The purpose of this study is to determine whether a strategy of maintaining serum potassium levels at ≥3.6 mEq/L is non-inferior to a strategy of usual treatment (≥4.5 mEq/L) on the occurrence of new onset atrial fibrillation after cardiac surgery (AFACS) in patients undergoing isolated coronary artery bypass graft (CABG) surgery.

At least one in three patients is affected by atrial fibrillation after cardiac surgery(AFACS), with most episodes occurring in the first five postoperative days. AF occurrence is associated with increased morbidity, short and long-term mortality, intensive care unit (ICU) and hospital stay and cost of care. Persistence of these associations after adjustment for potential confounding factors suggests that they may be causal. The incidence and prevalence of AF and its associated costs are expected to increase as the surgical population ages. Extensive effort is undertaken to prevent AF after cardiac surgery from occurring, but clinical practice in this area is highly variable and the evidence base for most interventions is sparse.

Potassium plays an important role in cardiac electrophysiology. Serum potassium concentrations ([K+]) are commonly low following cardiac surgery, and appear marginally lower in those suffering atrial arrhythmias in non-surgical cohorts. Despite an absence of proof that this association is causal, efforts to maintain serum [K+] in the 'high-normal' range (≥ 4.5 mEq/L), as opposed to just intervening if potassium drops below its lower 'normal' threshold (<3.6 mEq/L), are considered 'routine practice' for AF prevention in post-surgical patients in many centres across the world.

From the (unpublished) data from our Tight-K Feasibility Study, all 160 patients would have required at least one dose of potassium to supplement their levels to this high-normal range and 45.5% of all serum [K+] measurements were below 4.5mEq/L at some point. Data from the same pilot study show a median number of potassium doses given in the 'tight' group (high-normal serum potassium target) of seven, compared to a median of one, with most patients not receiving any potassium supplementation at all, in the 'relaxed' group. This demonstrated for the first time ever, that the practice does achieve a separation in serum potassium levels between the two groups, so the protocol is indeed effective in achieving higher serum potassium levels.

The efficacy of the practice of maintaining high-normal serum potassium levels for the prevention of AF after cardiac surgery, however, remains unproven and data supporting it are extremely limited, being derived from observational studies rather than randomised trials. Indeed, no data exist to demonstrate that maintaining a high-normal potassium level is beneficial in these circumstances, or that aggressive replenishment of potassium in these patients improves outcome.

Meanwhile, potassium supplementation may cause discomfort or harm. Routine central venous potassium administration in the early post-operative period, when oral supplementation is not possible, is time-consuming, costly and associated with clinical risk: rapid infusion can prove fatal, and leaving central venous catheters in situ for the sole purpose of potassium replacement increases infection risk. Oral replacement (when feasible) is commonly associated with profound nausea and gastrointestinal side effects, and is very poorly tolerated by patients. The annual costs of intravenous potassium exceed those for other drugs in many cardiac surgical units due to the large quantities administered. Nursing time (e.g. for drug checks and administration) will add to this cost.

The routine maintenance of serum [K+] ≥ 4.5 mEq/L is a costly practice of unproven efficacy that is unpleasant and may be hazardous for patients. The investigators shall address this issue, performing the first appropriately powered non-inferiority multicentre randomised trial of potassium supplementation in patients undergoing coronary artery bypass surgery.

Inclusion Criteria:

• Scheduled to have isolated CABG surgery
• Patient in sinus rhythm

Exclusion Criteria:

• Age less than 18 years
• Previous history of Atrial Fibrillation, Atrial Flutter and/or Atrial Tachyarrhythmia
• Pre-operative high-degree atrioventricular (AV) block (defined as Mobitz type 2 second degree AV block or complete heart block)
• Pre-operative serum [K+] greater than 5.5 mEq/L
• Current/previous use of medication for the purposes of cardiac rhythm management
• Dialysis-dependent end-stage renal failure
• Concurrent patient involvement in another clinical trial assessing cardiac rhythm post-operative interventions
• Unable to give informed consent