This week I am at the Trauma Care UK conference. If you’ve not been to this one then you should, it’s friendly, great value and aims to influence and engage with all aspects of trauma care from injury prevention right through to rehab. You can go for just the day if you want to focus on your area of interest or spread out and hear what the other parts of the trauma pathway are up to.
I’m delivering a short talk on the top 10 trauma papers over the last year or so which is always a pleasure to do. It gives me an excuse to go back over the literature and to ask myself if there is anything that we should be adopting in Virchester based on the latest evidence. I did a similar talk a couple of years ago which you can see here,
My usual approach is to search through the literature for those studies that can have an impact at the bedside, for papers that are relevant to the clinician and not just the researcher and if possible for those that might challenge our current practice.
Despite trauma being responsible for an enormous disease burden in our societies I was once again struck by the paucity of clinically relevant papers in the management of major trauma out there. If I was looking for theory or lab-based articles then I would have had no problem at all, there are many papers out there looking at stuff happening in test tubes, but I am naturally wary of these. Far too many therapies look good in the lab, but that does not transfer to clinical practice and so I’ve not really included them (though if you want to see some recommendations from Rich Carden on papers worth reading then I would start with these ones1–6 ).
So, my top 10 is intended to be real world, influential and pragmatic. You will no doubt disagree and I apologise if I have not included your personal favourite (or heaven forbid failed to include the paper you authored 😉 ). If you have better suggestions then please do add them in the comments.
1. The Diagnosis of Arterial Injury in trauma.7
I work in a trauma centre that sees more than its fair share of penetrating and blunt trauma. Many injuries, especially the penetrating ones, have the potential to cause vascular injury and as such it’s a real question as to when to perform angiography for these patients. For example, I recently had a patient with multiple stab wounds, those to the trunk and anterior neck were obviously attention grabbing and were going to be investigated by CT scan. The smaller wounds to the arms though? Perhaps they were less dramatic to look at but they might be harbouring a vascular injury too. Should we routinely image all of these wounds. The easy answer is yes. Just angio everything but for many reasons which I won’t go into here that’s not the right thing to do. Fortunately, there is a paper published in Academic Emergency Medicine that can help us.
In this systematic review and meta-analysis, the authors looks at which clinical and USS signs help predict significant vascular injury in patients with penetrating trauma. The papers are a little heterogenous in this review, and the authors state that there were biases and flaws which temper the results, but the bottom line appears to be that the next two statements are probably true for penetrating trauma.
- In patients with a normal physical exam (no hard or soft signs) and a normal ankle/brachial index then an arterial injury can be ruled out.
- In patients with a positive USS for arterial injury then it can be ruled in.
What I absolutely love about this paper is the way in which the authors have come to these statements based on an analytical and Bayesian approach to the evidence. This is evidence-based diagnostics in practice and if you want to understand how diagnostic tests really work then I would strongly recommend you read this paper.
We are doing a lot more Whole Body CTs as compared to a few years ago. You may remember the REACT-2 trial that recently suggested that it is acceptable to deliver a focused approach to CT scanning in trauma. That paper has led some to move away from WBCT to focused on the basis of a reduced radiation dose, but here at St Emlyn’s we were not convinced and we have not really changed practice.
However, there are other unintentional consequences of WBCT and that’s the identification of previously unknown pathologies. This is especially the case as we now WBCT more elderly patients and as a result it’s not uncommon for me to make a first cancer diagnosis on a trauma scan! They also pick up renal problems, funny looking lung nodules, liver changes etc. The list goes on and on, with the point being that once we’ve found these things we need to do something about it.
Great you might think, surely that’s great for the patient? Well perhaps not. There is a significant risk of over-diagnosis and the pursuit of unnecessary pathology. This was wonderfully illustrated at SMACC in Berlin when James Rippey told the stiory of what happened to his son after he found a retroperitoneal mass whilst trying out a new ultrasound machine. That story had months of angst, investigations and finally surgery for what turned out to be an incidentaloma!.
A re-analysis of the REACT-29 data in this paper8 looked at this and found some interesting results about incidental findgings. If you remember this was an RCT across 5 centres that randomised 1083 patients to focused or whole body CT after trauma. The bottom line for the differences were.
- For major incidental findings (may be life threatening) 4.3% vs. 1.7%
- For moderate incidental findings (may cause morbidity) 22.2% vs. 15.9%
- For minor incidental findings (hardly relevant) 31.8% vs. 23.8%
So this is important for two reasons. Firstly the chances of you finding incidental features on a CT scan is really high, probably higher than you think. Secondly, if that’s the case then what mechanisms do you have for the follow up of these patients.
