The primary purpose of this pilot study will be feasibility of recruitment. However, in order to further inform a full randomised controlled trial (RCT), information on the incidence of atrial fibrillation (AF) and other arrhythmias, hospital length of stay, resource use and morbidity will be collected.

Arrhythmias are common in critical care, with atrial tachyarrhythmias (and especially AF) being the most prevalent. This is especially true after cardiac surgery, with approximately 1 in 3 patients affected. The occurrence of new-onset post-operative AF is associated with increased short and long-term mortality, intensive care unit (ICU) and hospital stay and costs of care. This association appears causal, even after correction for confounding factors.

Potassium (K+) plays an important role in cardiac electrophysiology and abnormal levels may cause arrhythmias. Hypokalaemia, defined as a serum K+ <3.6 milliequivalents per litre (mEq/L) is thus associated with an increased incidence of ventricular arrhythmia after myocardial infarction. Low K+ levels are common following cardiac surgery, and appear marginally lower in those suffering atrial arrhythmias in this context. Despite an absence of proof that this association is causal, efforts to maintain serum [K+] in the 'high-normal' range (4.5 - 5.5 mEq/L) are considered 'routine practice' for AF prevention worldwide. The efficacy of such intervention remains unproven and data supporting this practice is extremely limited, being derived from observational rather than interventional studies. Indeed, no data exist to demonstrate that maintaining a high-normal potassium level is beneficial, or that aggressive replenishment of potassium in patients with heart disease necessarily leads to a better clinical outcome.

Furthermore, the method of potassium supplementation may be problematic. Oral replacement is not possible immediately post-operatively. Central venous administration is thus generally utilised in the early post-operative period. However, this practice is both time-consuming and costly: the intravenous administration of potassium carries recognised clinical risk, and is now prescribed in pre-diluted doses, stored securely for a safety purposes. Oral replacement is commonly associated with profound nausea, and is very poorly tolerated by patients. The investigators have estimated that the annual spending on potassium in cardiothoracic patients at Barts Health National Health Service (NHS) Trust is £100,000, compared to £16,500 for Milrinone (perceived as a high cost drug (2011-2012 prices). Additional costs relating to nursing time, drug checks, and intravenous connection and charting are also accrued. Central venous catheters may also be routinely left in situ solely for the purposes of parenteral potassium replacement; leading to an increased risk of line-related sepsis.

The routine maintenance of serum K+ in the high-normal range is thus a costly and unproven practice.

Inclusion Criteria:

1. All patients undergoing isolated elective coronary artery bypass graft (CABG)

Exclusion Criteria:

1. Age less than 18 years
2. Previous AF
3. Concurrent patient involvement in another clinical study assessing post-operative interventions
4. On-going infection/sepsis at the time of operation
5. Pre-op high-degree atrioventricular (AV) block
6. Pre-op serum K+ greater than 5.5 mEq/L
7. Current or previous use of medication for the purposes of cardiac rhythm management
8. Dialysis dependent end stage renal failure