Mike Wells from Scotland Writes:
Your two excellent podcasts on C spine imaging really got me thinking.
I work in the UK where resource constraints within our public
healthcare system mean that even if I wanted to, I would not be able
to obtain CTs as the first imaging port of call for all my neck trauma
patients. I can however argue individual cases with the radiology
department and therefore effectively need to try to choose high risk
I pulled the Canadian C spine and NEXUS studies and looked back
through their methodology and results. In both studies ordering a neck
CT was at the discretion of the treating physicians – but most
patients only got plain C spine films (in CCspine 436 patients got CTs
= 7% of total patients who were imaged; for NEXUS I could find data to
allow me to make this calculation). I also went through the further
NEXUS study looking at missed fractures – another way of looking at
their data is that in the 581 patients with technically adequate C
spine films, only 3 unstable fractures were missed – giving a
sensitivity of 99.4% for the unstable injuries which I am most scared
I absolutely agree with you though that very often plain films are
technically inadequate and that their sensitivity is therefore much
However I would argue that the real sensitivity we are interested in
is not that of C spine films alone, but rather than the sensitivity of
the combination of plain C spine films and clinical examination and
acumen. CTs in the NEXUS and Canadian studies were after all ordered
at clinician discretion. It's possible that fractures were missed in
the patients who weren't scanned but both studies did seem to attempt
follow up (NEXUS in particular checked local ‘event logs' although I'm
not clear on what these are).
So I think over here in the NHS I would argue that in ‘minor' trauma
patients failing the CCspine rule I am still obliged to use plain C
spine films as my first imaging step. On the basis of what you have
said I'll will set the bar higher in terms of making sure films are
technically adequate (over here we still use Swimmer's views, which I
detest). However for patients with adequate films and the roughly 3%
prevalence of fractures in the group failing Canadian C spine, I would
hope that my clinical exam would then identify those patients with
normal films but underlying injuries.
Utimately I think from my view what this is about is not the
sensitivity of plain films on their own – which I agree is
unacceptably low – but about the sensitivity of plain films + clinical
Please feel free to put this in your comments section if you wish!
Thanks again for your fantastic podcast and blog.