LOS ANGELES—Atrial fibrillation is three times more likely to be detected within six months of an ischemic stroke if, instead of 24 hours of Holter ECG monitoring, patients are monitored for 10 days post-stroke, and then again for 10 days at three and six months, according to an investigation presented at the International Stroke Conference 2016.
“Enhanced and prolonged monitoring should be considered for all stroke patients to improve the detection of atrial fibrillation,” said investigator Rolf Wachter, MD, a cardiologist at the University of Göttingen in Germany.
Patients in the trial had no history of atrial fibrillation and were age 60 or older. A total of 200 patients were randomized to the extended-monitoring protocol, and 198 were randomized to standard of care, which included a median of 24 hours of Holter monitoring. The median time from symptom onset to randomization was three days. All patients were enrolled by day five.
The study wasn’t limited to cryptogenic strokes. The first stroke may not have been related to atrial fibrillation, but could have been a recurrent stroke. “Our approach was to look for atrial fibrillation in all stroke patients, irrespective of etiology,” Dr. Wachter said.
At six months, atrial fibrillation had been detected in 27 patients (13.5%) in the extended-monitoring group, versus nine (4.5%) in the control group. The findings were largely the same when results were analyzed by age, sex, NIH Stroke Scale scores, and other criteria. Atrial fibrillation was detected in the majority of both groups within a month. Dr. Wachter did not elaborate on the frequency and duration of atrial fibrillation, but inclusion criteria required at least one episode lasting 30 seconds or longer.
The research raises the questions of how aggressive post-stroke atrial fibrillation monitoring should be, and when treatment should start.
Every patient with atrial fibrillation in the study received an oral anticoagulant, and almost all remained on their anticoagulant at one year. The oral anticoagulants used were not reported, and the study wasn’t powered to detect differences in clinical outcomes. Even so, “we saw very positive trends in the right direction” for prolonged monitoring, said Dr. Wachter.
Five patients in the intervention arm (2.5%) had a recurrent stroke, and three (1.5%) had transient ischemic attacks at one year, versus nine (4.5%) recurrent strokes and five (2.5%) transient ischemic attacks in the control arm. Six patients in the intervention arm (3%) and nine (4.5%) in the control group had died.
A total of 60% of the subjects were men, and the mean age in the study was 73. The mean NIH Stroke Scale score was 3 in the intervention group and 2 in the control group. The mean CHADS2 was 3.5 in both groups. Patients in the extended-monitoring group wore their Holter monitors for a median of 9.5 days at all three time points, but compliance decreased from 100% to about 75% by the third session.
—M. Alexander Otto