LAUSR

Direct-acting oral anticoagulants (DOACs) are the preferred agents for stroke prevention in patients with non-valvular atrial fibrillation (AF). DOACs may require dose adjustment based on several factors, including: age, renal function and body weight. An inappropriate DOAC prescription is defined as a deviation of the drug specific recommended dose as mentioned in the summary of product characteristics. Inappropriate DOAC prescription may consist of both under- and over-dosing, potentially exposing patients to harm. Therefore, we carried out the current study, with the aim of defining the prevalence and predictors of inappropriate DOAC prescription on first attendance of patients at a specialist AF clinic. We performed a retrospective analysis of the clinical database associated with a dedicated AF clinic in a large Irish hospital from August 2015 to March 2020. All new patients who were referred to the clinic and prescribed a DOAC prior to attendance were included. Data collected on patients included demographic and biochemical data in addition to clinical information on medical co-morbidities. In addition, the CHADS2VASc and HASBLED score was calculated for all patients. A multivariable logistic regression model was developed to assess for predictors of inappropriate DOAC dosing. Purposeful variable selection was used with univariate regression performed initially in order to identify predictors to include in the multivariable model. We included 367 patients in the analysis. An inappropriate DOAC dose was identified in 47 of 367 patients (12.8%). The majority of inappropriate DOAC doses were due to under-dosing (76.6%). Patients prescribed an inappropriate DOAC dose tended to be older (78.9±8.4 vs 69.0±10.5 years, p<.001), with higher creatinine (108.5±4.6 vs 88.9±1.3, p<.001). Patients prescribed an inappropriate DOAC dose also tended to have higher CHADS2VASc (3.8±1.7 vs 3.0±1.5, p=.001) and HASBLED scores (2.0±1.0 vs 1.6±1.0, p=.01) than patients prescribed an appropriate DOAC dose. DOAC choice did not differ between the inappropriate and appropriate DOAC dose groups. On univariate logistic regression analysis, several predictors of inappropriate DOAC prescription were identified, including age, renal function, history of falls, CHADS2VASc score and HASBLED score. However, in the multivariate logistic regression model, only increasing age (p<0.001) and decreasing renal function (p<0.001) remained significant predictors of inappropriate DOAC prescription. Over one in eight patients (12.8%) are prescribed an inappropriate DOAC dose on first attendance at a dedicated atrial fibrillation clinic. In the majority of cases, the inappropriate DOAC prescription was secondary to under-dosing. In our multivariable, logistic regression model, increasing age and decreasing renal function were significant predictors of inappropriate DOAC prescription. Type of funding sources: None. Table 1