AF in sepsis

Dear all, this weeks question(s).

It is not uncommon for septic patients on the unit to develop atrial fibrillation.

-In those circumstances which do not require immediate DCCV but treatment is thought to be necessary to improve cardiovascular parameters, what is the most appropriate strategy?

-Is this approach consistent for all patient cohorts?

-Should the arrhythmia persist, when (if at all) should anticoagulation be considered?


It’s a bit of an evidence black hole. 

I wrote one of my ECS on this topic (not claiming this is a definitive opinion!). We also now have an SOP for this (

B –

Walkley et al did quite a nice, easy to read review in Chest (2015).

In summary:

– There are no real major trials in the area, and pretty much all evidence is single-centre small stuff, expert opinion, or extrapolated from the non-critically ill.

– None of the antiarrhythmics have been proven to be conclusively superior, IV Diltiazem was superior to IV Amiodarone in achieving rate control (rate reduction of 30% or <120), at the expense of more hypotension, with 30% of the Diltiazem group requiring the drug to be stopped prematurely. In another small study (n=50) Amiodarone infusion had a higher rate of reversion to sinus rhythm than Digoxin at 24 hours (92 v 71%, p=0.048) with a shorter median time to reversion (2.5 v 6.8 hours). It is worth noting that this trial was in a coronary ICU, but half of the patients had class III or class IV failure and none appeared to have systemic sepsis which limits its generalisability somewhat. IV Magnesium was shown to be superior to Amiodarone in a small Australasian study (n=42) of critically patients with respect to conversion to sinus rhythm.

– The authors of the review put a lot of faith in a small study that showed that an IV Esmolol infusion was associated with improved mortality in patients with septic shock, and recommended Esmolol as their first-line agent for rate control. I believe this study’s findings should be interpreted with caution; given its single centre, primary outcome of whether Esmolol lowered the heart rate (mortality was a secondary outcome) and high mortality in both arms of the study (Esmolol 49.4% v 80.5% in the control arm). I’m not sure I would agree that Esmolol is so great without more evidence to support it.

– The authors fairly uncontroversially recommended correcting reversible causes for AF with rate control recommended over rhythm control where patients aren’t compromised. DCCV appears to have fairly poor success in Sepsis-induced AF, but is recommended (mostly expert opinion) where patients have significant compromise.

– In terms of anticoagulation, there is a demonstrable increase in stroke risk with AF+Sepsis (0.2%/day) but septic patients can be coagulopathic so the authors decided to sit on the fence and recommend an individual risk-benefit analysis. Which seems fairly reasonable for most things.

– One thing that does seem worth changing in practice is following up patients who developed AF in Sepsis as recurrence appears relatively common and may well be asymptomatic. It seems that much like CKD patients being followed up a nephrologist, AF follow up by a cardiologist may reveal a number with paroxysmal AF who would potentially benefit from anticoagulation in the longer term.

– I think one thing that also should be considered is whether we are also iatrogenically inducing AF in some patients with noradrenaline. The SEPSISPAM trial, although conducted with a degree of gallic laissez-faire did show that more patients developed AF in the high MAP target (65-70mmHg v 80-85mmHg). Although in reality the study was more like 70-75mmHg v 85-90mmHg due to protocol violations, receiving more Noradrenaline didn’t improve your survival and did increase your chance of AF (although it reduced your risk of requiring RRT if you had CKD).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s