Thanks to Theo for adding to our shared experience of this problem, this time from the Hunter viewpoint:
"Our experience is similar to Liverpool, but on a smaller scale.    Decision to intubate is often prior to Immunology input, and far be it   from me to argue whether a patient is Mallampati score 3 or 4 with   someone who actual can tell the difference.    ACE-I are the first line antihypertensive for those with DM2 and   vascular disease, and some of those patients tend to be bariatric, which   can make one anxious, both about trickiness of emergency airway support   and also about those patients being intubated (often OSA, so   hypercarbia tolerant, so difficult to wean, need high inspiratory   pressures for the chest wall, high risk of VAP etc etc).    So, yes, when the occasional ED doctor rings prior to intubating a   patient with tongue swelling (as it typically is with ACE-I angioedema),   and asks what the risk of progression / death actually is – what do we   say?    Yes, try icatibant two doses (perhaps 3 if they are Pickwickian) – but   then they are "refractory" (which must mean "severe" right?).  But we   all have patients that wake (as Daman points out, it often is morning /   overnight), and travel 4 hours (ok, travel 1 hour, see a resident, then a   med reg, then they call us, but still four hours) and there is no   progression over that time (and no resolution – which means "refractory"   - right?).  So what is their actual risk?  Obviously I don't know, but   if several colleagues recall fatal episodes, well then early intubation   rather than observation is warranted.  That is the reason that I ask.    I would query the ethnic background for the Liverpool experience – the   increased US mortality in those with African background may reflect a   bias, and perhaps the risk is higher in other groups also? Thanks ,Theo "
    
