h (see paper for references).
Reviewed by: Henry Goldstein
Article 3: Is nebulised hypertonic saline useful in managing acute bronchiolitis?
Angoulvant F et al. Effect of nebulized hypertonic saline treatment in emergency departments on the hospitalization rate for acute bronchiolitis: a randomised clinical trial. JAMA Pediatrics, 2017, doi: 10.1001/jamapediatrics.2017.1333
What’s it about?
Particularly in the winter months, bronchiolitis presentations and admissions pose a significant burden to our hospitals. Studies frequently examine ways to reduce the need for hospitalisation. There was a period, a few years ago, where hypertonic saline was thought to be effective in the management of acute bronchiolitis, however further studies questioned this evidence. This study is the largest single trial of nebulised hypertonic saline use in bronchiolitis.
The trial examined 777 infants, who were otherwise healthy, and who presented to the Emergency Department with acute bronchiolitis. Infants were randomised to receive either 4ml of nebulised hypertonic saline (3%) or 4ml of nebulised normal saline (0.9%). Treatments were given twice and over 20 minutes each time. The primary outcome was hospitalisation within 24 hours of presentation to the ED.
After 24 hours, 48.1% of the hypertonic saline group were admitted to hospital, compared to 52.2% of the normal saline group. This was not statistically significant.
Why does it matter?
There has been uncertainty about the utility of nebulised agents in bronchiolitis, and the choice of agent. This paper demonstrates clearly that hypertonic saline has no benefit over normal saline. Additionally it did show that patients in the hypertonic saline group demonstrated higher rates of mild adverse effects, such as worsening of their cough.
The Bottom Line
Nebulised hypertonic saline was no better than nebulised normal saline in preventing hospitalisation of infants with bronchiolitis presenting to the Emergency Department. In fact, it was associated with minor adverse effects. If you want to give an infant with acute bronchiolitis something nebulised, then make it normal saline.
Reviewed by: Tessa Davis
Article 4: Another tool to help us assess paediatric mental health
Cappelli M, Zemek R, Polihronis C, Thibedeau NR, Kennedy A, Gray C, Jabbour M, Reid S, Cloutier P. The HEADS-ED: Evaluating the Clinical Use of a Brief, Action-Oriented, Pediatric Mental Health Screening Tool. Pediatric Emergency Care. 2017 May 23.
What is it about?
Mental health disorders in children and young people are occuring in record numbers with seemingly more complex manifestations. Long waits for both emergency and mental health reviews in over-crowded waiting rooms are not ideal. Nor are some blanket policies to admit all cases overnight for a period of reflection, often to wards with young children – an environment not tailored to their needs. This study aimed to standardise the assessment of children and young people with mental health disorders presenting to ED.
Why does it matter?
This research team developed a clinical tool, HEADS-ED. It is a validated tool based on 7 variables (Home, Education, Activities, Drugs and Alcohol, Suicidality, Emotions and Discharge resources) with each scoring between 0-2. A previous study noted a sensitivity of 82% and specificity of 87% for predicting the need of admission or consult (when completed by crisis intervention workers).
This was a single centre effectiveness study that examines the utility and decision validity of the tool.
It occurred over an 11-month period in a tertiary paediatric hospital. ED staff review 2/3 of patients with the remainder referred directly to psychiatry (‘medically cleared’ at triage). Of 2,704 potential participants, 1,586 met the inclusion criteria with 665 (49%) having the HEADS-ED assessment document completed.
Of this cohort, 40% received a psych consult with 15% admitted. The overall mean score for the HEADS-ED was 5.6 (IQ 3-7). A higher mean score was associated with those referred for a consult (6.91) and requiring admission (7.2). There was a statistically significant difference between those admitted (7.2) and discharged (5.28) (p=.0000), and those referred for consult (6.91) and those not (4.7) (p=0.000). The results identified that a high score for any variable (i.e. 2) was significant in predicting the need for a psych consult or admission (except the alcohol or drugs).
In a previous paper, the authors had recommended a score of 8 (with 2 for suicide attempt) as the threshold for referral. In this population, 86% of participants who fulfilled this threshold were referred. 32.5% below this threshold were also referred for consult.
The study is single centre and there was poor percentage completion rate for the tool (<50%). The fact nearly a 1/3 of patients received input despite a score below the stated threshold may mean there are features of concern not picked up by the tool. Given the potential consequences of sending a patient home inappropriately are life altering is a prospective validation of this tool possible?
However the authors noted the referral rate of 40% is markedly less than the 90% they have seen in local units without HEADS-ED. A ROC curve analysis would be interesting, as would long term follow up of cases (i.e. re-attendance rates, need for psych input at later date etc)
The bottom line
It would be a bold decision to implement the tool as a referral guide but increasing the awareness of the HEADS approach is important. The ability of the tool to aid referral from triage, something we have tried locally without success, does seem to be an impressive attribute.
Reviewed by: Stephen Mullen (@smullen001)
That’s it for this month. Many thanks to all of our reviewers who have taken the time to scour the literature so you don’t have to. If you think they have missed something amazing then let us know.