A total of 25% to 40% of ischemic strokes are considered as cryptogenic stroke (CS) because etiology remain unclear, potentially because of misdiagnosis of Atrial Fibrillation (AF). Optimal monitoring duration to detect occult AF is debated but prospective data have demonstrated the benefits of long-term Holter Monitoring (HM).

The aim of our study was to compare, in a real life study, 7 days Holter Monitoring (7DH) to 3 days Holter Monitoring (3DH) after CS, firstly in term of AF detection performance but also in term efficiency.

We performed a retrospective pilot study including all consecutives patients who had 7DH, performed in our university centre, after CS. The primary end point was the presence of AF episode lasting at least 30seconds. When AF was detected, we newly interpreted the 7DH to consider if this AF episode happened during the first 3 days of monitoring, or later. Thus each patient was compared to himself and the conditions of the exam were similar between 7DH and 3DH.

From the 01/08/2015 to 01/03/2018, 217 patients with no history of AF had a 7DH for CS. Mean population age was 67±10 years old, with a median CHADSVASc score of 5 (4,6) and 22% of patients had a dilated left atrium. AF was detected in 13 patients (5.9%), among them 11 were detected in the first 72hours (5%) (P=0.157), which corresponded to one extra AF diagnosis every 108 HM. In term of efficiency, according to the availability, mean 7DH delay was 182±380 days, which would be shorter using 3DH without additional recorders. HM reimbursement price by the Public French medical insurance is fixed at 77 euros unrelated to the monitoring duration. Finally, time consuming for the practitioner were lower according to the 3DH.

This pilot study showed a very low incremental benefit of AF detection performance with a worse efficiency, considering the 7DH compare to a 3DH. However further investigations are necessary to achieve definitive conclusions.