Background and importance Increasing numbers of patients receiving a novel oral anticoagulant (NOAC) are undergoing elective surgery. The extent to which perioperative interruption of NOAC therapy is concordant with best evidence is uncertain.

Aim and objectives This study investigated whether inappropriate perioperative advice can lead to the occurrence of an ischaemic stroke. Furthermore, we examined the relation between inappropriate dosing, perioperative management and interactions.

Material and methods Data from all ischaemic stroke patients, previously treated with a NOAC, were retrospectively collected from the EVAS-BE-database (January to October 2019). The following data were retrieved: date of stroke, aetiology, previous stroke, posology and indication for NOAC, renal function, weight, age, concomitant drugs, surgery (indication, date, bleeding risk, preoperative advice), medication management post-stroke and discharge therapy. Concordance of perioperative anticoagulation management with regional and EHRA guidelines was rated by a clinical pharmacist according to the explicit risk of thrombosis and bleeding.

Results Of the 57 patients with an ischaemic stroke receiving a NOAC, nine (16%) had been planned to undergo surgery. The decision to interrupt anticoagulation was concordant with regional guidelines. Compared with EHRA guidelines: three cases stopped without indication (5 days–2 days), and three low and one high bleeding risk patient stopped too early. None of them were bridged.

Firstly, inappropriate dosing (30%) and posology (7%) based on the SmPC criteria was identified. Of the 17 inappropriately dosed patients, underdosing was the main factor (16 vs 1). Secondly, 16 patients (28%) showed one or more interactions with concomitant drugs. Due to the pharmacodynamic interactions, a higher risk of thrombosis was seen in two patients. Four patients showed a pharmacokinetic interaction; one was a decreased effect. Thirdly, the greatest risk in the perioperative phase seemed to be post-surgery in comparison with pre-surgery, seen in seven and two patients, respectively. Medication adherence was questionable in five patients (9%).

Conclusion and relevance The occurrence of ischaemic stroke in the perioperative phase in patients treated with a NOAC was a major problem. The main issue seemed to be the discordance between our regional and EHRA guidelines regarding perioperative NOAC management. Apart from the perioperative transition phase, other reasons for occurrence were inappropriate dosing, drug interactions and non-compliance.

Conflict of interest No conflict of interest