Author: Charlotte Kennedy / Editor: Govind Oliver / Codes: SLO10 / Published: 28/08/2018

Warning

The content you’re about to read or listen to is at least two years old, which means evidence and guidelines may have changed since it was originally published. This content item won’t be edited but there will be a newer version published if warranted. Check the new publications and curriculum map for updates

Academic Emergency Medicine (EM) is a small but rapidly growing field. There are now numerous large randomised control trials being run by Emergency Departments across the globe, we are trailblazers in areas such as staff wellbeing research and many senior EM academics are now house-hold names (well, amongst other EM folk anyway!). But what about the next generation of up-and-coming researchers? What are the topics that interest them? And what can we expect from them in the near future?

The Rod Little prize is awarded every year by the Royal College of Emergency Medicine to a United Kingdom (UK) based trainee for their research as presented at the Annual Scientific Conference. This year the conference was held in conjunction with EUSEM and here we present short summaries of the top-scoring abstracts nominated for the prize.

“The lack of correlation between the Injury Severity Score and the need for life-saving interventions in trauma patients in the United Kingdomby Jamie Vassallo et al. (ST3 Emergency Medicine, University Hospitals Plymouth and Post-Doctoral Research Fellow, Academic Department of Military Emergency Medicine)

Most people in EM will be familiar with the frustration of trying to ensure that injured patients end up in the right hospital: a hospital which can address their needs but that also minimises transport times and keeps them as close as possible to their loved ones. In the UK trauma care was re-organised into regional trauma networks in 2012. Under this system, seriously injured patients (i.e. those defined as ‘major trauma’) are transported directly to a Major Trauma Centre, which has all the resources and specialities required to address their complex needs, whilst the remainder go to the nearest Trauma Unit. The identification of a ‘major trauma’ patient is based on their predicted Injury Severity Score (ISS). However, the ISS is calculated retrospectively using the extent and seriousness of a patient’s injuries. What we really want to know when triaging injured patients at scene is what treatment they will need and whether the receiving hospital is capable of delivering it. This study looks at whether the eventual ISS correlates with resource use and whether these two things are linked to mortality.

The study used data collected from 9 years’ worth of records kept on the Trauma Audit Research Network (TARN) database (2006 – 2009). In total it included 127,233 patients and of these 24,791 (19.5%) required a lifesaving intervention. Examples of lifesaving procedures include intubation and ventilation, transfusion or emergency surgery. The list of all 32 procedures was decided on by expert consensus prior to the study. The authors found that only 49.3% of patients that received a life-saving intervention had an eventual ISS of >15 as per the traditional definition of major trauma. In addition to this, statistical analysis showed only a weak correlation between the ISS and the need for a lifesaving intervention. Of those that died, 61% had an ISS > 15 but this didn’t necessarily mean that an intervention was performed: only 39.5% of these patients had a lifesaving procedure conducted. The authors conclude that the way major trauma is currently defined needs reviewing and have proposed that an alternative metric may need to be sought for trauma triage.

Jamie’s take home message

 “When we see a trauma patient in our ED, the assimilation process that goes through our mind is ‘what do I need to do for this patient’, not ‘what is their ISS’.  This study, in keeping with work done 3 decades ago shows that the two bear little correlation to one another.  Whilst the ISS allows for a nice ‘epidemiological’ definition, it may not be the best way to characterise these patients, nor the best way to validate the decision rules (triage tools) used to prioritise these patients to Major Trauma Centres.  Further work looking at this is starting later this year by Dr Gordon Fuller on the NIHR funded Major Trauma Triage Tool Study (MATTS).  Exciting times – watch this space!”

“Can emergency physician gestalt rule in or out acute coronary syndrome: validation in a large prospective diagnostic cohort study” by Govind Oliver et al. (ST3 Emergency Medicine, NIHR Academic Clinical Fellow, Emergency and Intensive Care Research Group (EMERGING), Manchester Royal Infirmary)

Chest pain is an extremely common ED presentation but less than 10% of patients with suspected cardiac chest pain actually have an acute coronary syndrome. There are a number of different rule out strategies in use across the UK, most of which include the use of serial troponins in combination with the ECG and clinical findings. But how good are we as a profession at predicting who has had a heart attack? If we think the cause of the chest pain is non-cardiac, are we usually right? This study looked at clinician gestalt in patients with suspected cardiac chest pain and compared it to diagnosed acute myocardial infarction (AMI).