In this study the authors were unable to determine if there were important clinical outcomes for patients as to quote
‘We could not demonstrate a significant difference in follow-up, which could be explained by low follow-up rates in
general and poor documentation of incidental findings and their management in trauma patients.’
That’s a really concerning statement. If you’ve worked in EM for long enough then you will have no doubt experienced the patient who comes back 6-12 months later with an inoperable cancer which was present on a chest X-Ray that was taken in your department, but which was missed. Those are tough days for all involved and my feeling is that this has an even greater potential to take place in the era of WBCT.
There is something of a nihilism about significant head injury in older adults. At this important site we try not to be tribalistic about inter-specialty conversations but there is no doubt that you can hear some of our neuro-surgical colleagues switch off and in some cases just interrupt with a ‘NO’ when you mention that your patient with a brain injury is aged over 75. It’s not all neurosurgeons and there are some exceptions who take a different approach but………,. if you’ve worked in EM for long enough then you will know how challenging and aggressive those conversations can be.
But is that nihilism really justified? Perhaps they are right and we should just give up on the basis of age? This question and others related to the outcomes of older patients with TBI is addressed in a TARN based paper in the EMJ in the last year10. Hawley et al looked retrospectively at all patients admitted to an MTC in the West Midlands (Coventry as it happens) who met TARN criteria (you can read more about TARN and the great work it does here).
Of the 4413 patients with major traumaI, 1389 were over 65 and of those 45% had a TBI, most commonly from a low energy fall. Interestingly and very positively most of these patients survived (78%) despite 80% of injuries being classified as moderate or severe. 28% of patient had surgery (I don’t know if this reflects a conservative or aggressive management style on the neurosurgery spectrum).
Predictors of mortality were unsurprisingly….
- Increasing age (17% aged 65-74, 19% aged 75-84, 30% aged over 85)
- Injury severity (unsurprisingly).
They also noted that head injuries in older adults can be missed if there are other pathologies and I interpret that as a reason to keep my threshold low for CT in older patients with a significant injury in another body area.
According to an important site, older adults that have received a severe head injury, will need special cares at home and assistance with their daily life.
Hang on though, you want to know about outcomes…. I think you’ll find these better than expected.
- 78% survived
- 55% went directly to their own home
- 57% discharged with a good Glasgow Outcome Score
The bottom line here is that the outcome from TBI in older patients is a lot better than most people would credit and that we should not be quite as nihilistic as some of our colleagues might suggest.
4. Mode of transportation and penetrating trauma.11
As I mentioned earlier, I work in a trauma centre with more than its fair share of penetrating trauma. It’s interesting work but it can also be awful when we see young and old lives taken too soon as a result of violence. Most of our patients come by ambulance, but a significant minority come by private transport. This can be alarming for the waiting room as patients may come through the door in all sorts of states. A patient walking through the main waiting room holding their intestines in their hands tends to be somewhat alarming to the gentry of Virchester.
Anyway, the point is that patients don’t always come by ambulance and that means that they do not benefit from prehospital care, or does it? On occasion I’ve been really frustrated by incredibly long scene times for patients who have been stabbed or shot even when they are very close to the ED. I struggle to understand scene times of over an hour for patients with penetrating injuries and shock who are within 10 mins of a trauma centre but it can and does happen. I was therefore really interested in this paper from the US that looked at outcomes in patients with penetrating trauma who self presented vs. arrived by ambulance.
This was again a retrospective data set (and yes I know there are problems with this) that examined outcomes from the National Trauma Data Bank over 12 years in the US. It’s big with over 103,000 patients included in the analysis. They compared outcomes for patients transported by private vehicle vs. ground EMS and found that the outcomes for private vehicle were much better.
- Private vs. Ground EMS 10.1% vs. 11.6%
- GSW only 17.2% vs. 19.3%
- Stab wounds only 2.4% vs. 2.9%
The results were adjusted in multivariate modelling to take account of a whole bunch of variables that we know might affect outcome such as physiological changes, ISS, age, sex, race, insurance etc. The overall mortality without risk adjustment was 2.2% for private vs. 11.6% for EMS. That difference is not so great when adjusted for known factors as seen above (which is why it’s important to do the adjustments).
All results were statistically significant and with no change when sub analyses conducted for large vs. small systems.
