Inconsistent centralised versus non-centralised ischaemic stroke aetiology.

Area of Science:

• Neurology
• Epidemiology
• Clinical Research

Background:

• The Trial of Org 10 172 in Acute Stroke Treatment (TOAST) system is the standard for classifying ischemic stroke aetiology.
• Limited studies have validated the accuracy of routine stroke aetiological diagnosis against the TOAST criteria.

Purpose of the Study:

• To investigate the agreement between centralised and non-centralised (site-reported) stroke subtypes in the Third China National Stroke Registry (CNSR-III).
• To analyse the influence of classification consistency on hospitalisation evaluation and secondary prevention strategies.

Main Methods:

• Included 12,180 ischaemic stroke patients from the CNSR-III with complete diffusion-weighted imaging data.
• Utilised multivariable Cox proportional-hazard regression models to identify factors associated with subtype classification consistency.
• Conducted sensitivity analyses on subgroups with complete information.

Main Results:

• Agreement between centralised and non-centralised subtypes varied: large-artery atherosclerosis (77.4%), small-vessel occlusion (40.6%), cardioembolism (38.7%), and other determined aetiology (12.2%).
• Patient- and hospital-level factors influenced classification inconsistency.
• Inconsistency impacted secondary prevention; only 15.3% of newly diagnosed cardioembolism patients received oral anticoagulants post-centralised subtyping.

Conclusions:

• Significant inconsistency exists between centralised and non-centralised stroke subtyping in China.
• Inaccurate aetiological subtyping can result in inadequate secondary prevention, especially for cardioembolic stroke.
• Standardising classification is essential for optimising patient care and secondary prevention strategies.