Incidence and Determinants of Traumatic Intracranial Bleeding Among Older Veterans Receiving Warfarin for Atrial Fibrillation | Atrial Fibrillation | JAMA Cardiology | JAMA Network
[Skip to Navigation]
Sign In
Figure 1.  Derivation of the Study Cohort
Derivation of the Study Cohort

Of 99 552 patients initially identified through administrative data, 31 951 were included in the final sample. AF indicates atrial fibrillation; ICD-9-CM, International Classification of Diseases, Ninth Revision, Clinical Modification; and INRs, international normalized ratios.

Figure 2.  Kaplan-Meier Plot of Traumatic Intracranial Bleeding Events Over Time
Kaplan-Meier Plot of Traumatic Intracranial Bleeding Events Over Time

Based on the Weibull distribution (Weibull shape, 1.28; 95% CI, 1.27-1.29), there was a modest increase in the hazard of traumatic intracranial bleeding over time.

Figure 3.  Rate of Traumatic Intracranial Bleeding and Stroke, Stratified by CHA2DS2-VASc Score
Rate of Traumatic Intracranial Bleeding and Stroke, Stratified by CHA2DS2-VASc Score

The observed rate of stroke increased as the CHA2DS2-VASc score increased, while the rate of traumatic intracranial bleeding remained constant across CHA2DS2-VASc score categories. CHA2DS2-VASc indicates congestive heart failure, hypertension, age 75 years or older, diabetes mellitus, stroke/transient ischemic attack, vascular disease, age 65 to 74 years, and female sex.

Table 1.  Study Sample Characteristics
Study Sample Characteristics
Table 2.  Unadjusted and Adjusted Models
Unadjusted and Adjusted Models
Original Investigation
April 2016

Incidence and Determinants of Traumatic Intracranial Bleeding Among Older Veterans Receiving Warfarin for Atrial Fibrillation

Author Affiliations
  • 1Leon H. Charney Division of Cardiology, Department of Medicine, New York University School of Medicine, New York
  • 2Veterans Affairs New York Harbor Healthcare System, New York
  • 3Massachusetts Veterans Epidemiology Research and Information Center, Veterans Affairs Boston Healthcare System, Boston
  • 4Department of Biostatistics, Boston University School of Public Health, Boston, Massachusetts
  • 5Section of Geriatrics, Department of Medicine, Yale School of Medicine, New Haven, Connecticut
  • 6Department of Chronic Disease Epidemiology, Yale School of Public Health, New Haven, Connecticut
  • 7Section of Cardiovascular Medicine, Department of Internal Medicine, Yale School of Medicine, New Haven, Connecticut
  • 8Center for Outcomes Research and Evaluation, Yale–New Haven Hospital, New Haven, Connecticut
  • 9Robert Wood Johnson Foundation Clinical Scholars Program, Yale School of Medicine, New Haven, Connecticut
  • 10Department of Health Policy and Administration, Yale School of Public Health, New Haven, Connecticut
  • 11Division of Aging, Department of Medicine, Brigham and Women’s Hospital/Harvard Medical School, Boston Massachusetts
JAMA Cardiol. 2016;1(1):65-72. doi:10.1001/jamacardio.2015.0345

Importance  Traumatic intracranial bleeding, which is most commonly attributable to falls, is a common concern among health care professionals, who are hesitant to prescribe oral anticoagulants to older adults with atrial fibrillation.

Objective  To describe the incidence of and risk factors for traumatic intracranial bleeding in a large cohort of older adults who were newly prescribed warfarin sodium.

Design, Setting, and Participants  Retrospective cohort study at the US Department of Veterans Affairs (VA). Participants included 31 951 veterans with atrial fibrillation 75 years or older who were new referrals to VA anticoagulation clinics (for warfarin therapy) between January 1, 2002, and December 31, 2012. The dates of the core analysis were March 2014 through May 2015, and subsequent ad hoc analyses were performed through December 2015. Patients with comorbid conditions requiring warfarin were excluded.

Main Outcomes and Measures  The primary outcome was hospitalization for traumatic intracranial bleeding. Secondary outcomes included hospitalization for any intracranial bleeding or ischemic stroke. We used International Classification of Diseases, Ninth Revision, Clinical Modification codes to identify the incidence rates of these outcomes after warfarin initiation using VA administrative data (in-system hospitalizations) and Medicare fee-for-service claims data (out-of-system hospitalizations). Clinical characteristics, laboratory results, and pharmacy data were extracted from the VA electronic medical record. For traumatic intracranial bleeding, Cox proportional hazards regression was used to determine predictors of interest selected a priori based on prior known associations.

