Background and importance In a multidisciplinary hospital with 426 beds, anticoagulant treatments have a high risk of iatrogenism and prescription error. We decided to focus on the direct oral anticoagulant (DOAC) treatments.

Aim and objectives We analysed prescriptions to evaluate the rate of correct prescription. We wanted to assess the level of patient knowledge and impact of a pharmaceutical interview (PI) on this degree of knowledge.

Material and methods A prospective study including 38 patients from 1 July 2020 to 31 August 2020 was conducted. First, we evaluated the relevance of dosage of DOAC during pharmaceutical analysis. Then, patients’ knowledge of DOACs was assessed by a questionnaire before and after PI. This questionnaire comprised nine items concerning: general notions about DOAC, drug administration, over- and under-dosing and drug interactions. A statistical test was performed to compare the data (α = 2.5%).

Results The mean age of the 38 included patients was 83 years and the sex ratio was 1. Patients received apixaban (53%), rivaroxaban (45%) and dabigatran (2%). DOAC were used to prevent stroke in adult patients with non-valvular atrial fibrillation for 95% of them. Data were lacking to allow a correct pharmaceutical analysis: patient weight was not indicated in the patient file in 90% of prescriptions of apixaban. A dosage error was noted in 9 prescriptions and 4 prescriptions were changed following pharmaceutical intervention. Knowledge assessment was carried out in 31 patients. For 7 patients, communication difficulties, cognitive and psychiatric disorders made this assessment impossible. Therapeutic knowledge before and after PI was 52% and 67%, respectively. We observed a statistically significant improvement in patients’ knowledge of their DOAC treatment concerning general notions about the drug (+10%), administration (+24%), over- and under-dosage (+13%) and drug interactions (+23%).

Conclusion and relevance This study reveals patients’ poor knowledge of their DOAC treatment. However, performing PI statistically improves patient knowledge. It would therefore be interesting to systematically carry out these PI. It would also be interesting to develop a city–hospital link in conjunction with pharmacists for optimised patient follow-up.

Conflict of interest No conflict of interest