The vitamin K antagonists (VKAs) are currently the most eective therapeutic class for the prevention of cerebrovascular events in atrial fibrillation (AF) patients. However, several studies showed an under-prescription of this therapy. The aim of the study was to assess the prescription of VKAs in non-valvular AF (NVAF) patients and factors influencing the non-prescription of such treatment.
We conducted a prospective, observational study in an emergency department (ED). Patients with high thromboembolic risk NVAF and not receiving VKAs beforehand were included. Calculation of CHA2DS2-VASc and HAS-BLED scores was performed. An analytic study was conducted in order to identify independent predictors of the under-prescription of VKAs.
RESULTS
During study, 176 patients were enrolled, the mean age was 67±13 years and 66% were women. The mean CHA2DS2VASc and HASBLED scores were 2.88 ± 1.55 and 1.52 ± 1.05, respectively. Among our cohort, VKA was prescribed in 36% of cases. Age >70 years (OR=1.59, 95%CI[1.11-2.21],p<0.001), creatinine level ≥110 μmol/l (OR=2.54,95%CI[1.20–5.37],p=0.01) and aspirin use (OR=1.7,95%CI [1.08-2.67],p=0.02) were independently associated with under-prescription of VKAs. Bedside, the main causes reported by the emergency physicians (EP) were: factors related to patient characteristics (n=38,34%), factors related to emergency physician (n=62,55%), factors related to the patient environment (n=20,17%) and factors related to the drug (n=22,23%).
Our results showed that the prescription of VKAs was low in ED. e reasons of VKA under-prescription are linked usually to several factors inherent to patient and to the adherence of EP to new recommendations.
KEY WORDS : Non valvular atrial brillation - Anticoagulation - Underuse - Emergency Departments.
Atrial brillation (AF) is the most common signicant cardiac rhythm disorder encountered in emergency departments (EDs) being present in 3.6 to 7% of general emergency visits (1, 2). It affects 2 to 5% of the general population, and increases substantially with age (3-5). Patients with AF have at least a 5-fold increased risk of ischemic stroke (6).
Approximately, 15% of all strokes are associated with this dysrhythmia, and the association increases with age (7,8). Stroke prevention is therefore crucial to reduce mortality and disability in patients with AF, and especially in those with the highest risk of embolic events (9).
Major randomized primary prevention trials conducted in patients with AF have shown that oral anticoagulation (OAC), reduce the risk of embolic events (9-13).
The vitamin K antagonist (VKA) is currently the most prescribed therapeutic class for the prevention of embolic events. However, hemorrhagic stroke is the major complication of VKA, it come to the forefront of iatrogenic accidents and are responsible for approximately 1.1 to 7.4% per year independently of age and from 1.1 to 13% per year in elderly patients (14). It is therefore essential to accurately assess the benet / risk ratio of VKA to each patient to whom anticoagulation is proposed.
Diverse studies conducted in the hospital setting, in general practice, and in EDs have found that anticoagulation is prescribed to less than 55% of eligible patients [14]. e reasons of this underuse remain unclear. EDs play a substantial role in AF management, notably, concerning stroke prophylaxis. On the other hand, a signicant proportion of AF patients attending EDs are at high risk of stroke and are not receiving anticoagulants (15). erefore, these patients may benet from adequate prescription to improve their prognosis and long-term quality of life.
In this perspective, we conducted this study whose objectives were to assess the prescription of VKAs in non-valvular AF (NVAF) patients and factors inuencing the non-prescription of this therapy.
We conducted a prospective, observational and mono-centric study in an ED over a period of two years (April 2013 to April 2015).
Patient Selection
We included consecutive patients older than 18 years, attended to the medical area of the ED during the study period, with the diagnosis of AF eligible for anticoagulation by VKA ccording to guidelines of the European Society of Cardiology (ESC) (3).
AF was documented in an electrocardiogram (ECG) obtained when the treating physician considered it necessary during clinical evaluation. Exclusion criteria for the study were valvular AF, patients receiving OAC, low risk of stroke (CHA2DS2-VASc=0), contraindications to VKA and hemodynamic instability.
Data collection
Data collected from patients included demographic data, comorbidities, disability, type of AF (rst episode, paroxysmal, persistent and permanent), risk factors for stroke according to theCHA2DS2-VASc scores, bleeding risk–prediction factors according to the HAS-BLED score, symptoms that caused ED consultation, clinical presentation and ED evaluation, arrhythmia management (rhythm or rate control), patients’ outcome (symptom relieving and type of cardiac rhythm at discharge) , nal disposition (discharge, admission, or death) and stroke prophylaxis prescribed in the ED (anticoagulation, antiplatelet therapy, or both).
Patients with a CHA2DS2-VASc score ≥ 1 are classied as at high risk of stroke, and therefore anticoagulation is recommended in patients without contraindications. VKA was prescribed in a doseadjusted approach to achieve an international normalized ratio (INR) target of 2.5 (2-3). If anticoagulation was not prescribed, the treating physician was asked to explain the reason for not precribing it. Patients were divided into two groups: VKA + Group: patients received VKA and VKA – Group. The two groups were analyzed and compared in order to identify factors independently associated with VKA under-prescription.
Finally, the reasons why physicians did not give VKA were described with detailed arguments on the benet/ risk of OAC.
Follow-up for adverse events (ischemic stroke, hemorrhagic complications) and death was performed during a three months period. We used telephone follow-up, in addition to record review.