Results  The study population comprised 31 951 participants. The mean (SD) patient age was 81.1 (4.1) years, and 98.1% were male. Comorbidities were common, including hypertension (82.5%), coronary artery disease (42.6%), and diabetes mellitus (33.8%). During the study period, the incidence rate of hospitalization for traumatic intracranial bleeding was 4.80 per 1000 person-years. In unadjusted models, significant predictors of traumatic intracranial bleeding included dementia, fall within the past year, anemia, depression, abnormal renal or liver function, anticonvulsant use, labile international normalized ratio, and antihypertensive use. After adjusting for potential confounders, the remaining significant predictors for traumatic intracranial bleeding were dementia (hazard ratio [HR], 1.76; 95% CI, 1.26-2.46), anemia (HR, 1.23; 95% CI, 1.00-1.52), depression (HR, 1.30; 95% CI, 1.05-1.61), anticonvulsant use (HR, 1.35; 95% CI, 1.04-1.75), and labile international normalized ratio (HR, 1.33; 95% CI, 1.04-1.72). The incidence rates of hospitalization for any intracranial bleeding and ischemic stroke were 14.58 and 13.44, respectively, per 1000 person-years.

Conclusions and Relevance  Among patients 75 years or older with atrial fibrillation initiating warfarin therapy, the risk factors for traumatic intracranial bleeding are unique from those for ischemic stroke. The high overall rate of intracranial bleeding in our sample supports the need to more systematically evaluate the benefits and harms of warfarin therapy in older adults.


Advanced age is a powerful risk factor for thromboembolic stroke in patients with atrial fibrillation (AF), and oral anticoagulation reduces this risk by almost two-thirds in patients at risk.1,2 However, up to half of the eligible older adults with AF are not treated with anticoagulant therapy due to health care professionals’ concerns about potential treatment-related harms.3-5 Multiple studies4-9 have shown that a primary concern among health care professionals is perceived fall risk and the sequelae of traumatic intracranial bleeding, which can have devastating consequences. However, despite such concerns, the incidence and determinants of this outcome among older adults with AF who are prescribed oral anticoagulants remain largely unknown. Most prior investigations are small and report that the outcome is rare.10,11

Accordingly, we sought to investigate the long-term incidence of and risk factors for traumatic intracranial bleeding among a large cohort of older adults with AF after the initiation of oral anticoagulation with warfarin sodium. To accomplish this end, we assembled a retrospective sample of patients from the US Department of Veterans Affairs (VA) health system, which contains key clinical variables not available in many other large data sets, and then used Medicare fee-for-service claims data to capture subsequent out-of-system hospitalizations. We hypothesized that the incidence rate of traumatic intracranial bleeding would be higher than previously reported in clinical trials and that selected known risk factors for falls and bleeding would independently predict this event. As secondary outcomes, we analyzed the incidence of any intracranial bleeding (traumatic and nontraumatic) and ischemic stroke.

Data Sources

The VA is the largest integrated health system in the United States and has a well-developed clinical information system that has been used extensively for large-scale epidemiologic investigations.12-16 For our study, we obtained VA medical and administrative data through the VA Corporate Data Warehouse, including outpatient encounters and inpatient hospitalizations (dates and International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] codes), laboratory results, and pharmacy dispensing. Because we anticipated that almost our entire cohort would also be covered by Medicare17 and hospitalized at non-VA institutions, we obtained Medicare enrollment information and claims data (fee for service) for inpatient hospital stays from the Centers for Medicare & Medicaid Services through the VA Information Resource Center (Hines, Illinois). Veterans’ data were linked to Medicare claims data using unique scrambled Social Security numbers.

Data were stored on VA Informatics and Computing Infrastructure servers and analyzed through a secure network at the Massachusetts Veterans Epidemiology Research and Information Center (Boston, Massachusetts). Study approval and a waiver of informed consent were obtained from institutional review boards at the Veterans Affairs Boston Healthcare System and the Veterans Affairs New York Harbor Healthcare System.

