Clinical Question

To investigate the efficacy and safety of early compared with delayed initiation of direct oral anticoagulants (DOACs) in patients with acute ischaemic stroke associated with atrial fibrillation.

Inclusion Criteria

• Adults (≥18 years) with atrial fibrillation confirmed by ECG or medical records.
• Clinical diagnosis of acute ischaemic stroke with symptoms lasting >24 hours.
• Brain imaging (CT or MRI) to exclude intracranial haemorrhage and non-stroke diagnoses.
• Eligibility for anticoagulation with a DOAC, with the treating physician uncertain of the optimal timing to start.
• Ability to be followed up for 90 days after trial entry.

Exclusion Criteria

• Coagulopathy.
• Evidence of recent or current anticoagulation with a vitamin K antagonist leading to an INR of 1.7 or higher at randomisation.
• Clinically significant thrombocytopenia (platelet count <75×10^9 platelets per L).
• Other coagulopathy or bleeding tendency judged to contraindicate anticoagulation.
• Severe haemorrhagic transformation of the acute infarct (parenchymal haematoma type 2).
• Acute intracranial haemorrhage unrelated to the acute infarct.
• Contraindication to DOAC use (e.g., severe renal impairment [creatinine clearance <15 mL/min], cirrhosis [Child Pugh classification B or C], alanine aminotransferase >2x upper limit of normal, or concurrent medication with a notable DOAC interaction).
• Known allergy or intolerance to Factor Xa and direct thrombin inhibitor.
• Definite indication for use of a vitamin K antagonist (e.g., mechanical heart valve).
• Pregnant or breastfeeding.
• Brain imaging evidence of non-stroke pathology judged likely to explain clinical presentation.
• Did not agree to provide consent to study procedures.
• Any other contraindication to early anticoagulation as judged by the treating clinician.
• Any other reason that the treating clinician considered would make the patient unsuitable to enter OPTIMAS.

Baseline Characteristics

Arms

Outcomes

Criticisms

• Open-label design, which means participants and treating clinicians were not masked to treatment assignment, potentially introducing performance bias (though outcome adjudication was blinded).
• The study did not randomly assign participants to start anticoagulation between 4 and 7 days after onset, limiting the ability to observe optimal timing within this specific early period.
• The low number of symptomatic intracranial haemorrhage events limited statistical power for this specific safety outcome.
• The 95% CI for the primary outcome, while meeting non-inferiority, includes a maximum adjusted risk difference of 1.2 percentage points, which might be considered clinically meaningful by some.
• The trial excluded patients with severe parenchymal haematoma type 2, limiting definitive guidance for this rare, high-risk subgroup.
• The study did not centrally adjudicate cardiovascular mortality data.
• Brain imaging analyses (e.g., infarct volume, haemorrhagic transformation subtypes) will be reported separately, meaning full imaging-related insights are not yet available in this publication.

Funding

British Heart Foundation.