Real-World Traumatic Intracranial Bleeding in Elderly, Warfarin-Treated Patients a Bigger Problem Than Seen in Trials

Traumatic intracranial bleeding is more common among older US military veterans starting warfarin for A-fib than what has been reported in clinical trials, with risk factors that differ from those for ischemic stroke, a new analysis shows.

Implications:  Real-World Traumatic Intracranial Bleeding in Elderly, Warfarin-Treated Patients a Bigger Problem Than Seen in Trials

Hospitalization for traumatic intracranial bleeding occurred at a rate of 4.80 per 1,000 person-years, high enough to support “the need to more systematically evaluate the benefits and harms of warfarin therapy in older adults,” John Dodson, MD (New York University School of Medicine, New York, NY), and colleagues report in a study published online March 9, 2016, ahead of print in JAMA Cardiology.

They identified several variables associated with a greater risk of bleeding, with dementia being the strongest.

“While we were unable to generate a clinical prediction tool to evaluate risk given poor model discrimination, we still believe that the individual factors we identified may potentially be used in patient-centered discussions about the benefits and harms of warfarin therapy in older adults,” they write, noting the need to validate the findings in other populations.

Those discussions may be particularly important in elderly patients because many are not treated with oral anticoagulation despite its proven effectiveness at lowering the risk of thromboembolic stroke. A primary concern limiting use in this group is traumatic intracranial bleeding related to falls.

“However, despite such concerns, the incidence and determinants of this outcome among older adults with [A-fib] who are prescribed oral anticoagulants remain largely unknown,” the authors point out.

To explore the issue, they performed a retrospective study of 31,951 veterans with A-fib who were 75 years or older (mean age 81.1 years; 98.1% men), had records in the US Department of Veterans Affairs (VA) system, and were new referrals to VA anticoagulation clinics for management of warfarin therapy between 2002 and 2012. Patients taking any of the newer direct oral anticoagulants were excluded.

Absolute rates of hospitalization for traumatic intracranial hemorrhage (primary outcome) were 0.54% at 1 year and 2.10% at 3 years.

After adjustment for potential confounders, several factors emerged as risk factors for traumatic bleeding.

Risk Factors for Traumatic Intracranial Bleeding

The rate of bleeding was higher in patients with at least two of those risk factors than in those with none (6.38 vs 3.36 per 1,000 person-years).

A model incorporating all of the risk factors had poor discrimination for the prediction of traumatic intracranial bleeding, however.

The researchers also found that the rate of any intracranial bleeding—both traumatic and non-traumatic—was 14.58 per 1,000 person-years, which was comparable to the rate of ischemic stroke (13.44 per 1,000 person-years). That similarity “highlights the difficult risk-benefit trade-off that healthcare professionals and older patients face when considering warfarin therapy,” Dodson and colleagues write.

They also showed that although the rate of ischemic stroke increased with higher CHA2DS2-VASc scores, the rate of traumatic intracranial bleeding remained relatively constant. Dodson told TCTMD in an email that the poor correlation between the score and the risk of traumatic bleeding was one of the most unexpected findings of the study and indicates that the risk factors for ischemic stroke and bleeding “appear to be distinct.”

Explaining Risk Factors

Dodson and colleagues point to several potential explanations for the strong relationship between dementia and traumatic bleeding, including difficulty in following medication instructions or dietary noncompliance.

But Dodson highlighted two possible mechanisms in particular. “First, patients with dementia are at considerable risk for falls (and therefore head trauma),” he said. “Second, structural changes in the brain (eg, amyloid deposition) may make clinically overt bleeding more likely after a patient experiences head trauma. These factors likely act synergistically to create a particularly high-risk phenotype.”

The link between labile international normalized ratio (INR) and traumatic bleeding is consistent with results of prior studies. The high rate of labile INR in the current study (55.9%) “underscores the challenge of optimal warfarin management in practice,” Dodson said.

The other risk factors have all been associated with either falls or bleeding in previous analyses.

Absent from the list was a history of falls in the past year, and Dodson said that might have been a consequence of the nature of the information that was available. “I think our ascertainment of fall history was challenging, given our use of administrative data,” he said. “My clinical suspicion is that if we were able to take a detailed fall history in every patient, falls would be an important predictor of risk given what prior studies have found; however, this is speculative.”

Era of Increasing Use of Non-Warfarin Options

The authors note that the study was limited by excluding patients taking any of the newer direct oral anticoagulants, which have been shown to carry a lower risk of intracranial bleeding compared with warfarin.

“When we started this study, direct oral anticoagulants were newly approved, and we therefore did not have a sufficient length of observation to generate a large cohort of patients taking these medications,” Dodson said. “I think this is a critical question for a future study: determining the comparative risk of direct oral anticoagulants versus warfarin in terms of traumatic intracranial bleeding.”

Commenting on the study, Kelvin Ng, MBBS (Hamilton General Hospital, Hamilton, Canada), noted in an email to TCTMD that even though use of direct oral anticoagulation is on the upswing, “there exists a large proportion of patients already established on warfarin anticoagulation, of which the majority are elderly. Most guidelines do not advocate swapping from warfarin to a direct acting oral anticoagulant if the patient is already well established on warfarin.”

Nevertheless, because of the lower risk of intracranial bleeding, direct oral anticoagulants “may have a greater role in [the elderly] if patient factors, in particular medication compliance, are favorable,” Ng said, pointing out, however, that physicians should be cautious about using a set age threshold to define elderly patients.

“Chronological age is really a surrogate marker for other variables and frailty, which can be difficult to quantify accurately,” he said.


  • Dodson JA, Petrone A, Gagnon DR, et al. Incidence and determinants of traumatic intracranial bleeding among older veterans receiving warfarin for atrial fibrillation. JAMA Cardiol. 2016;Epub ahead of print.

Related Stories:

Todd Neale is the Associate News Editor for TCTMD and a Senior Medical Journalist. He got his start in journalism at …

Read Full Bio
  • The study was supported by a grant from the National Institutes of Health/National Institute on Aging and by a T. Franklin Williams Scholarship Award (funding provided by Atlantic Philanthropies, the John A. Hartford Foundation, the Alliance for Academic Internal Medicine–Association of Specialty Professors, and the American College of Cardiology).
  • Dodson and Ng report no relevant conflicts of interest.