Study Sample

Because the risk of warfarin-related bleeding is highest immediately after the initiation of therapy,18 we created an inception cohort of veterans 75 years or older who were started on warfarin therapy between January 1, 2002, and December 31, 2012. The dates of the core analysis were March 2014 through May 2015, and subsequent ad hoc analyses were performed through December 2015. Ambulatory patients within the VA system who are prescribed warfarin are followed up in dedicated anticoagulation clinics, which can be identified by an administrative “stop code” (code 317). These codes are used to characterize the distinct types of services a patient receives during each VA encounter. We identified all patients for whom this code was entered at least once within our study period. Our index date was defined as the date of first warfarin dispensing. Among this population, we then identified patients who had a diagnosis of AF (ICD-9-CM code 427.3) within 1 year before the index date (one inpatient or 2 outpatient encounters). To reduce the potential for out-of-system anticoagulation clinic use, we also required that patients had to have an international normalized ratio (INR) available at least every 3 months between the time of their first and last warfarin prescriptions. We also excluded patents with no INRs exceeding 1.2 because it was unlikely that they actually took warfarin. To ensure that patients were existing users of VA services, we required at least 1 outpatient visit or outpatient pharmacy dispensing between 6 months and 2 years before first warfarin dispensing.

Because risk-benefit considerations with anticoagulation vary by condition, we excluded patients if the following concomitant diagnoses (ICD-9-CM codes) were also present: pulmonary embolus (code 415.1), deep venous thrombosis (code 453.4), mechanical mitral valve (code 35.24 or V43.3), or mechanical aortic valve (code 35.22 or V43.3) within 1 year before first warfarin dispensing. Patients who were prescribed direct oral anticoagulants available during the study period (dabigatran, rivaroxaban, or apixaban) were also excluded.

Derivation of the study cohort using the above criteria is shown in Figure 1. Of 99 552 unique patients 75 years or older with at least 1 warfarin prescription within the VA between January 1, 2002, and December 31, 2012, a total of 42 838 had AF, and 41 022 had AF without another diagnosis that would require long-term anticoagulation. After applying the remainder of our exclusion criteria, our final sample included 31 951 patients.


Our primary outcome was hospitalization for traumatic intracranial bleeding, which was determined by the use of ICD-9-CM codes entered for hospitalizations either within VA hospitals (VA administrative data) or at outside hospitals (Medicare fee-for-service claims data). The use of ICD-9-CM coding is a previously accepted method of identifying hospitalizations from large data sets.19,20 In claims data, ICD-9-CM codes for intracranial bleeding are explicitly specified as traumatic or nontraumatic. We selected our ICD-9-CM codes based on prior studies10,21 and consensus among an expert panel of health care professionals (J.A.D., D.R.G., M.E.T., H.M.K., and J.M.G.) (eTable 1 in the Supplement). To increase the likelihood that these events were attributable to falls, we excluded all patients with a concomitant motor vehicle crash code (ICD-9 CM codes E810.0-E819.9).22

Secondary outcomes included (1) hospitalization for any intracranial bleeding (traumatic or nontraumatic) and (2) hospitalization for ischemic stroke. To identify secondary outcomes, we followed an algorithm using ICD-9-CM codes entered at discharge (eTable 1 in the Supplement).

Predictors and Covariates

Our predictors of interest were prespecified and based on prior studies showing an association with either falls or bleeding. For falls, we required that risk factors had to be reported in at least 2 independent studies22-28 and be available through administrative data. Our risk factors included dementia, arthritis, fall within the past year, visual impairment, anemia, dizziness, diabetes mellitus, depression, low body mass index, anxiolytic use, antipsychotic use, anticonvulsant use, antihypertensive use, and polypharmacy (≥4 chronic medications). Our bleeding risk factors were based on the previously validated HAS-BLED model29 (hypertension, abnormal renal or liver function, stroke, prior bleeding, drug or alcohol use, labile INR, and age).

