INTRODUCTIONCardiac implantable electronic devices (CIEDs) are becoming more and more common due to an aging global population and expanded criteria for implantation. During 2009, over 1.3 million new or replaced implantable cardiac devices were implanted worldwide, with around a quarter of these occurring in the United States (US) alone.(1,2) South Africa has experienced substantial growth with an almost 25% rise in CIEDs implanted between 2005 and 2009, as opposed to only 8% in the US.(1) In addition, there is a greater availability of external surface monitoring and implantable loop recorders (ILR). Out of this vast increase in continuous arrhythmia monitoring arises the challenge of how to approach the abundance of data that identifies subclinical arrhythmia, and germane to this discussion, atrial fibrillation (AF).As the most pervasive sustained arrhythmia encountered in clinical practice, AF has risen in age-adjusted incidence over the last half century -a trend which may be plateauing over the last decade.(3,4) After age 40 the lifetime risk of developing AF is almost 1 in 4.(2) As many as 40% of patients are entirely asymptomatic -also termed subclinical AF -and the arrhythmia may only come to the attention of the patient and provider as an incidental finding.(5) Population screening of those over 65 years of age would detect subclinical AF in an estimated 1.4% of patients -of whom, more than two thirds would be at high risk for stroke based on clinical risk prediction models.(6) AF shares many risk factors -including age, male gender, heart failure and coronary disease -with indications for a permanent pacemaker or implantable cardioverter defibrillator (ICD) and, as such, the detection of clinically silent paroxysms of AF by device interrogation is not uncommon. In patients with CIEDs placed for unrelated indications, without a prior history of permanent AF, 43% had 5 minutes or more of AF detected over 2 years of follow-up in a pooled analysis of 5 large prospective trials (n=10 016).(7) Therefore, adoption of a strategy for dealing with device-detected subclinical AF is vital for clinicians.In the midst of several studies which demonstrate the con-