Screening for atrial fibrillation with or without general practice involvement: a controlled study

There has been a drive to increase atrial fibrillation (AF) detection in general practice. However, one-off, opportunistic testing can miss paroxysmal AF and requires significant resource. Paroxysmal AF can be detected through screening that involves repeated ECGs over a period of time, although it is unclear whether screening would need to be led by general practice, and how much support participants require. We aimed to investigate whether AF screening can be delivered remotely by a centralised administration instead of general practice, and to determine the level of support required.

We undertook a controlled comparator study with secondary randomisation in three English general practices. We invited people aged ≥ 70 years to use a hand-held ECG device four times daily for three weeks. Participants were allocated to practice-led or administrator-led screening, with administrator-led support randomised to three different levels. We compared quantity and quality of ECGs obtained in each arm. The primary outcome was proportion of screened participants who recorded ≥ 56 adequate-quality ECGs (2/3 of possible ECGs).

Of 288 screened participants, 59 participants received practice-led screening with a telephone consultation to explain the device. The remainder received administrator-led screening: 81 were automatically given a consultation; 74 were offered a consultation, and 74 were not offered a consultation. Most screened participants (280/288, 97.2%) recorded ≥ 56 adequate-quality ECGs. This proportion did not vary significantly between practice-led and administrator-led screening (100.0% vs. 98.8%), or between support levels (94.6% to 98.8%). Practice-led screening led to slightly more adequate-quality ECGs (mean: 83.9 vs 78.3, p < 0.001).

AF screening can be successfully delivered remotely, outside general practice, with minimal support.