For medical conditions in our model, we used previously validated ICD-9-CM coding algorithms12-14,30-32 when available (eTable 2 in the Supplement). For pharmacy variables, we used categories within the VA pharmacy system.33 Patients were classified as receiving a medication regimen if they had a prescription within 6 months before cohort entry. Medication status was updated each month thereafter, with individuals counted as discontinuing medication if there was no refill for 6 months after the supply of their prior fill ran out. For the laboratory variable of abnormal renal or liver function (including aspartate aminotransferase, alanine aminotransferase, bilirubin, and creatinine), values were based on the most recent available result within 6 months before cohort entry. For the INR variable, we used the linear interpolation method by Rosendaal et al,34 which assumes a linear relationship between 2 INRs and assigns a specific daily INR for each patient. The mean time in therapeutic range (INR, 2.0-3.0) is then determined. We initially divided time in therapeutic range into the following 3 categories based on the classification scheme by White et al35: poor control (<60%), moderate control (60%-75%), and good control (>75%). We considered patients with poor control as having labile INR.36

Statistical Analysis

We reported the incidence rate of our primary outcome (hospitalization for traumatic intracranial bleeding) as the number of events per 1000 person-years. Patients were censored at the time of hospitalization for a first event, or if they did not experience an event, they were censored at death or at the end of the study period (December 31, 2012). We then examined bivariate associations between our predictors of interest and covariates (described above) with the outcome. Subsequently, we used Cox proportional hazards regression to evaluate the independent association of our predictors of interest with the outcome. In addition to adjusting for all prespecified predictors in this model, we adjusted for age, sex, self-identified race/ethnicity, and common comorbidities (peripheral vascular disease, coronary artery disease, hyperlipidemia, chronic obstructive pulmonary disease, and heart failure), which were identified from ICD-9-CM codes entered within 1 year before patients’ entry into the cohort (eTable 2 in the Supplement). Within our Cox proportional hazards regression model, medications were updated monthly, and time in therapeutic range was updated every 3 months. These factors were treated as time-varying covariates. Otherwise, baseline variables were used. We reported the strength of associations using hazard ratios (HRs) for point estimates with 95% CIs. The proportional hazards assumption was tested for each predictor in our multivariable model and was found to be valid. For purposes of clinical applicability, we then ran 2 separate logistic regression models for the outcomes of 1-year and 3-year traumatic intracranial bleeding, including any variable with P < .20 in unadjusted analyses. We reported C statistics for these outcomes to evaluate model discrimination for the prediction of traumatic intracranial bleeding at 1 year and 3 years.

For descriptive purposes, we also reported our primary outcome without truncating patient follow-up at the time of first event to account for repeated events in a single patient. For our secondary outcomes (any intracranial bleeding and ischemic stroke), to capture multiple events, we reported the incidence rates without censoring patients after their first event. For stroke, we also classified the observed rates of stroke and traumatic intracranial bleeding based on predicted stroke risk using the CHA2DS2-VASc (congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, stroke/transient ischemic attack–vascular disease, age 65-74 years, female sex) score.37

Study Cohort

Of the 31 951 patients in our final cohort, the mean (SD) age was 81.1 (4.1) years, 98.1% were male, and 8.4% were of nonwhite race/ethnicity. Frequent comorbidities included hypertension (82.5%), coronary artery disease (42.6%), and chronic obstructive pulmonary disease (25.5%). Polypharmacy was common, with 87.0% of patients using at least 4 chronic medications. The mean (SD) time in therapeutic range for the entire study sample was 55.5% (22.7%). Most patients (55.9%) were classified as having labile INR. On average, among patients with labile INR, 42.7% of INRs were less than 2.0, 39.4% were 2.0 to 3.0, and 17.9% were greater than 3.0. Other study sample characteristics are summarized in Table 1.

Primary Outcome

Over a median follow-up period of 2.97 years, the incidence rate of traumatic intracranial bleeding was 4.80 per 1000 person-years. More than three-quarters of events (76.0%) were identified from Medicare claims data, and the remainder occurred within the VA system. A Kaplan-Meier curve illustrating traumatic intracranial bleeding events over time is shown in Figure 2. Based on the Weibull distribution (Weibull shape, 1.28, 95% CI, 1.27-1.29), there was a modest increase in the hazard of this outcome over time. The absolute event rates for traumatic intracranial bleeding were 0.54% at 1 year and 2.10% at 3 years. When we included multiple events per patient (without censoring), the incidence rate of traumatic intracranial bleeding was 6.16 per 1000 person-years.

