Top 10+1 trauma papers 2016-17. St.Emlyn’s

This week I am off to one of the most enjoyable conferences I can get to in the UK. The Trauma Care conference is usually held around the Midlands and brings together everyone involved in trauma care for 4 days of learning and fun. My role is usually to go and present the top 10 trauma papers of the year and so it is again this year (2016-17 and a couple sneakingin fom late 2015). In previous years I’ve really enjoyed this¹ and I’m looking forward to it again in 2017.

The system is fairly well established now. I use a combination of regular PubMed searches, together with #FOAMed sites such as REBELEM, The Bottom Line, RESUS.ME, EMLitofNote, and of course the wider St.Emlyn’s team to scan the literature from 2016 up until the present day. I look for articles that ideally have three features. I want them to be able to change what we do, be of high quality and tio be interesting to the audience. This is usually an easy tasks as there are lots of papers out there, but not so much this year. I struggled to find those game changing papers around trauma which really surprised me. Trauma is a huge topic, with an enormous societal impact and yet the number of high quality RCTs is pretty damn small, especially so around the initial management and resuscitation of patients. Having said that I think we have found 10 papers that are interesting and useful to know about. If you find some more then let me know. We can always add in some extras 😉

Here’s my top 10.

1. The REACT-2 trial.² Do we always need a whole body CT (WBCT)? I worry about the amount of radiation we use, especially in younger trauma patients. Past studies have shown that WBCT can reduce mortality but I’m sure you’ve had the experience of ordering one on the basis of mechanism of injury only to find minimal or even no injury. You’ve just exposed your patient to lots of radiation and you should feel bad. Well perhaps not as you really don’t want to miss anything either. In this RCT trauma patients were randomised to either WBCT or a more focused appoach. 1403 patients were randomised on fairly liberal criteria (as compared to Virchester)  and in terms of mortality there was no difference. There was slightly less radiation in the selective group but relatively little difference equivalent to on average 0.3mSv which in a 20 year old would cause one extra cancer for every 28000 scans (you read that right so let’s not get our knickers in a twist).
2. Impact Brain Apnoea.³ A bit of an indulgence this as it’s a paper I co-authored. It’s also in memory of John Hinds who we lost last year in a tragic motorcycle accident. I like this paper for lots of reasons, but in particular the realisation that  important pathophysiological mechanisms can be lost to history only to re-emerge as technology (in this case motorsport) make them relevant again. If you don’t already know impact brain apnoea is a condition where a blow to the head with minimal anatomical damage leads to apnoea and death in the absence of intervention. I’ve seen it at the racetrack as have others involved with prehospital care close to the time of impact. Bottom line here is that we need to ensure that patients with apnoea post trauma get their As and Bs sorted quickly and probably by bystanders. Maybe IBA should be part of ALS training in schools?
3. The RESCUE-ICP⁴ trial looked at the use of the use of decompressive craniotomy for patients with refractory intracranial hypertension post head injury. They only got into the trial if level 1 and 2 treatments had failed (and interestingly that included hypothermia in some – and we suspect that does not work⁵). It was a trial I was vaguely involved in in the past and although conducted in 20 countries, 71% of patients were recruited here in the UK. Unsurprisingly the 408 patients randomised were largely young men with diffuse injury accounting for 2/3rds of patients. The results of this trial have been really quite controversial and I’m sure your interpretation will have more to do with your personal values than the numbers. In brief there was a clear survival advantage to craniotomy with 48.9% of patients surviving vs. 26.9%. However, there was no improvement in good outcomes. All the patients who would otherwise would have died were left with disability. Our fabulous friends at thebottomline have crunched the numbers and shown that for every 100 patients treated with craniotomy you would expect 22 more survivors, 6 would be in a vegetitive state, 8 would have a lower severe disability and 8 would have an upper severe disability or better (on GOS-E scale⁶). So the question is, should we swap death for disability, and in many cases severe disability. These are tough ethical questions and it’s far from clear whether we know who best to offer this therapy. I’ll probably mention the STITCH trial⁷ here as well (although it’s 2015 really). It’s a randomised controlled trial of initially conservative management of patients with intracranial traumatic bleeds. The patients were those where the treating clinician was in equipoise about whether a craniotomy was indicated or not. Patients were randomised either to early surgery (within 12 hours) or conservative management. The trial