If you’re an emergency physician practicing in the UK you may be a bit surprised to hear about a trial published this week looking at the use of open reduction and internal fixation of rib fractures, and more specifically flail chests. Surgical management of rib fractures is something that took place 40-60 years ago, but in the modern world it’s arguably not routine. So many thanks to @traumagasdoc for asking the question after it came up at a trauma meeting in London.
A quick look around the internet reveals some rather alarming photos from trauma.org of external fixation devices of the past. These external fixation devices remind me of some medieval torture devices and I guess it’s no surprise that they fell from favour.
On the other hand we know that there is a significant morbidity associated with flail chest in trauma. In addition to the underlying lung contusion the pain and potential mechanical instability of a large flail segment can lead to pneumonia, pain and a prolonged hospital stay. Pain, a known complication and a prolonged hospital stay would serve as just reasons to perform an open fixation for many orthopaedic injuries – so why not the ribs?
Following a short debate on Twitter it seems that surgical management has not dissapeared everywhere with colleagues in Australia and Yorkshire talking about it’s use in 2013. Perhaps our headline is a little disingenuous, in some places it has already returned (if it ever went away).
So, it’s interesting, it’s potentially useful, but what about some evidence that it works? Well lucky us there is a paper in the Journal of Critical Care on exactly this topic. Abstract below
So perhaps it is time to reconsider the possibility of surgical fixation for patients with blunt trauma.
[DDET What kind of paper is this?] This is a before – after study. The authors have used surgical fixation from 2009-2010 and compared outcomes against historical controls. The main outcomes were days on the ventilator and length of stay on ICU and in hospital. [/DDET]
[DDET Tell me about the patients] Good question, especially as this is not a randomised controlled trial. As we are looking at an intervention (surgical fixation) then an RCT would appear to be the right design. However, in this study they have compared historical patients with patients after deciding to use surgical fixation. Clearly this is not ideal as ‘other’ things may have happened over time that may affect the outcomes in this study. It’s a real problem when we appraise historical control studies – basically what else might have gone on to confound the results??
Patients were considered for surgical rescue therapy (SRF) if standard analgesia and physio support failed and patients ended up heading for ventilatory support. This is a reasonable group in my opinion, SRF is clearly invasive and has risks so focusing it as a therapy on those failing standard management seems wise.
We also need to consider the exclusions in this study as they are a little odd. Patients with a low GCS and those that died in hospital. The reasons for this are unclear, especially in hospital mortality as death is a jolly important outcome for patients and in severe injury studies I feel that it has an important place when looking at outcomes. They do state that they excluded death reporting as there were no deaths in the SRF group – but I’m not sure this is wise.
Lastly, the numbers here are very small. 11 controls and 10 intervention patients.
[DDET What does surgical fixation entail?] It seems fairly simple (to a non thoracic surgeon like me), involving a thoractomy and then a combination of plate/IM nailing of the ribs. I’m no surgeon – and it may be much more complex than it sounds, but perhaps not. [/DDET]
[DDET What do the results tell us?] The authors state that the 2 groups were similar in age, injury severity score, intensive care unit LOS, hospital LOS, total number of rib fractures, and total segmental rib fractures. However, the numbers are so small we would not expect to see a significant statistical difference unless the groups were very different indeed. For example, the average ISS score was 10 points higher in the non-surgical group, this difference was too small to be statistically significant, but it might be clinically significant.
The operative group demonstrated a significant reduction in total ventilator days as compared with the nonsurgical group (4.5 [0-30] vs 16.0 [4-40]; P = .040) with patients with SRF coming off the ventilator with a median of 1.5 days (0-8 days).
In fact all the outcomes they looked at were better following surgical fixation. Patients appeared to do better but as the numbers are so small no statistical difference was demonstrated.
So the results are impressive at first look….but….. [/DDET]
[DDET How confident can we be about the findings?] This is a key question. I got quite excited about this paper as it is current, interesting and potentially addressing a serious problem in our trauma patients. However, this is a paper with many biases, and of most concern I would mention the following.
- The before/after design will always be exposed to the potential that we are seeing the effects of time rather than intervention.
- The small numbers mean that we may be seeing a type 1 error.
- In the control group patients were not selected for therapy. In the intervention group some patients were declined for surgical fixation. This is not then a comparable group.
- Although the differences in groups are not statistically different, they look as if they may well be and I would want to know a lot more about the patient characteristics and their injury patterns.
[DDET So what does this really tell us?] Well, let’s not throw the baby out with the bath water. There are good pathophysiological reasons why this ‘should’ work and this early study suggests that there ‘might’ be a benefit. There are also papers out there that suggest an economic benefit to surgical treatment.
In addition, a recent meta-analysis of papers comparing surgical and non-surgical fixation suggests a benefit for patients in terms of ICU LOS and complications of the injury. However, there were only 2 RCTs in this analysis amongst 11 papers analysed.
As my good friend Bernard Foex tells me, this is a hypothesis generating study. I’d love to see more & even better I’d love to see an RCT, but that may not happen. The authors give a fair and balanced discussion of their findings with links to other case series in the literature. It’s reasonably well balanced and they mention many of the concerns raised in this review. I would agree that we would benefit from seeing data from a larger series of patients, and ideally from a well constructed clinical trial.
If nothing else I’m pleased to see the debate come again and for our surgical colleagues to keep thinking and researching techniques that may benefit our patients.
So, back to @traumagasdoc and his original question. We have answered the question about whether people are doing it (they clearly are) but this study does not as yet tell us whether it’s a brilliant idea or just a hypothesis. [/DDET]
- Surgical rib fixation for flail chest deformity improves liberation from mechanical ventilation. J Crit Care. 2013 Sep 24. pii: S0883-9441(13)00288-8. doi: 10.1016/j.jcrc.2013.08.003. [Epub ahead of print] Doben AR, Eriksson EA, Denlinger CE, Leon SM, Couillard DJ, Fakhry SM, Minshall CT.
- Surgical fixation vs nonoperative management of flail chest: a meta-analysis. J Am Coll Surg. 2013 Feb;216(2):302-11.e1. doi: 10.1016/j.jamcollsurg.2012.10.010. Epub 2012 Dec 5. . Slobogean GP, MacPherson CA, Sun T, Pelletier ME, Hameed SM.
- Rib fracture fixation for flail chest: what is the benefit? J Am Coll Surg. 2012 Aug;215(2):201-5. doi: 10.1016/j.jamcollsurg.2012.02.023. Epub 2012 May 4.Bhatnagar A, Mayberry J, Nirula R.