What does this mean for us here in the UK with a different population and a different trauma set up? I think there are principles that we can apply here. For me it makes me question what can be achieved in the prehospital environment for those patients who are at most risk of dying from penetrating trauma. Unless definitive procedures are being performed at the scene (e.g.thoracotomy for traumatic cardiac arrest) then we really need to think hard about what we do. I don’t want to see any more patients with long scene times as a result of interventions that we know don’t make a difference. Penetrating injury may well need a scoop and run policy (puts hard hat on and runs for cover).
In adults, and especially older adults we are pretty liberal with the use of WBCT. When recently challenged about whether we should CT a 96 year old who had fallen down the stairs, or whether we should just focus on the areas where we thought the patient had pathology it was difficult to resist the sarcasm reply of ‘You’re right a WBCT might increase her risk of cancer in about 20 years time’ I resisted on that occasion, but it’s a very different story in kids, and especially young children where we have real concerns about the over use of WBCT in a population who do have time and opportunity to develop radiation induced cancers as a result.
We’ve covered this on St Emlyn’s before, recommending that you follow the guidance of the RCR13 in taking a more focused and thoughtful approach to WBCT in kids. I work in the busiest paediatric trauma centre in the UK and even then I rarely, very rarely, do a WBCT in children. The rare occasions are when a child is intubated and ventilated with a mechanism that could result in multisystem injuries and where their predicted mortality is really high (e.g. a child hit at high speed on a motorway who is now unconscious).
What about the evidence though? Guidelines and good will are one thing, but we could do with some numbers to back up our opinions.
This is another study from the National Centre Trauma Bank14 that Natalie reviewed on the blog recently. In essence they looked at trauma outcomes for those patients in the paediatric cohorts who received focused or WBCTin practice. They used propensity score matching to ensure that the comparison was as like for like as possible.
The bottom line here is that there was no significant difference in mortality between those who received a WBCT and those who had focused investigations. In this group of patients who have a lifetime ahead of them it thus makes sense to be thoughtful and cautious in the use of WBCT in kids.
However, data from TARN also tells us that in the UK non-specialist paediatric trauma centres are more likely to WBCT than specialist Paed MTCs (and outcomes are better in the specialist units). So we have more evidence, we have the guidelines, so we need to deliver.
6. Blood pressure thresholds in head injury15
I suspect that like me you aim for an arterial pressure of 90mmHg in your head injured patients. This dogmatic value has been around for many years and in a way it makes sense, we know that in traumatic brain injury the cerebral perfusion lacks it’s normal autoregulation and becomes more pressure dependent.
In the UK I’ve often been told that keeping the MAP up, elevating the head of the bed and taking the collar off are well established therapies, but where does the 90mmHg come from, are we talking about mean or systolic pressures and is it real? If we delve into the recommendations in the past we will often find that although the evidence is there for lower BPs linked to worsening outcomes then the evidence underlying it has been weak.
It’s important as for many trauma patients there will be challenges in maintaining BP as the tools to do so are limited. More crystalloid anyone? Well we can do that, but if they are also bleeding then we don’t think that’s a good idea either.
In this paper from the US they looked at outcomes for patients with moderate to severe brain injury as compared to their systolic blood pressures in the prehospital setting. Amongst 3844 patients they plotted survival against the lowest SBP. If 90mmHg was an important number then you would expect an inflection in mortality at or around the 90mmHg point.
What they found is that there is no such inflection. In fact the relationship between SBP and outcome is linear with mortality with the odds ratio of death increasing by about 18% for every 10mmHg decrease in SBP.
The 90mmHg appears to be dogma in this study, but what does that mean in practice? What this study does not tell us is whether intervening to a higher or lower BP is of benefit. It might be that we cannot influence this and it is just a feature of the underlying illness, we can’t be sure at this time. For me I think this does ask questions of the 90mmHg threshold and it will encourage me to think twice before being satisfied with such a value in my TBI patients. I also think that it will influence my management of multisystem injured patients where in the past I have wrestled with the importance of controlled BP for bleeding vs higher BPs for head injury. In that group I think this will redress the balance in favour of the head injury unless there is obvious uncontrolled bleeding from another source.
