Holter monitoring in the diagnosis of stroke mechanism

BACKGROUND: Atrial fibrillation (AF), an important and treatable cause of ischaemic stroke, can occur as a sustained or a paroxysmal arrhythmia. Continuous cardiac rhythm monitoring (Holter monitoring) is often performed in stroke patients to identify paroxysmal AF, which is an indication for warfarin anti-coagulation in this patient population. AIM: The aim of this study was to assess the clinical utility of Holter monitoring in detecting occult AF in patients with possible cardioembolic stroke. METHODS: The medical records of ischaemic stroke patients consecutively hospitalized at a single academic centre during a one-year period were reviewed. Data regarding patient demographics, stroke characteristics, electrocardiography and echocardiography results and duration and findings of Holter monitoring were abstracted. The primary outcome was yield of newly diagnosed AF on Holter monitoring. RESULTS: Of 465 consecutive patients admitted with a diagnosis of new ischaemic stroke, 210 underwent Holter monitoring. The mean duration of monitoring was 22.8 +/- 4.0 h. Previously undiscovered AF was -identified in five cases (2.4%), all of which represented non-rheumatic AF. In three cases, the Holter test was negative despite AF documented on an admission electro-cardiogram. CONCLUSIONS: Holter monitoring can identify occult paroxysmal AF, assisting targeted secondary prevention in patients with new ischaemic stroke. However, the standard 24-h duration of monitoring probably under-estimates the prevalence of paroxysmal AF in this population. Prospective studies are indicated to evaluate the value of longer monitoring periods in stroke populations.