Recommendation 1 – Beta-blockers to be continued through the operative procedure (2010)
We recommend that patients who have been receiving a beta-blocker before cardiac surgery have that therapy continued through the operative procedure in the absence of the development of a new contraindication (Strong Recommendation, High Quality Evidence). We suggest that patients who have not been receiving a beta-blocker before cardiac surgery have beta-blocker therapy initiated immediately after the operative procedure in the absence of a contraindication (Conditional Recommendation, Low Quality Evidence).

Values and preferences (2010)
These recommendations place a high value on reducing postoperative AF and a lower value on adverse hemodynamic effects of β-blockade during or after cardiac surgery. It is also noted that inherent to a strategy of prophylaxis, a number of patients will receive beta-blocker therapy without personal benefit.

Recommendation 2 – Amiodarone for patients with contraindications to beta-blockers (2010)
We recommend that patients who have a contraindication to beta-blocker therapy before or after cardiac surgery be considered for prophylactic therapy with amiodarone to prevent postoperative AF (Strong Recommendation, High Quality Evidence).

Values and preferences (2010)
This recommendation places a high value on minimizing the patient population exposed to the potential adverse effects of amiodarone and a lower value on data suggesting that amiodarone is more effective than beta-blockers for this purpose.

Recommendation 3 – Consider IV magnesium, colchicine, biatrial pacing when beta-blocker and amiodarone contraindicated (2016, updated from 2010)
We suggest that patients who have a contraindication to beta-blocker therapy and to amiodarone before or after cardiac surgery be considered for prophylactic therapy to prevent POAF with intravenous magnesium (Conditional Recommendation, Low Quality Evidence) or colchicine (Conditional Recommendation, Low Quality of Evidence) or with biatrial pacing (Conditional Recommendation, Low Quality of Evidence).

Values and preferences (2016, updated from 2010)
This recommendation places a high value on preventing POAF using novel therapies that are supported by lower-quality data; with a higher value on the lower probability of adverse effects from magnesium versus colchicine. The use of biatrial pacing needs to be individualized by patient and institution, as the potential for adverse effects may outweigh benefit based on local expertise.

Recommendation 4 – High risk and sotalol or combination prophylaxis (2010)
We suggest that patients at high risk of postoperative AF receive prophylactic therapy to prevent postoperative AF such as sotalol or combination therapy including ≥2 of a beta-blocker, amiodarone, intravenous magnesium, or biatrial pacing (Conditional Recommendation, Low- to Moderate Quality Evidence).

Values and preferences (2010)
This recommendation recognizes that data confirming the superiority of combinations of prophylactic therapies are sparse.

Recommendation 5 – Consideration of OAC for postoperative AF >72 hours (2010)
We suggest that consideration be given to anticoagulation therapy if postoperative continuous AF persists for >72 hours. This consideration will include individualized assessment of the risks of a thromboembolic event and the risk of postoperative bleeding (Conditional Recommendation, Low Quality Evidence).

Values and preferences (2010)
This recommendation places a higher value on minimizing the risk of thromboembolic events and a lower value on the potential for postoperative bleeding. Because the risk of postoperative bleeding decreases with time, the benefit-to-risk ratio favours a longer period without anticoagulation in the postoperative setting than that suggested in other settings.

Recommendation 6 – Temporary epicardial pacing electrode wires at surgery (2010)
We recommend that temporary ventricular epicardial pacing electrode wires be placed at the time of cardiac surgery to allow for backup ventricular pacing as necessary (Strong Recommendation, Low Quality Evidence).

Values and preferences (2010)
This recommendation reflects the relative ease of placement of epicardial temporary pacing wires at the time of surgery as well as the potential for significant morbidity associated with postoperative bradycardia.

Recommendation 7 – Post-op AF with rapid response: beta-blocker, CCB, or amiodarone (2010)
We recommend that postoperative AF with a rapid ventricular response be treated with a beta-blocker, a non–dihydropyridine calcium antagonist, or amiodarone to establish ventricular rate control. In the absence of a specific contraindication, the order of choice is as listed (Strong Recommendation, High Quality Evidence).

Values and preferences (2010)
This recommendation places a high value on the randomized controlled trials investigating rate control as an alternative to rhythm control for AF, recognizing that these trials did not specifically address the postoperative period.

Recommendation 8 - Rate-control or rhythm-control strategy for post-op AF (2016, updated from 2010)
We recommend that postoperative AF may be appropriately treated with either a ventricular response rate-control strategy or a rhythm-control strategy (Strong Recommendation, Moderate Quality Evidence).

Values and preferences (2016, updated from 2010)
This recommendation places a high value on the randomized controlled trials investigating rate control as an alternative to rhythm control for AF, including one trial specifically addressing the cardiac postoperative period. Choice of strategy should therefore be individualized based on the degree of symptoms experienced by the patient.

Recommendation 9 – Reconsideration of ongoing therapy 6-12 weeks post-op (2010)
We recommend that, when anticoagulation therapy, rate-control therapy, and/or rhythm control therapy has been prescribed for postoperative AF, formal reconsideration of the ongoing need for such therapy should be undertaken 6-12 weeks later (Strong Recommendation, Moderate Quality Evidence).

Values and preferences (2010)
This recommendation reflects the high probability that postoperative AF will be a self-limiting process that does not require long-term therapy.