In our unadjusted model, antihypertensive use was the strongest risk factor for traumatic intracranial bleeding (HR, 2.63; 95% CI, 2.06-3.35), although this effect was no longer significant after multivariable adjustment (HR, 1.15; 95% CI, 0.89-1.49) (Table 2). Other significant risk factors included dementia (HR, 2.11; 95% CI, 1.53-2.92), fall within the past year (HR, 1.72; 95% CI, 1.21-2.44), anemia (HR, 1.35; 95% CI, 1.10-1.65), depression (HR, 1.49; 95% CI, 1.22-1.82), abnormal renal or liver function (HR, 1.42; 95% CI, 1.00-2.01), anticonvulsant use (HR, 1.30; 95% CI, 1.01-1.69), and labile INR (HR, 1.40; 95% CI, 1.10-1.80). After adjusting for potential confounders, dementia, anemia, depression, anticonvulsant use, and labile INR remained significant predictors, with the strongest association seen with dementia (HR, 1.76; 95% CI, 1.26-2.46). The incidence rates for traumatic intracranial bleeding were 3.36 per 1000 person-years in patients with none of these risk factors (n = 9422) and 6.38 per 1000 person-years in those with at least 2 risk factors (n = 9271). Using logistic regression models to evaluate traumatic intracranial bleeding at 1 year and 3 years and retaining risk factors with P < .20 in adjusted analyses, we found that model discrimination was poor (C statistic, 0.55 at both 1 year and 3 years).

Secondary Outcomes

The incidence rate of any intracranial bleeding (traumatic or nontraumatic) was 14.58 per 1000 person-years. More than half (57.2%) of the intracranial bleeding events were nontraumatic. Among the 1317 patients who experienced any intracranial bleeding event, 407 (30.9%) had more than 1 episode. Over the same period of observation, the incidence rate of ischemic stroke was 13.44 per 1000 person-years.

Traumatic Intracranial Bleeding Based on Stroke Risk

After dividing our sample into 3 categories of stroke risk based on CHA2DS2-VASc scores (2, 3-4, or ≥5), more than one-third of patients (40.3%) fell into the category of very high stroke risk (CHA2DS2-VASc score, ≥5). The observed rate of ischemic stroke per 1000 person-years increased as the CHA2DS2-VASc score increased (rate of 4.59 for a score of 2, rate of 7.63 for a score of 3-4, and rate of 14.90 for a score of ≥5) (Figure 3). Conversely, across CHA2DS2-VASc scores, the observed rate of traumatic intracranial bleeding per 1000 person-years was constant (rate of 3.85 for a score of 2, rate of 4.81 for a score of 3-4, and rate of 5.00 for a score of ≥5).


To our knowledge, our study represents the largest investigation to date focused on traumatic intracranial bleeding among older adults (≥75 years) initiating warfarin therapy for AF. Among the 31 951 patients in our sample, we found that the rate of traumatic intracranial bleeding was 4.80 per 1000 person-years. As expected, this rate was considerably higher than that in previous clinical trials,38,39 but it was in line with a smaller population estimate from Medicare beneficiaries previously described by Gage et al.10 While we found that traumatic intracranial bleeding was a less common event than ischemic stroke, when we analyzed all intracranial bleeding events (traumatic and nontraumatic), the rates were similar (13.44 per 1000 person-years for ischemic stroke and 14.58 per 1000 person-years for any intracranial bleeding).

Identified Risk Factors for Traumatic Intracranial Bleeding

Several variables among those we analyzed placed patients at increased risk of traumatic intracranial bleeding. Most notably, the presence of dementia was associated with a doubling of risk, which was only mildly attenuated after adjusting for other factors. Prior studies40,41 have shown that patients with dementia are prescribed oral anticoagulants less frequently than patients without dementia, presumably due to concerns over harm or futility. Dementia is known from prior studies to increase fall risk considerably. In a cohort of 2015 older adult nursing home residents, the relative risk for dementia and falls was 1.93 after adjusting for other characteristics.28 While not all falls lead to traumatic intracranial bleeding, the presence of cerebral amyloid angiopathy in dementia subtypes (eg, Alzheimer disease) is known to increase bleeding risk,42 which may act synergistically with falls to predispose patients to traumatic intracranial bleeding. Other factors, such as difficulty following medication instructions or dietary noncompliance, may also have a role through excessive anticoagulation or increased INR variability. In this context, we believe that our findings further support the perceived increased risk of warfarin therapy in patients with dementia and can potentially be used in individualized discussions with patients and caregivers when considering the risk-benefit trade-off.