was sadly stopped early because of recruitment issues, but an analysis of the available data showed a non statistically significant benefit in terms of Glasgow Outcome Score for early surgery (10.5% so an NNT of 10) at 6 months. The secondary outcome of mortality was better with an 18% difference and this did reach statistical signifance and an NNT of 6 (though wide confidence intervals on this). However, like the RESCUE-ICP trial there was little improvement in patients with a good outcome. All the additional survivors were left with disability (though less severe than in RESCUE-ICP). It’s a great shame this trial was stopped early, that there were changes to the protocol as it ran and that many patients did not receive treatment as randomised. It looks as though we might have to do this again, BUT it does tell us that conservative management may not be as safe as some (in my experience) referrals to neurosurgery suggest!
4. Coagulopathy is a significant cause of mortality and morbidity for our trauma patients (see paper 11), but how to measure and treat it? Traditionally we use tests such as PT, APTT, FDP to track coagulopathy, but there is good evidence that they take too long and don’t really tell us everything we need to know. Thromboelastography (TEG/ROTEM) is a way of looking at clot activation, formation and destruction in real time and has been a standard of care in cardiac surgery for some time. In trauma it kind of makes sense to do the same but there has been little fantastic evidence for this in trauma and other non elective settings⁸. An interesting RCT from a US trauma centre⁹ looks at using either TEG or traditional markers to resuscitate patients. 111 patients who were eligible for the major transfusion protocol were randomised, and they found a difference in survival of 36.4% vs 19.6%. That’s impressive but perhaps a little too good to believe (maybe). However, this is a good step on the road to better evidence.
5. UK Traumatic Cardiac Arrest stats¹⁰ are up next with an excellent review of the epidemiology and outcomes here in the UK. Ed Barnard and colleagues have put together a great paper using the TARN database from 2009-2015. In that period 705 patients out of 227,944 on the database had a traumatic cardiac arrest with an overall 30-day survival of 7.5%. Interestingly the success rate of survival prehospital was greater than that in hospital (not sure why) and no patients survived who were in cardiac arrest at scene and in the ED. These stats can look pretty depressing but it’s not a million miles away from the success rates of medical causes of cardiac arrest. There may well be a question of inclusion bias in this study as some prehospital cardiac arrests won’t make it into the TARN database but despite this it’s fair to say that we should be aggressive in our management of these patients.
6. In terms of what to do, then controversy exists about whether closed chest compressions are effective in traumatic cardiac arrest. The logic is that if the heart is empty and/or there are bits of broken rib/chest around then there is no point and indeed a likelihood of harm. Looking at the stats above one suspects that an RCT is not going to happen anytime soon (too few patients) and so we must use wisdom and observational data to decide what to do. Here at St.Emlyn’s we’ve advocated thoracotomy and direct cardiac massage for traumatic cardiac arrest¹¹, but a paper in the Journal of Trauma questions this. In an observational study in a US level 1 trauma centre researchers used ETCO2 to determine the effectiveness of closed vs open cardiac massage. ETCO2 is a widely used marker of CPR effectiveness and interestingly the authors found no difference, and no difference in survivors either.  That could be a small numbers issue and open techniques looked slightly better in penetrating trauma. So what should we do? I suspect we should carry on as we are, onthe basis of the pathophysiological arguments until something better comes along. As Rich Carden suggests on the St.Emlyn’s blog, that may be REBOA, but we don’t really have the evidence for that either as yet!
7. If you want to start a fight on Twitter then mention something about emergency RSIs….. a whole plethora of opinions will gush forth and to be honest most of it is just that, opinion. I’ve gotten a little bored of it at times, but that does not stop the St.Emlyn’s team looking out for any evidence that can help us make better decisions about issues such as drug choice. A regular ‘firestarter’ issues is the Roc vs Sux debate. You will no doubt have heard the phrase ‘Roc Rocks, Sux Sucks’… well maybe. We could do with some evidence and there is some out there on BestBets¹² , Cochrane¹³ and #FOAMed sites^14,15 Unsurprisingly they all sort of disagree with each other. There are no fabulous RCTs out there and so we have to look at observational data such as this paper reviewed by Swami from Pantawala in the US¹⁶. It is observational but in 260 patients with traumatic brain injury they saw a signficant difference in mortality amongst those with severe head injury (44% vs 23%). It’s interesting and will get the Roc fans excited, but in all honesty, and even though there is a plausible pathophysiological mechanism, this may well just be association rather than causation.
