JAMA Netw Open. 2026 Oct 1;9(10):e2640377. doi: 10.1001/jamanetworkopen.2026.40377.
ABSTRACT
IMPORTANCE: Direct oral anticoagulants (DOAC) are recommended for selected patients with cancer-associated venous thromboembolism, but their implementation remains heterogeneous.
OBJECTIVE: To describe temporal trends in DOAC initiation for cancer-associated venous thromboembolism in France and to identify factors associated with their use.
DESIGN, SETTING, AND PARTICIPANTS: This nationwide population-based cohort study included adults without recorded contraindications to DOACs and a first treated venous thromboembolism episode between January 1, 2016, and December 31, 2022. Data were collected from the French National Health Data System, a nationwide claims database covering approximately 99% of the population. Analysis was performed from January to August 2025.
EXPOSURES: Patients were classified into active cancer and no active cancer groups. Covariates included calendar period, patient characteristics, cancer type and treatment, venous thromboembolism characteristics, and initial prescriber setting.
MAIN OUTCOMES AND MEASURES: The primary outcome was DOAC initiation within 30 days after the index venous thromboembolism event. Bayesian structural time series models assessed temporal changes across 2016 to 2018, 2019 to 2020, and 2021 to 2022. Multivariable logistic regression identified factors associated with DOAC initiation among patients with active cancer.
RESULTS: Among 683 184 patients with incident treated venous thromboembolism (median [IQR] age, 67 [53-77] years; 347 421 female [51.0%]), 91 954 patients (13.5%) had active cancer while 591 230 (86.5%) had VTE not associated with active cancer. DOAC use increased from 15.3% (95% CI, 15.0%-15.7%) in 2016 to 2018 to 32.6% (95% CI, 32.1%-33.2%) in 2021 to 2022 among patients with active cancer, and from 63.5% (95% CI, 63.3%-63.7%) to 81.1% (95% CI, 80.9%-81.2%) among patients without active cancer. DOAC initiation increased beyond expected secular trends in patients with active cancer, with relative increases of 20.5% (95% credible interval, 13.0%-27.8%) in 2019 to 2020 and 41.3% (95% credible interval, 32.9%-50.2%) in 2021 to 2022. DOAC initiation was less frequent among patients with comorbidities, history of bleeding, gastrointestinal or gynecologic cancer, active brain metastases, or systemic treatment, and when the initial prescriber was a hospital-based physician rather than a general practitioner.
CONCLUSIONS AND RELEVANCE: In this cohort study of patients with incident venous thromboembolism, DOAC use increased substantially in patients with active cancer but remained lower than in patients without active cancer and varied across clinical and care-setting characteristics. This lower uptake may reflect both delayed implementation of recent evidence and the need to individualize anticoagulant selection in complex situations.
PMID:42912389 | DOI:10.1001/jamanetworkopen.2026.40377