7. Apnoeic Oxygenation in the ED.16
There has been quite a kerfuffle about the use of apnoeic oxygenation in emergency17 RSI over the last few years. It’s been popular amongst #FOAMites and has appeared in some form or another at many conferences. It’s a bit cool and trendy in a way, popping the nasal cannulae on the patient to apnoeically oxygenate the patient during the period between induction and ventilation. I’ve used it in quite a lot of patients as it seems to be a good idea and the chances of harm seemed very small, but we have been sceptical. When tested in the real world the impact of the technique on ICU (the FELLOW trial is reviewed here) was not as impressive as some had hoped, but what of the ED?
In the ENDAO trial 206 patients were randomised to apnoeic oxygenation or not in the ED settings. Admittedly not all of these were trauma patients, and I know this is a trauma papers top 10, but I think it is arguable still relevant as this is a high risk procedure in trauma care.
They looked at a principle outcome of lowest oxygen level recorded during the RSI procedure. Of the patients intubated there was no significant difference between the two groups which would make you think that there is no point in doing it. That is certainly one interpretation but there are cautions. Firstly both groups were really well preoxygenated and so we would not necessarily think that there would be a difference and secondly the incidence of significant desaturation was tiny and not enough to show a difference.
What does this mean for me in the ED? Basically that if the patient can be preoxygenated and does not have pathology that is likely to make them fall off a cliff in terms of sats then I guess we don’t need it. On the other hand, if I’m worried, if preoxygenation is tough or if I think the intubation is going to be prolonged then I will continue to use the technique. You might decide that as a service you want to make everything the same, all the time, and always do this and that’s fine. For me, I’ll consider it always but use it thoughtfully.
8. IV contrast in CT scanning.18
Again, this is not strictly trauma, but it is really important to those of us who order imaging for our trauma patients. The question of contrast induced nephropathy has been a real challenge at times. The concern is that patients who have a reduced renal function are at risk of renal failure if we give them contrast. This is embedded in the psyche of many radiology departments and it’s not possible to get a contrast enhanced scan without evidence of the eGFR and a contract written in the blood of your children that you will take all responsibility for any renal impairment that results. In Virchester our really vulnerable bed base often has patients admitted for 24 hours IV hydration in order to facilitate a CT scan if they are known to have a low eGFR. In our trauma patients then the really sick ones are easy as no-one argues the case, but in the subacute patients in whom we want a CT, but not immediately there is often a delay and disagreement about when, how and if we can get a CT based on the patient’s baseline renal function.
Is this true though? Does contrast really cause renal failure and if so what can we do about it?
In this study published in the NEJM they looked at the rates of AKI, renal failure, need for dialysis after a contrast enhanced CT scan. They used propensity matching to adjust for Illness severity and underlying pathology. The used case controls of patients in the ED who did not get contrast, but who also had blood results available in similar time periods to compare.
- No difference in the rates of Contrast Induced Nephropathy between those given contrast and those who did not receive contrast. (OR 0.85-1.08)
- No difference in AKI (OR 0.87-1.16)
- No association with CKD, dialysis or transplantation if given contrast.
They did exclude patients with a creatinine greater than 4mg/dL (which is about 350 umol/L in UK units). Although such patients were not included in the primary analysis they did report them in a separate table and although the numbers are smaller there was again no obvious difference. There are some concerns in that with this being a retrsospective study there may have been effects such as avoidance of contrast in high risk patients or use of more crystalloid amongst some patients prior to scanning, the results may be biased, but if so there is little evidence to support this. The authors rightly conclude that there is at least sufficient uncertainty to warrant an RCT of contrast in these patients.
The bottom line appears to be that there is little evidence for contrast induced nephropathy and that our suspicions about this condition in the past were probably a result of the underlying pathology rather than the contrast. As the evidence currently stands you should just get on with that CT scan.
9. Does every traumatic Pneumothorax need a chest tube?19
When I was a junior doc every traumatic pneumothorax got a chest drain, usually a nice big one, about a 32F would do inserted in the 5th intercostal space with a modicum of local anaesthetic and a lot of pain. In retrospect I do wonder whether inserting the drain was more for my benefit in getting the experience than for the patient who really did not enjoy the process.
In recent years we have been much more conservative about putting drains in patients with traumatic pneumothoraces, especially the teeny, tiny ones we only see on CT scan. That’s even been the case in those patients undergoing positive pressure ventilation, who are traditionally seen as exceptionally high risk. You can read more about our top tips for chest drainage here.20
This paper19 gives our current custom and practice some evidential back up as it’s a review of 602 patients admitted through a single UK trauma centre over a four-year period. Within that time 277 of them were treated conservatively and of those 277 only 25 subsequently needed a chest drain. It’s not that this was a simple group either, the average ISS was 26 and the in patient mortality in this group was 9%.
