Stopping Anticoagulation in LTC Patients with Atrial Fibrillation: No Rise in Stroke but Higher Mortality

Among Medicare beneficiaries with atrial fibrillation (AF) entering long-term care (LTC), discontinuing anticoagulation (AC) was not associated with a higher risk of stroke or bleeding, although discontinuers had higher mortality, according to a study published in JAMA Network Open.

The investigators noted that AC is the main treatment to prevent stroke in AF, but bleeding events and falls lead many patients to discontinue it, and many LTC residents have life-limiting illness or disability that affects AC decisions. According to the authors, professional society guidelines recommend AC for most older adults but do not specifically address LTC residents or discontinuation, and trial data in this setting are unavailable.

The comparative effectiveness study used target trial emulation with clone censor weighting to analyze Medicare claims for 15,426 residents (mean age, 82.6 years; 62.6% female) who were receiving AC when they entered LTC between April and December 2020. All had elevated stroke risk by CHA2DS2-VASc score; residents with other AC indications, such as a prosthetic heart valve or venous thromboembolism, and those receiving hospice care were excluded. The analysis did not identify a significant difference in stroke risk between discontinuers and continuers (subdistribution hazard ratio, 1.07; 95% CI, 0.93-1.23), and bleeding risk was also similar between groups. Mortality, however, was higher among discontinuers (hazard ratio, 1.27; 95% CI, 1.23-1.31), with estimated annual death rates of 46.8% and 39.8%, respectively.

Results were similar among residents taking direct oral anticoagulants at LTC entry, who made up 75.2% of the cohort. The negative control outcomes of pneumonia and urinary tract infection showed no significant excess risk among discontinuers, which the authors said argued against unmeasured confounding by frailty or illness severity. Still, they noted that unmeasured or residual confounding may explain the excess mortality, that stroke and bleeding were identified from hospitalization diagnosis codes without imaging or laboratory confirmation, and that enrollment overlapped with the COVID-19 pandemic.

The authors concluded that high mortality and the apparent excess of deaths among discontinuers leave the clinical implications unclear, and they said an individual’s health goals at this life stage require careful consideration when deciding whether to continue AC. They added that the analysis remains observational and called for research into the excess mortality and replication in more recent LTC cohorts.

Source: Kapoor A, Chen Z, Patel J, et al. Stroke and bleeding risk in long-term care residents with atrial fibrillation who discontinue anticoagulation. JAMA Netw Open. 2026;9(10):e2640964. doi:10.1001/jamanetworkopen.2026.40964