Back

Overview of Direct Oral Anticoagulation Trends in the North Bronx: Patient and Systemic Factors Contributing to Medication Nonadherence

Khasnavis, S.; Moon, J.-Y.; Lerner, D.; Leiderman, E.

2025-05-02 health systems and quality improvement
10.1101/2025.04.28.25326611 medRxiv
Show abstract

IntroductionAnticoagulation nonadherence has historically been attributed to myriad factors in patient populations across the world. While direct oral anticoagulants (DOACs) have demonstrated advantages such as fewer interactions with medications and freedom from routine lab monitoring compared to other anticoagulants, nonadherence persists and the underlying contributors vary by geography. MethodsThis retrospective review used all available electronic medical records on patients receiving active DOAC therapy from primary care centers in the Bronx between 2017-2024. Adherence and nonadherence groups were determined by prescription fill status and provider documentation. The two groups were compared by age, gender, race, ethnicity, insurance type, diagnosis, pharmacy type, employment status, comorbidities, home medications, and primary language. Univariable and multivariable logistic regressions were applied between the groups and categories. P values < 0.05 were deemed significant. ResultsThe cohort had 938 patients with nonadherence reported in 227 (24.2%) patients. Nonadherence was more common in younger patients (OR 1.20, CI 1.04-1.39, p=0.010) and in males (OR 1.44, CI 1.04-1.99, p=0.026). It was more common in black Non Hispanics (OR 1.98, CI 1.17-3.47, p=0.014) and in the employed (OR 1.92, CI 1.09-3.57, p=0.032). It was also more frequent among patients prescribing from hospital pharmacies (OR 2.63, CI 1.89-3.57, p=0.001) and in English speakers (OR 1.69, CI 1.11-2.63, p=0.016). ConclusionDOAC nonadherence in the Bronx is significantly associated with younger age, male gender, black Non Hispanic origins, employment, hospital pharmacy prescriptions, and English primary language. Other categories were borderline significant or not significant and warrant future investigation.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.