Real-World Health Care Costs Based on Medication Adherence and Risk of Stroke and Bleeding in Patients Treated with Novel Anticoagulant Therapy

BACKGROUND: With the lack of real-world evidence, the challenge for drug reimbursement policy decision makers is to understand medication adherence behavior among users of novel oral anticoagulants (NOACs) and its effect on overall cost savings. No study has examined and quantified the burden of cost in high-risk patients taking NOAC therapy. OBJECTIVE: To examine the association of cost with adherence, comorbidity, and risk of stroke and bleeding in patients taking NOACs (rivaroxaban and dabigatran). METHODS: A retrospective cohort study used deidentified data from a commercial managed care database affiliated with Optum Clinformatics Data Mart (January 1, 2010-December 31, 2012). Patients aged 18 years and older with ≥ 1 diagnosis of atrial fibrillation/flutter, > 1 NOAC prescription, 6-month pre-index and 12-month post-index continuous enrollment, and CHA2DS2-VASc score ≥ 1 were included. Adherence was calculated using proportion of days covered (PDC ≥ 80%) over an assessment period of 3, 6, and 12 months and compared based on level of comorbidity, stroke, and bleeding risk. The adjusted annual health care costs per patient (drug, medical, and total) were calculated using multivariable gamma regression controlling for demographic and clinical characteristics and compared across groups based on adherence over 12 months, baseline level of comorbidity, and risk of stroke and bleeding. RESULTS: Of 25, 120 NOAC patients, 2, 981 patients were included in the final cohort. Based on a PDC threshold of ≥ 80%, the adherence rate over 3, 6, and 12 months was 72%, 65%, and 54%, respectively. For all time periods, the level of adherence significantly increased (P < 0. 001), with an increase in stroke risk (based on CHA2DS2VASc scores of 1, 2-3, and 4+) ; comorbidity (Charlson Comorbidity Index scores of 0, 1-2, and 3+) ; and risk of bleeding (HAS-BLED scores of 0-1, 2, and 3+). Adjusted all-cause total cost calculated for a 12-month period was significantly lower (29, 742 vs. 33, 609) among adherent versus nonadherent users. Drug cost was higher (5, 595 vs. 2, 233) among adherent versus nonadherent patients but was offset by lower medical costs (23, 544 vs. 30, 485) costs. The overall cost significantly increased for patients with a high risk of bleeding and a high level of comorbidity. CONCLUSIONS: Adherence to NOAC therapy led to a reduction in overall health care cost, since higher drug costs were offset by lower medical (inpatient and outpatient) costs among adherent patients. Cost information based on adherence and risk of stroke and bleeding can help formulary decision makers to assess risk-benefit and help clinicians in developing interventions to reduce patient burden. DISCLOSURES: Funding to acquire the data source was provided by the University of Rhode Island College of Pharmacy, Kingston, to support PhD dissertation work. Deshpande is currently an employee of Pharmerit International.