The data came from the Bedside Evaluation of Sensitive Troponin (BEST) study, which was a large multicentre cohort study conducted across 18 different centres in the UK. Clinician gestalt was measured on a 5 point Likert scale which ranged from ‘definitely’ to ‘definitely not’ AMI. Clinicians weren’t blinded to the clinical history, ECG or point of care troponin results. AMI was diagnosed using ECG findings and two serial troponin assays taken 3 hours apart. Over one thousand three hundred patients were included in the analysis, of which nearly 15% had an AMI. Clinician gestalt for ruling out AMI using ‘definitely not’ as a cut off showed a sensitivity of over 98% and a negative predictive value of 95.0%, although the specificity was very low at just under 5%. For ruling in AMI using ‘definitely’ as a cut off, the specificity was over 98% and the positive predictive value was just over 60% (with a sensitivity of 18%). Cut offs using ‘probably’ or ‘probably not’ had much poorer sensitivities and specificities. Overall, it appears our clinician gestalt may be better than we realised in predicting cardiac events, although only when clinicians are confident about the diagnosis and have seen the results of early troponin testing.

Govind’s take home message

“Decision rules can be incredibly helpful in guiding clinical management, allowing us to understand the statistical risks and probabilities for the patient in front of us. The accuracy of clinician gestalt has not been looked at on this scale; interestingly Emergency Physician gestalt is highly accurate at ruling in and ruling out Acute Myocardial Infarction in patients that present with chest pain. I look forward to breaking down and presenting the analysis of this data further, please watch this space.”

Contact Govind about this work: [email protected]

“Point-of-care-testing for procalcitonin in diagnosis of bacterial infections in young infants” by Thomas Waterfield et al. (ST8 Paediatric Emergency Medicine Royal Belfast Hospital for Sick Children & Doctoral Research Fellow Queen’s University Belfast)

Sometimes the decision to treat an infant with intravenous antibiotics is easy: they are febrile, look unwell and have a clear source of infection. Often however, it’s more difficult than that and relies on the clinician’s judgement whilst awaiting the results of blood cultures. Procalcitonin has been increasingly investigated as a possible new marker for sepsis in both adults and children. This study looked at the diagnostic accuracy of procalcitonin when performed as a point of care test in febrile children less than 3 months old. 

The study was a prospective observational diagnostic study, conducted at a single site and including 124 children. 3.2% of children were found to have an invasive bacterial infection, as demonstrated on reference standard (blood culture or quantitative PCR). Procalcitonin was tested whilst the patient was in the Emergency Department using 0.5mls of whole blood and results were available within 20 minutes. Analysis showed the area under the ROC curve to be 0.97 with a 95% confidence interval (CI) of 0.94 – 1.0, representing a very accurate diagnostic test. When using a cut-off of 0.5ng/ml the test demonstrated a sensitivity of 100% and a specificity of 91%. 

The authors also examined the ability of procalcitonin to predict non-invasive bacterial infection, which was present in 10.3% of children studied. However, the test did not perform as well for this subset of children, with an area under the curve of 0.81 (CI 0.61 – 1.0), a sensitivity of 71% and a specificity of 97%. Still, the results in invasive bacterial infection are promising and we hope to see this validated in larger studies in the future.

Thomas’ take home message

 “Point-of-care testing for procalcitonin is highly accurate at identifying infants with invasive bacterial infection requiring immediate treatment in the Emergency Department using a cut-off of 0.5ng/ml.  Conversely in those infants who appear well and have a procalcitonin result of less than 0.5ng/ml it is appropriate to withhold immediate parenteral antibiotics and observe them in the Emergency Department pending the rest of their diagnostic assessment.”

Contact Thomas about this work: [email protected] or on twitter  @DrTomWaterfield 

The above work highlights how diverse and exciting research in Emergency Medicine can be. We hope it also shows that research can be done in different ways, accessed through different channels and undertaken at different stages of a clinical career. We congratulate all the authors on their shortlisting and wish them the very best of luck in their continuing work. Let’s hope we see more from them in the near future!