A 76-year-old man with hypertension and sinus node dysfunction presents to your office for his first pacemaker check two months after device implantation. The interrogation reveals that he had one mode switch episode for atrial fibrillation (AF) that lasted ten minutes. The patient has no prior history of AF and denied having any symptoms on the day of the episode. He has no contraindication to anticoagulation. Do you start him on oral anticoagulation?
The current recommendations regarding the need for anticoagulation in patients with AF are based on the presence or absence of non-AF risk factors for stroke, such as those included in the CHADS2 scoring system. Patients with paroxysmal AF and risk factors are advised to start oral anticoagulation in the absence of a contraindication, regardless of the frequency or duration of AF. However, given the risks of oral anticoagulation, it is always difficult to know whether or not to start such a patient on an anticoagulant when they are found to have a very brief episode of asymptomatic AF during a pacemaker check. This clinical decision commonly arises in the device clinic. Should a single episode of AF found on a pacemaker check in a patient with other risk factors for stroke give that patient a lifetime indication for oral anticoagulation? Shouldn’t the burden of AF correlate with the risk of stroke?
In the January 2012 issue of the New England Journal of Medicine, the ASSERT Investigators published the results of their study designed to evaluate whether subclinical episodes of AF detected by a pacemaker were associated with an increased risk of stroke in patients who had no prior documented AF.1 They monitored 2,580 patients over age 65 with hypertension and a pacemaker, but no history of AF, for episodes of AF on their pacemaker. Atrial fibrillation was defined as any atrial arrhythmia with an atrial rate over 190 beats per minute for more than six minutes. Ten percent of patients had subclinical AF detected by their device in the first three months, and these patients were 2.5 times more likely to have a thromboembolic event. Furthermore, subclinical AF was an independent predictor of stroke.
Based on the ASSERT data, should the patient described above, with a CHADS2 score of 2 and a single episode of AF lasting more than six minutes noted at the time of his pacemaker check, be prescribed oral anticoagulation? A more careful look at the ASSERT data suggests that using a six-minute duration cutoff to label a patient as being at high risk for stroke ignores the results of additional analyses performed by the investigators — data that is unfortunately not mentioned in the abstract. When patients with episodes of AF were stratified by the duration of the episode, only those in the highest quartile, those with episodes that lasted more than 18 hours, had a risk of stroke that was statistically significant. In contrast, the frequency of episodes did not seem to be important.
It is always difficult to make a yes-or-no decision regarding the need for an intervention based on a specific cut-off point when using a risk assessment formula. However, based on the ASSERT trial, it seems reasonable to consider anticoagulation in patients with subclinical AF detected at the time of a pacemaker interrogation if it lasted more than 18 hours. The harder decision at this time might actually be which drug to start.
- Healey JS, Connolly SJ, Gold MR, et al. Subclinical atrial fibrillation and the risk of stroke. N Engl J Med 2012;366:120–129.