8. If you’re working in the UK or US then you will know that Bariatric trauma is a real issue. The larger patient can present problems in assessment and management and it’s something we are seeing more frequently. It even affects our ability to do procedures due to a change in body size. Access to the chest is a notable problem, as it’s notably easier to get a chest drain in a skinny male than a very large female, I’m not making any judgements here, it just is. Similarly getting access using the Intra-osseous route can also be challenging. Swami (on the REBELEM site) found an interesting paper¹⁷ this year looking at exactly this using USS which shows that as your patient approaches a BMI of 43, or you can’t feel the tibial tuberosity then a standard 25mm (blue) IO may not be adequate and you should reach for the 45mm (yellow) one.
9. On a similar theme debates about the ‘right’ location for the decompression of tension pneumothoraces has raged for years. Personally my approach is to find a needle big enough to reach the pleural cavity and to just go for it, though I recognise that’s a little unscientific. A number of trials have been published on this and so it is good to see a collation of them in a systematic review and meta-analysis on this in 2016¹⁸. Overall we have 15 studies and over 6000 measurements o look at with 15 studies examining chest wall thickness and 13 looking at the effectiveness of needle thoracostomy at mid clavicular line, anterior axillary line and mid axillary line. The bottom line is that you are least likely to fail in the anterior axillary line (13% for AAl, 31% for MAL, 38% for MCL). For me it was also fairly sobering to read the complication rates from this technique which are potentially as high as 9% if you use abig needle (8cm) in the MCL, maybe the big needle and a good strong arm technique is not so wise after all. Read a more detailed review from Salim here¹⁹.
10. Lastly, and because the conference is being held on the Friday I think it’s appropriate for us to have a look at the weekend effect in major trauma.Unless you’ve been living under a rock then you will know that the idea of a ‘weekend effect’, in that you are more likely to die if admitted during the weekend is somewhat controversial. What about trauma patients though? Is there an effect for those patients with severe injury? David Metcalfe and colleagues looked across 22 UK major trauma centres and found no difference in mortality (8.4% vs 7.9% p=NS but actually better for the weekend) irrespective of whether the patient is admitted at night or during the day²⁰. This was leapt upon by the more politically minded as evidence that there is no weekend effect for any patients, arguably that’s not the case, and may be precisely the opposite. Trauma centres have been designed to deliver 24 hour cover, 7 days a week. In Virchester it now means that as a consultant I am resident on call and respond to all trauma patients within 10 minutes of arrival. Perhaps this paper tells us that if you have a well resourced system, with consultant involvement (from all specialities) early in the disease process then you get better outcomes. So watch this space for all our emergency care, not just trauma, perhaps the revelation is that if you fund and support an urgent care system things get better….. now who could possibly have believed that!
11. OK, I lied there is one more ‘REALLY?’ paper to finish off. We all know that trauma patients are at risk of coagulopathy and that this needs to be identified and managed well. In many centres clinicians rely on PT and APTT tests (which are flawed) and in other units clinicians are starting to use thromboelastography (TEG and ROTEM²¹). A rather interesting paper on ‘Impact Thromboelastometry’ has been spotted by Rich Carden. The concept here is that a tube containing blood with a poor clot is more likely to smash if dropped on the floor. In this study on Impact thromboelastometry ²² used blood diluted with saline to simulate coagulopathy. They then dropped them from a 75cm height onto the cap end to see if they broke. Interestingly they showed that the more dilute the blood then the more likely they are to be coagulopathic. An interesting finding, it’s certainly point of care and simple, but it’s not time to start chucking blood around the resus room.

So at the end of a review of papers over the last year or so I’m again struck by the paucity of really high quality data out there that is going to make a change to my day to day practice as an emergency physician in a UK trauma centre. I’m not despondent though. There are a number of trials out there and recruiting that may well change what we do in the next few years, of note CRYOSTAT²³ will be recruiting soon and CRASH-3²⁴ should be out in the next year or so and that’s great as who would want to think that there was nothing left to know 🙂

vb

S

@EMManchester

Before you go please don’t forget to…

• Subscribe to the blog (look top right for the link)
• Subscribe to our PODCAST on iTunes
• Follow us on twitter @stemlyns
• See our best pics and photos on Instagram
• PLEASE Like us on Facebook
• Find out more about the St.Emlyn’s team

References