Objective: To evaluate anticoagulation practices in patients with atrial fibrillation (AF) admitted for acute stroke in the neurology department of an emergency hospital.
Methods: From the 595 adult subjects (age range 28-104 years) admitted in 2016 for stroke, we selected 159 patients with previously diagnosed AF, and looked for predictors of stroke recurrence.
Results: AF had been diagnosed in 25% of first-stroke patients (n=118) and in 33% of those with recurrent stroke (n=41). As much as 69% of first stroke and 61% of recurrent stroke patients were not on oral anticoagu-lant therapy. In the first-stroke subgroup, all AF-nor-motensives (n=20) had a pre-stroke CHA2DS2-Vasc score ≥1, none were on non-antivitamin K oral anti-coagulants (NOACs), and 35% were on antivitamin K (AVK), having an admission INR<2 in most cases. First-stroke AF-hypertensives (n=98) were anticoagulated in only 31% of cases, with either NOACs (3 pa-tients), or AVKs (27 patients), the latter mostly having an admission INR <2. Among recurrent-stroke AF patients, NOACs were used in 4 patients and AVKs in 12 patients, these latter generally having an admission INR<2. On logistic regression neither CV risk factors nor comorbidities were significantly associated with recurrent stroke. The only significant association was seen with the CHA2DS2Vasc score (actually a composite of the above). In-hospital mortality was higher in AF with stroke recurrence versus first-stroke – 37% (15 patients) versus 2% (29 patients) (p .158) and higher among non-anticoagulated AF patients (31% vs. 24%, p=0.162), while no deaths occurred in NOAC users. From the initial AF cohort, there were 115 survivors at discharge. Among these, 23 patients (20%) remained naïve to anticoagulation, 50 patients (43%) were star-ted anticoagulation with either NOACs (29 patients) or AVKs (21 patients). In 11 previously anticoagulated patients (10%) this therapy was stopped. In AF pati-ents previously on AVK and with admission INR<2, the switch to NOACs was recommended at discharge in 10% of cases (2 out of 20 patients).
Conclusions: In more than half of this cohort of AF-stroke patients, prevention of embolism with anticoagulation was warranted by the CHA₂DS₂-VASc score decision algorithm, preferably with a non-AVK, but was non-optimal. The smaller percentage of AVK users were unprotected due to sub-therapeutic use, again underlining the importance of guideline recommended first-line non-AVK option for anticoagulation therapy. The practice was slightly improved after stroke occurrence. Mortality was higher in recurrent stroke, but comparison did not reach significance probably due to small cohort size