Objectives To assess whether daily short-term rhythm strip recordings on top of routine clinical care could increase the atrial fibrillation (AF) detection rate in the hospitalised elderly.

Methods A hand-held device storing a bipolar ECG during 1 min was used for daily rhythm recording in hospitalised elderly patients.

Results During 2 months, all patients admitted to the Department of Geriatric Medicine were screened (n=327). Five patients refused to participate in the study and 70 patients were unable to hold the device due to severe mental (n=46) or motor impairment (n=24). In the remaining 252 patients, 1582 recordings were successfully obtained after 1624 attempts with a median acquisition time of 1 min (min 1, max 9, IQR 1–2 min). The rhythm strips were not reliable interpretable due to artefacts in three patients or an implantable cardiac pulse generator in another 28 patients. Detailed clinical information was available in 214/221 patients. Mean age was 84±6 years. On top of 71 (33%) patients with AF identified by routine clinical care (history, n=64 or de novo detected during current hospitalisation, n=7), review of all rhythm strips identified another 28 patients (13%) with AF. All these patients had a CHA₂DS₂VASc score ≥2. A contraindication for anticoagulation was present in only 8/28 (25%) of identified patients.

Conclusions On top of routine clinical care, daily short-term rhythm strip recordings identified another 13% of elderly hospitalised patients with AF, leading to an overall prevalence of 46% in hospitalised patients. This can have significant therapeutic implications with respect to initiation of anticoagulation.