Authors: Rob Hirst, Andy Neill, Dave McCreary Codes: EnC2, RP1, SLO4, TP1 / Published: 04/09/2023

Clinical question

  • Can we treat DKA with sub cut insulin only.


  • DKA is a common ED presentation and is a lovely pathology to treat if not more than a little unpleasant for the patient. Many presentations meet criteria for DKA but aren’t really that sick and maybe they don’t need all the infusions swe use for the sicker DKA patients. This paper looks at whether we can use the insulin in a less invasive (and much more familiar way to the patient) to manage DKA.


  • The SQuID protocol (subcutaneous insulin in diabetic ketoacidosis): Impacts on ED operational metrics


  • Griffey et al September 2022


  • this is a pre post study from the states in a single centre. Though it all seems retrospective. It seems that sub cut insulin was already accepted and they simply looked back at how they did. This is less than ideal as a way of proving something works.
  • They enrolled mild to moderate but excluded severe, (identified by a pH<7 and HCO<10) so they could still be nasty enough compared with historical controls
  • Protocol involved 2L fluid and glucose every 2 hrs (as opposed to eveery hr). They give 0.2units/kg of short acting insuling on arrival and then adjustable amounts depending on where the glucose end up.
  • They went to ward level care instead of a critical care unit (which was required if an insulin infusion was needed)
  • Primary outcome was operational here. They looked at ED LOS primarily. (the big issue here is that ED LOS is determined by many more things than clinical issues, it may be there are lots of ward beds and less ICU beds)


  • 180 patient with roughly half and half in standard vs the squid. But remember these were retrospective and i’m unclear why some got the squid and others the IV insulin. They compare that 180 with around 320 historic controls (who all got the IV insulin)
  • Those on the sub cut protocol spent less time in the ED. This is hardly surprising


  • This is not a study to prove much I’m afraid but the concept is intriguing and is crying out for a non inferiority trial
  • People with DKA do really well, we can treat this and they almost all get better. So this is the poster child for a condition where we probably over do it in some and coudl reduced the intensity of treatment without adverse outcomes.
  • I know in many hosptials i’e worked an insulin infusion has limited placement options within the hospital and in my own insitution many of them end up in my HDU as the altenrnatives are fairly limited. if we could safely treat this with less intensity then that would be great.
  • They all end up in some HDU like environment in my place. i don’t know what experience elsewhere is but it seems to be standard of care for these patients to end up with art lines as a means of regular sampling without further needles. . One of my now retired colleagues pointed out that this was all a little silly as we have no need for arterial access here and they’d be much better with some sort of durable multilumen IV access like a CVC or a PICC. So this is a little campaign of mine to stop putting art lines in these poor people and instead put in a CVC or even better a bedside PICC or mid line.