Data Analysis
Statistical analysis was carried out with SPSS (version 18.0) statistical software package. Continuous variables are presented as means ± standard deviation (SD) and discrete variables as absolute values and percentages. An univariate analysis comparing the two groups was performed, with the chi -square test with Yates’ correction or Fisher’s exact test when appropriate, odds ratio (OR) with 95% condence intervals (CI), and the unpaired t-test.
The logistic regression analysis with no prescription OAC as dependent variable was there run. The analysis was performed with a binary logistic regression and “enter” method, with an entry criterion of 0.05 and a removal criteria of 0.10. Dierences were considered to be statistically signicant with P
Characteristics of the study population:
Between April 2013 and April 2015, 126,800 patients were presented to the ED, 249 of these patients were found to have AF and 176 patients were included in the study (Figure 1).
The mean CHA2DS2-VASc score was 2.88±1.55. The mean HASBLED score was 1.52±1.05. A high bleeding risk (HASBLED score ≥3) was found in 18% of patients.
Only sixty-four patients (36%) were received VKA at ED. Anticoagulation prescription according to the CHA2DS2-VASc score was illustrated in Figure 2.
Performed in 17% of patients and rate control (in patients with AF >48 h) in 40% of patients. Eighty-eight percent (88%) of patients were discharged home from ED.
Factors associated with the non-prescription of VKA: Patients who were not receiving VKA were older (69 vs. 61 years, p<0.001), more likely to have permanent AF, their HAS-BLED score was higher (3.09 ± 1.64 vs. 1.17 ± 0.8, p<0.001) and had more than three drugs per day in their medication prescription (37 vs. 20%, p=0.01).
The prevalence of ischemic stroke at three months was higher in the VKA- group (7 vs. 0%, p=0.02), and the incidence of hemorrhagic complications was higher in VKA + group (4 vs. 0%, p= 0.04). The rate of death was similar in the 2 groups (3 vs. 4%, p=0.5).
Factors independently associated with lack of oral anticoagulation prescription in eligible patients on logistic regression analysis were showed in table 2.
Our study has shown that the prescription rate of VKAs was low (36%). Factors independently associated with the prescription were related to the characteristics of the patients advanced age, using aspirin and past medical history of renal failure).
The leading reason why patients were not receiving OAC was a «potential contra-indication» (especially advanced age), followed by «lack of knowledge of recommendations», cognitive impairment and physical dependence of patient and «fear of bleeding». The medical management of AF (other than antiarrhythmic drugs) is primarily focused on reducing the risk of stroke. The prevention of thromboembolic events is based on the establishment and monitoring of anticoagulation now mainly based on the use of VKA.
This preventive treatment is a well codied theory, based on the thromboembolic risk assessed by the CHA2DS2-Vasc score (3). Although the VKA prescription rates increased over time since the publication of AFASAK, BAATAF, SPAF, CAFA, SPINAF studies between 1989 and 1992, an under-prescription of VKA has been observed in patients with NVAF at high thromboembolic risk (11-13,16).
VKA prescription rates in patients with AF vary from 28-64%. In elderly patients (from 75 to 80 years), it vary from 35 to 46% (17,18).
We found that despite a mean age of patients more than 65 years, with high risk of thromboembolism, the prescription of VKA was around 36%. ese data were consistent with the results of many recent studies (19-21).
Patients not receiving VKA were older (69 vs. 61 years, p<0.001), had higher prescription medication including more than 3 drugs per day (37 vs. 20%, p=0.01) and higher HASBLED score (3.09 ± 1.64 vs. 1.17 ± 0.8, p <0.001).
The number of ischemic stroke was higher in the VKA- group, and bleeding events were more in the VKA + group. The mortality rate was similar in both groups.
Advanced age (>70 years) was independently associated with the lack of anticoagulant prescription to eligible patients. These data are consistent with those in other studies in which the elderly population was often excluded from anticoagulant treatment (15, 22-27).
The fear of bleeding is one of the main barriers in these patients, and may override the perception of these patients’ risk of stroke (28-30). It was provided as one of main explanations by our physicians. They probably consider that it is more serious to have severe bleeding because of properly prescribed and managed OAC, than to have severe embolic events in the absence of OAC. These explanations had already been suggested (3,28,30).
One of the major reasons provided in SAFE II by GPs, and sometimes cardiologists, not to prescribe OAC, was that they thought there was no indication (28,29). is reason was provided by 36% of physicians in our study. This finding supports the hypothesis that the main reason why NVAF patients are not receiving OAC in practice is a lack of knowledge of trials and guidelines.
Because the guidelines’ recommendations have been demonstrated to be widely applicable in daily practice (3,27,32), and previous studies in local settings have demonstrated the usefulness of specic training to improve management’s adequacy, educational eorts to increase physicians’ adherence to the guidelines appear warranted (36).
Previous cognitive impairment, falls or gait disturbances and living conditions were often considered as major contra-indications for OAC. These factors have not been clearly evaluated until now and should be only considered on an individual basis when the risk of such a condition appears higher than that of the natural history of the disease.
Because stroke is the major risk of AF and because OAC can drastically reduce this risk, primary and secondary prevention of ischaemic stroke among patients with AF by an appropriate use of OAC is a crucial issue to improve quality of care.
The main limitation of this study is the small number of patients; this does not reect the real rate of the VKA prescription in the other ED and cardiology department. Another limitation is that our data lack objective assessment of cognitive status, falls or gait disturbances and living conditions.
In is study, the prescription of VKA in patients with high thromboembolic risk NVAF was low in ED. e reasons of VKA under-prescription are linked usually to several factors inherent to
patient and to the adherence of EP to new recommendations.