Amongst the 62 patients who had positive pressure ventilation then again 90% of the patients did not need a chest drain.
The authors used a propensity score adjustment to take account of known factors leading to advserse outcomes and again there was no difference (a hazard ratio of 1.1).
The bottom line appears to be that if the patient does not obviously need a tube in the early stages of their resuscitation and you are only thinking about it because of something you’ve seen on the CT or CXR, then maybe they don’t need it at all and if you can observe them carefully with the ability to intervene quickly should problems arise then conservative management is the way forward.
10. Human Factors are key to Trauma team success.
I’m delighted to end with a paper that arguably is more influential and potentially more applicable to all our practices than any of those previously listed. When we talk about trauma, or even emergency medicine in general we often focus on interventions, drugs and procedures whilst leaving the influence of ourselves and our behaviours out of the mix. This is crazy as every decision, every action, every inter-action even must involve our cognitive processes expressed through our actions and behaviours.
Every resus is affected by how we function as individsuals and teams. If you don’t believe me then read a few critical incident reviews and you’ll soon find that it’s the cognitive and behavioural aspects of our job that most often leave ourselves and our patients in trouble. The flip side is that when teams work well it’s not just luck, it’s because we know how to think, behave and interact to the benefit of everyone there. The bottom line is that human factors in their wider sense really count in the resus room.
My last paper is therefore an excellent review by the wonderful Chris Hicks and Andrew Petrosniak, trauma team leaders from Toronto in Emergency Clinics of North America.
Trauma is, and always will be, a team sport. Everything we do in the resus room is done in the company of others and that dynamic must be learned, monitored and reviewed. In this paper the authors build on the work of Cliff Reid and others in developing the idea of managing yourself, the team, the environment and the system to optimise patient care.
This article is something that everyone who works in the resus room should read and share. We should all stop and think about how we are the intervention in many cases. Just as we would optimise our patient for a procedure, so should we optimise ourselves to respond. This is a great place to start, and of course if you want to know more then there is plenty of links in the #FOAMed world, and also an excellent paper in press on the Zero -Point survey to guide you into the preparation phase of any resuscitation.
The bottom line here is that if you want to make the biggest difference to the trauma care of your patients, then the place to start is with the improvement of you.
It has been another lean year for high quality trauma papers. I am still underwhelmed by the numbers of high quality, methodologically sound studies that seek to advance trauma care over the next decade. Although that’s a disappointment it’s not an excuse to develop our practice. There are plenty of lessons to be learned from the papers above to improve patient diagnostics, patient care and the performance of you and your team.
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
8 thoughts on “Top 10 trauma papers 2017-2018 for @traumacareUK conference. St Emlyn’s”
Pingback: Trauma Care Conference 2018 – Wednesday – Trauma Resuscitation Anaesthesia
Pingback: LITFL Review 328 | Edwin M. Thames
Pingback: Top 10 Trauma Papers 2018 – The Resus Room
Pingback: St Emlyn's April 2018 blog and podcast round up. - St.Emlyn's
Pingback: Junior Doctor Teaching – Head Injuries – Barnsley Emergency Department Education
Do you think the conclusions from the whole body CT scans in kids study was a touch strongly worded? In the adjusted analysis, the absolute risk difference was reported as -0.2% with a 95% confidence interval ranging from -0.6% to 0.1%. Though technically not statistically significant, this is somewhat of an imprecise confidence interval, suggesting that if we assume this population is representative (do you think overall mortality of 0.9% is the trauma population we are interested in?), it could actually be that whole-body CT scans actually reduce mortality by 0.6%.
Could this effect be magnified in a higher risk population? The confidence intervals are even wider in these subgroups. I’m struggling to understand how there could be a signal for improvement in overall mortality for whole body CT scans, yet in the higher risk groups (GCS < 9, hypotensive), seemingly all of a sudden whole body CT scans are harmful (whereas selected CT scans have a signal towards improved mortality). It somewhat defies logic unless there is some bias leading to under resuscitation of whole body CT patients. A whole body CT scan in my shop would not add substantially to the amount of time spent at the CT department.
Sometimes I think the never CT paediatric trauma patients narrative is a bit over the top… I think clinical judgment trumps all. If you are worried about your patient, just get the scan.
Pingback: #SASEM2019: Top 10+1 papers • St Emlyn's
Pingback: Del 1: Den ældre patient i akutmodtagelsen – Akutmedicineren.dk