After multivariable adjustment, patients with labile INR in our cohort were also more likely to experience traumatic intracranial bleeding. This finding mirrors prior investigations linking labile INR with all-cause bleeding.29,35 For example, a study35 of patients enrolled in the SPORTIF (Stroke Prevention Using an Oral Thrombin Inhibitor in Atrial Fibrillation) III and V trials found that patients with labile INR were more than twice as likely to experience major bleeding as patients with more than 75% of values within therapeutic range. The high prevalence of labile INR among our population (55.9%) may be in part due to advanced age. Studies have shown that older patients have a greater anticoagulant response to warfarin despite administration of a lower weight-related dose43 and require a longer period to return to therapeutic range when their INR is supratherapeutic.44 In practice, there is considerable variability in the effectiveness of warfarin management between health systems, and labile INR therefore represents a potentially modifiable risk factor. The degree of therapeutic warfarin values in our sample was similar to that in an investigation by Rose et al,45 who studied 124 551 veterans receiving warfarin over a 2-year period and found a mean time in therapeutic range of 58%, with values ranging from 38% to 69% at different sites. Other reports have shown considerably better control of INR. For example, in the Swedish AuriculA registry, the mean time in therapeutic range among 18 391 patients was 76%.46

While the strength of association was lower for other independent risk factors, all have been shown in prior studies to be associated with either falls or bleeding. For example, depression may predispose patients to falling through daytime sleepiness, impaired attention, or psychomotor slowing.47 Furthermore, selective serotonin reuptake inhibitors, the most commonly used antidepressants, have been shown to increase bleeding risk.48 Anemia in older adults has been associated with a doubling of fall risk,26 which may be due to its association with fatigue and poor muscle strength, and has also been implicated as an independent risk factor for bleeding.49 The use of anticonvulsant medications has been associated with fall-related fractures, possibly due to adverse effects of dizziness or unsteady gait,50 as well as resistance to warfarin, which may make it challenging to maintain INRs in therapeutic range and place patients at risk for a rapid increase in INR if the anticonvulsant is discontinued for any reason.51

Comparative Risk of Ischemic Stroke

Despite anticoagulation with warfarin, the rate of ischemic stroke in our population (13.44 per 1000 person-years), was more than twice as high as the rate of traumatic intracranial bleeding. While this finding underscores the burden of stroke in patients 75 years or older with AF, the difference between our observed rates of ischemic stroke and traumatic intracranial bleeding was considerably less than that reported in prior studies.10,11,52 For example, in the previously noted study of Medicare beneficiaries by Gage et al,10 the incidence of ischemic stroke was almost 7 times as common as traumatic intracranial bleeding. As expected, we also found that the rate of ischemic stroke increased with higher CHA2DS2-VASc score. However, there was no concomitant increase in the rate of traumatic intracranial bleeding, which underscores that the risk factors for this outcome are unique. In addition, as previously noted, the rate of any intracranial bleeding (traumatic or nontraumatic) was similar to the rate of ischemic stroke. This observation highlights the difficult risk-benefit trade-off that health care professionals and older patients face when considering warfarin therapy.

Strength and Limitations

There are several strengths of our study, including our linkage of VA data with Medicare claims data to capture out-of-system hospitalizations as well as our ability to analyze variables that are unavailable in many other administrative data sets. However, our findings must also be interpreted within the limitations of our study design. First, we were unable to analyze fall risk factors, such as slow gait or balance impairment, which are generally not coded during routine clinic visits.24,25 In addition, we could not precisely characterize the severity of each risk factor, and certain risk factors of interest (eg, problem drug or alcohol use) may have been underidentified given our requirement that they were explicitly coded during inpatient or outpatient visits. Second, our study represents a single health care system with an almost entirely male population; therefore, our findings will need to be evaluated in other groups, including samples with a larger proportion of women (in whom risk factors may differ). Third, because we excluded patients who had infrequent INRs due to the potential for incomplete data, we may have excluded noncompliant patients with higher bleeding risk. Similarly, we excluded patients with AF who never received warfarin, and we cannot comment on bleeding rates within this subset. Fourth, we were unable to capture information on hospitalizations for patients enrolled in Medicare managed care. While only 23.1% of patients were enrolled in one of these plans for more than 1 year during the study period, we may still have underestimated the incidence rates of our outcomes. Fifth, our study period predated the common use of direct oral anticoagulants. These medications are increasingly used for stroke prevention in AF and may have a different risk-benefit profile from that of warfarin in older adults. For example, the reliable effect of these medications, taken regularly, would obviate the labile INR variable, which we found to be significantly associated with bleeding risk. Therefore, we believe that the comparative risk of direct oral anticoagulants vs warfarin, in the context of traumatic intracranial bleeding among older adults in clinical practice, warrants further investigation in future observational studies.


