There has already been some great discussion online so be sure and check it out.
- EMCrit interviews the lead author
- HippoEM Blog has Weingart and Orman discussing it.
So there’s no shortage of great editorialising already out there.
That being said, I wrote most of this on the plane on the way home from SMACC so figured I may at least get it out there.
Sepsis is big and sexy in critical care. It covers a huge range of patients and has turned individual conditions into one clinical syndrome. This in itself is controversial. Why should we treat pneumonia and gall bladder sepsis with the same bundle. While sepsis evangelism has undoubtedly raised awareness and probably improved care are we dumbing it down? Anyhow, that’s a little bit off topic.
For now the big focus in ED sepsis research is on working out which parts of a bundle of care that includes very different interventions actually makes a difference. Do we need central venous oxygen saturations or is it as simple as antibiotics and fluids? (Never mind the question of which and how much fluid…)
One of three trials (ARISE and PROMISE still to come) looking at early sepsis management, this is the US led version.
- multi centre RCT, non blinded (as expected, it’s difficult to blind an Edwards catheter)(though as is good practice, the investigators were blinded when doing the analysis)
- note that these were big academic centres and a requirement to get your centre into the study was use of lactate to screen for cryptogenic sepsis and adherence to the SSC guidelines. The SSC guidelines are not endorsed by Australasia as far I’m aware so this in itself might be a marked difference in practice. What do you think?
- to get into the trial you had to be early in your sepsis course and have had a decent bolus of fluids
- three groups:
- a protocol based on rivers EGDT model
- a protocol based on international opinion
- usual care which would be whatever the doc would normally do.
- These protocols are as always in a supplementary appendix which you have to search for separately. This is an real personal bug bear of mine – everyone is reading this electronically and keeping the appendix separate for space saving reasons seem poor form
- Lead investigators could enrol and manage patients in both protocol based groups but not in the usual care group. I understand why but being looked after by a leading researcher in sepsis is an intervention in itself. [See EMCrits discussion of this on the podcast]
- primary outcome was in hospital death.
- they powered the study on the assumption that they’d have a 30-45% mortality (as was the case in the Rivers trial), as happens in a lot of trials this overestimated how sick their patients would be.
- 31 centres screening 12000 pts and got 1300.
- no difference in primary outcome (60 day in hospital mortality) 21% v 18.2% v 18.9% for EGDT, protocol, usual care respectively.
- they report a significant difference in fluid administration but I don’t think the difference is clinically significant. [EMCrit agrees even though the lead author sees a big difference in the three groups] We’re talking a litre difference between the most and the least with the other group half way between. If one group was 6l and one was 2l then that would seem to me a clinically significant difference. Note that the paper reports the numbers as 2.3, 3.3 and 2.8. These figures don’t include the mandatory fluid bolus that was pawrt of the inclusion criteria. So in reality it was more like 4-5 l in each group (HT EMcrit and WessexICS for pointing this out top me.
- there was a bump in renal failure in the protocol based group but this may well be just noise
- the only other big difference that caught my eye was differences in transfusion with 10% in the EGDT group and minimal in the other.
- there was slightly more use of vasopressors in the two protocol used groups but not much.
- there was an understandably big difference in central line usage. The EDGT protocol mandated a central line whereas the other two left it up to the doc essentially.
- one big issue here is that it might be difficult to find clear water between the groups. If group 2 was developed following international consensus then it will draw on the principles of group 1 Group 3 will likely be the same as group 2 (a doc doing usual care will hopefully be practising in a manner consistent with international standards) This failure to have a major difference between all 3 groups will make it difficult to find an outcome difference between the three.
- as a result I didn’t expect a big difference in outcomes as I don’t really think they were treated that differently.
- maybe if the population had been sicker you might be able to draw out a difference.
- when i deal with a septic patient I use a number of factors to guide my resus. This includes what happens to the lactate and even an ScvO2 if the line is in. Most docs who manage sepsis have some goal in mind that they’re targeting. Maybe all the goals are roughly equal, or if not equal then at least ‘good enough’ to get these outcomes