Gillinov 2016

Screen Shot 2018-08-01 at 12.50.05.png

Rate versus rhythm control for atrial fibrillation after cardiac surgery

Included 523 patients with new postoperative atrial fibrillation (POAF) without hemodynamic compromise after cardiac surgery
POAF was defined as having atrial fibrillation lasting for at least 60 minutes during the first seven days after CABG or valve surgery
Randomized patients to rate control (using medications to achieve a HR <
100 bpm) or rhythm control (using primarily oral amiodarone with or without rate controlling medications).
Among those randomized to rhythm control, direct current cardioversion (DCC) was recommended after 24 to 48 hours if sinus rhythm was not achieved. Additionally, those randomized to rhythm control were
continued on oral amiodarone for up to 60 days.
Cross-over from rate to rhythm control was allowed if the provider felt that conversion to sinus rhythm would improve hemodynamics or alleviate symptoms. This cross-over occurred in 26.7% of rate control patients.
Primary endpoint (total days in hospital within the first 60 days of randomization) was not different between rate and rhythm control (6.4 vs. 7.0 days, p=0.76). The trial met its intended power to detect a difference of 2 days.
There was no statistical difference in the incidence of atrial fibrillation at hospital discharge between rate and rhythm groups (89.9% vs. 93.5%, p=0.14). At 60 days, rate control patients were less likely to have been free of atrial fibrillation within the previous 30 days (93.8% vs. 97.9%, p=O.02, NNT 24)
There were no differences observed between the two groups regarding hospital readmission rates, receipt of anticoagulation (warfarin) at discharge, duration of anticoagulation (about 45 days), mortality, ischemic stroke, bleeding, or serious adverse events
Among rate control patients, the most common reason for protocol non- adherence was due to ineffective heart rate control. Among rhythm control patients, the most common reason to discontinue therapy (amiodarone) was due to side effects of the drug itself.
Unfortunately, the study does not describe the type or severity of amiodarone side effects requiring drug discontinuation.
In those randomized to rhythm control, the typical amiodarone regimen was 400 mg PO TID x 3 days, then at least 200 mg daily for 60 days. Patients unable to take PO were allowed to receive IV therapy.
For both prevention and treatment of POAF, there is considerable practice variation regarding route (IV vs. PO), dose, and duration of loading dose before changing to a daily maintenance dose.
It is not known whether a different amiodarone strategy would have produced different results in this trial.
All POAF preventative measures were left to the discretion of each site. Due
to wide variation in preventative measures (including pre-op
amiodarone), it isn’t clear whether the efficacy of rate or rhythm control could be impacted by POAF preventative measures.
Like in the AFFIRM trial (a similar study question for non-surgical
outpatient atrial fibrillation), cross-over from rate to rhythm was fairly common with about 25% of patients assigned to rate control receiving amiodarone or DCC.
Although this fact reduces the internal validity of the trial, it does improve the external validity in that this is commonly done in clinical practice.
The selection of hospital duration is interesting because it was intended to be a surrogate for complications related to POAF. Unfortunately, length of stay endpoints can be difficult because many factors (unrelated to the study question) can make length of stay highly variable