This paper received a fair amount of coverage from social media, and I would really recommend that you listen to this podcast (also available in iTunes) where Mark Wilson (who is a neurosurgeon but also a HEMS doctor) discusses his work, as well as giving some fascinating insights into management of TBI from the neurosurgical perspective.
The headline result from this systematic review and meta-analysis was that a patient with an extradural haematoma and BFDPs who goes to theatre has only a 29.7% mortality rate, and if they survive will in 54.3% of cases have a favourable outcome.
The group included papers from the last 20 years meeting predefined criteria. Unfortunately, the lack of primary data excluded 11 of 16 possible studies which is a problem seen all too often. There are campaigns for researchers to be required to make all their data available to the journals at time of submission (motivated by a desire to reduce research fraud!) which will hopefully help avoid this in the future. For outcome, they used the Glasgow outcome scale; more information about that can be found here.
The papers remaining were observational retrospective cohort studies, and the numbers small. The paper's conclusions are based on only 57 patients with subdural haematoma, and 25 with extradural (which explains the wide confidence intervals).
The elephant in the room is that these patients were ones who the neurosurgeons, for unknown reasons, selected to operate on. We don't know if they were younger, what the time interval to surgery was etc. Selection bias is therefore a massive feature and there are clearly confounding factors.
Death is also a slightly odd measure in these studies, as it is with work considering patients post cardiac arrest. These patients don't neccesarily die despite full and active management, they die because they have active management withdrawn, usually on an expectation of poor functional outcome. So death therefore becomes a surrogate for "the team don't think they will get better" which opens up to self-fulfilling prophecies etc.
Greater credance could be applied to the results if they mirrored those found elsewhere. Whilst not using the methodology of a meta-analysis, the brain trauma foundation have produced guidelines regarding early prognostication in TBI. They are currently being updated as they take their information from papers pre-1995 (so pre widespead CT scanning for example). They too highlight the difference in prognosis between extra and subdural haematoma; reporting results such as a mortality in postoperative patients with extradural and BFDPs of 56%, compaired with 88% in the subdural group.
So where are we left? To a certain extent that depends on whether you are an optimist and a risk taker, as well as where your threshold for an 'acceptable outcome' lies. If I have BFDPs and a subdural, please leave me alone. If I have an extradural, please operate but do it quickly (regardless of whether there's evidence it must be better!). However, if I don't recover my neurology quickly, paliative care is fine by me. Whether you want me to have ICP monitoring, decompressive craniotomy etc, I'll leave to the neuro-intensivists and surgeons.
Finally, just some facts for exams etc:
The proper name for unequal pupils is anisocoria.
A difference is a difference of 1mm or more.
A response to light is a 1mm change or more.
A dilated pupil is 4mm or bigger.
Other causes of dilated pupils include anticholinergics, orbital trauma, and various 'eye diseases'.
Pupils should be assessed post haemodynamic resuscitation for reliability.
Whilst surgery is the only way to 'rectify' the situation, you should know about hyperosmolar therapy, hyperventilation etc.
These are, as always, just my impressions so please comment below to let everyone know what your take on the paper or the topic is..... I might also see if I can get an expert or two to comment to see how this effects their practice if at all.