Pharmacist Interventions to Reduce Modifiable Bleeding Risk Factors Using HAS-BLED in Patients Taking Warfarin

Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia and is associated with a 5-fold increase in the risk of ischemic stroke and the risk increases with age.^1-3 Oral anticoagulation (OAC) therapy effectively reduces the risk of ischemic stroke in patients with nonvalvular AF. However, OAC therapy carries a bleeding risk.⁴

Several bleeding risk scores have been developed and validated for patients with AF who are taking warfarin: HEMORR₂HAGES (Hepatic or renal disease, Ethanol abuse, Malignancy, Older age, Reduced platelet count or function, Re-bleeding, Hypertension, Anemia, Genetic factors, Excessive fall risk, Stroke), ATRIA (Anticoagulation and Risk Factors in Atrial Fibrillation), and HAS-BLED (Hypertension, Abnormal renal and/or liver function, Stroke, Bleeding history or predisposition, Labile international normalized ratio [INR], Elderly, and Drugs and/or alcohol excess concomitantly).^4,5 All 3 bleeding risk scores demonstrate only modest ability to predict clinically relevant bleeding in patients taking warfarin. The HAS-BLED score was superior to HEMORR₂HAGES and ATRIA for predicting any clinically relevant bleeding and was the only bleeding risk score that demonstrated significant predictive performance for intracranial hemorrhage.⁵ Compared with HEMORR₂HAGES, the HAS-BLED score is simpler to use and to assess risk factors that can be gathered from medical history or routinely tested in patients with AF.⁴ Unlike HAS-BLED, HEMORR₂HAGES and ATRIA do not consider medications that could increase the risk of bleeding.

Despite the availability of validated bleeding risk scores, clinical application of these measures should not be used to exclude a patient from OAC therapy for patients who reach a threshold score. Rather, current guideline and expert consensus agree with the recommendation to use bleeding risk scores to identify risk factors and address those factors that are modifiable to reduce the risk of anticoagulant-associated major bleeding.^6-8

The authors identified modifiable bleeding risk factors using the HAS-BLED score and evaluated pharmacist interventions to correct these factors in patients with nonvalvular AF who are taking warfarin. To the authors’ knowledge, there have been no published studies evaluating interventions to reduce modifiable bleeding risk factors identified by the HAS-BLED score.

Methods

Clinical pharmacy specialists (CPSs) in the primary care (PC) clinics at the Clement J. Zablocki VAMC (CJZVAMC) in Milwaukee, Wisconsin, have prescriptive authority within their scope of practice to manage smoking cessation and diseases, including anticoagulation, diabetes mellitus, heart failure, hypertension, and dyslipidemia. Patients who are on OAC therapy, including warfarin, receive comprehensive anticoagulation management from PC CPSs, including prescribing OAC therapy, education, dosage adjustments, and laboratory monitoring.

Patients were included in the HAS-BLED risk scoring and intervention if their warfarin therapy was managed by a PC CPS, had an active warfarin prescription with a diagnosis of nonvalvular AF in their problem list, and had ≥ 1 modifiable risk factor(s) from the HAS-BLED risk score. Modifiable risk factors evaluated were systolic blood pressure (SBP) > 160 mm Hg, an active prescription for VA or non-VA (which generally indicates over-the-counter [OTC] medication use) aspirin, clopidogrel, or a nonsteroidal anti-inflammatory drug (NSAID). Excess alcohol consumption was not listed as a modifiable risk factor in this assessment because nearly all the anticoagulation patients already receive regular recommendations to minimize alcohol use from the PC CPSs.

Patients were excluded from analysis if they had an indication for warfarin use other than nonvalvular AF, such as atrial flutter, acute/chronic deep vein thrombosis or pulmonary embolism, history of venous thromboembolism, peripheral vascular disease, or aortic or mitral mechanical valve. Patients also were excluded if they were on antiplatelet therapy for unstable coronary artery disease (CAD), experienced acute coronary syndrome within the past 1 year, history of stent placement, carotid endarterectomy, carotid stenosis, or noncardioembolic stroke and no other modifiable risk factors. Last, patients were excluded if clinic SBP readings were > 160 mm Hg but there was documented white coat hypertension or home SBP readings < 160 mm Hg.

The following definitions or measurements were used for assessing the HAS-BLED bleeding risk score⁴:

• Uncontrolled hypertension: most recently charted SBP > 160 mm Hg;
• Abnormal renal function: dialysis, renal transplant, serum creatinine > 2.26 mg/dL;
• Abnormal liver function: chronic hepatic disease, biochemical evidence of significant hepatic derangement (bilirubin > 2 × upper limit of normal and/or AST/ALT/alkaline phosphatase > 3 × upper limit of normal);
• Stroke: including history of transient ischemic attack;
• Bleeding history or predisposition: history of major bleeding (intracranial and/or any bleeding requiring hospitalization and/or causing a decrease in hemoglobin (Hgb) level of > 2 g/dL and/or requiring blood transfusion), anemia (males: Hgb < 13 g/dL; females: Hgb < 12 g/dL);
• Labile INR: percentage of INRs in therapeutic range < 60% (using the CJZVAMC anticoagulation management tool, which calculates percentage of INRs in goal reported since the first visit);
• Geriatric: age > 65 years at initial assessment;
• Concomitant drug use (VA prescription or non-VA medication list): aspirin, clopidogrel, or NSAID; and
• Alcohol in excess: > 8 alcohol servings per week from chart documentation of the patient’s self-report.