Demographic and clinical profile of cardioembolic stroke patients in Western Sydney

Abstract

Background Cardioembolism (CE) contributes to a large proportion of ischaemic stroke. Aims To evaluate the demographic and clinical profile of CE stroke in Western Sydney. Methods A retrospective analysis of ischaemic stroke patients presenting to Westmead Hospital (January–October 2016) was performed. Strokes were classified by TOAST (Trial of Org 10172 in Acute Stroke Treatment) criteria into different categories. Clinical and demographic data were collected on all stroke patients, and differences between CE and other stroke causes were identified. Results Two hundred and twenty‐eight consecutive patients (70.9 years; 53% male) were identified. By TOAST criteria, 21 (9%) had large‐artery atherosclerosis, 94 (41%) CE, 10 (5%) small‐vessel disease, 2 (1%) other aetiology and 101 (44%) undetermined aetiology. A significant proportion of CE stroke patients had cardiovascular risk factors including hypertension (66%), hypercholesterolaemia (50%), diabetes (26%) and ischaemic heart disease (28%). The majority (81%) of patients with CE had atrial flutter/flutter. CE stroke, compared with other types of stroke, was more common in females (56 vs 41%, P = 0.022) and patients with CE stroke were more likely to have previous cerebral ischaemia (34 vs 21%, P = 0.026), suggesting increased recurrence in this group. Of the patients with atrial flutter/flutter (n = 56), the majority (87%) had a high CHA2DS2‐VASC score (≥2); however, a significant proportion (55.4%) were not on anticoagulation. Conclusions Cardioembolic stroke remains a significant burden in Western Sydney, and it is likely that a significant proportion may be preventable, as evidenced by the substantial presence of modifiable cardiovascular risk factors, and inadequate anticoagulation of patients with atrial arrhythmias.

Direct oral anticoagulant use in hospitalized patients with atrial fibrillation across body mass index categories: design and rationale for a retrospective cohort study

“…As eMR data represent routinely collected clinical data, there is a level of uncertainty to consider in sample size calculations. Based on the sample size of previous studies 36 – 40 that have explored a similar topic within the same health district, a sample size of approximately 300 patients is anticipated. The Pearson product–moment correlation coefficient will be used to examine the relationship between the rate of rehospitalization and BMI.…”