In summary, we found that the rate of traumatic intracranial bleeding among older adults with AF initiating warfarin therapy was higher than previously reported in clinical trials. After multivariable adjustment, several factors placed patients at increased risk of traumatic intracranial bleeding, including dementia, anemia, depression, anticonvulsant use, and labile INR. 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. Our findings should be validated in other data sets, particularly given the underrepresentation of women in our sample.

Back to top
Article Information

Accepted for Publication: December 15, 2015.

Corresponding Author: John A. Dodson, MD, MPH, Leon H. Charney Division of Cardiology, Department of Medicine, New York University School of Medicine, 227 E 30th St, Translational Research Bldg Room 851, New York, NY 10706 (

Published Online: March 9, 2016. doi:10.1001/jamacardio.2015.0345.

Author Contributions: Drs Dodson and Gaziano had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Dodson, Gagnon, Tinetti, Krumholz, Gaziano.

Acquisition, analysis, or interpretation of data: Dodson, Petrone, Gagnon, Gaziano.

Drafting of the manuscript: Dodson, Tinetti, Gaziano.

Critical revision of the manuscript for important intellectual content: Petrone, Gagnon, Krumholz.

Statistical analysis: Gagnon, Petrone.

Obtained funding: Dodson, Gaziano.

Administrative, technical, or material support: Gagnon, Gaziano.

Study supervision: Dodson, Gagnon, Tinetti, Krumholz, Gaziano.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Krumholz reported being a recipient of research agreements from Medtronic and from Johnson & Johnson (through Yale University) to develop methods of clinical trial data sharing and reported being chair of a cardiac scientific advisory board for UnitedHealthcare. No other disclosures were reported.

Funding/Support: This project was supported by grant R03AG045067 (Dr Dodson) 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). Dr Krumholz is supported by grant U01 HL105270-05 (Center for Cardiovascular Outcomes Research at Yale University) from the National Heart, Lung, and Blood Institute. Support for US Department of Veterans Affairs/Centers for Medicare & Medicaid Services data is provided by the Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development, Health Services Research and Development, Veterans Affairs Information Resource Center (project numbers SDR 02-237 and 98-004).

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.

Go  AS, Hylek  EM, Phillips  KA,  et al.  Prevalence of diagnosed atrial fibrillation in adults: national implications for rhythm management and stroke prevention: the Anticoagulation and Risk Factors in Atrial Fibrillation (ATRIA) Study.  JAMA. 2001;285(18):2370-2375.PubMedGoogle ScholarCrossref
Gage  BF, Waterman  AD, Shannon  W, Boechler  M, Rich  MW, Radford  MJ.  Validation of clinical classification schemes for predicting stroke: results from the National Registry of Atrial Fibrillation.  JAMA. 2001;285(22):2864-2870.PubMedGoogle ScholarCrossref
Antani  MR, Beyth  RJ, Covinsky  KE,  et al.  Failure to prescribe warfarin to patients with nonrheumatic atrial fibrillation.  J Gen Intern Med. 1996;11(12):713-720.PubMedGoogle ScholarCrossref
Beyth  RJ, Antani  MR, Covinsky  KE,  et al.  Why isn’t warfarin prescribed to patients with nonrheumatic atrial fibrillation?  J Gen Intern Med. 1996;11(12):721-728.PubMedGoogle ScholarCrossref
Glader  EL, Sjölander  M, Eriksson  M, Lundberg  M.  Persistent use of secondary preventive drugs declines rapidly during the first 2 years after stroke.  Stroke. 2010;41(2):397-401.PubMedGoogle ScholarCrossref
Sellers  MB, Newby  LK.  Atrial fibrillation, anticoagulation, fall risk, and outcomes in elderly patients.  Am Heart J. 2011;161(2):241-246.PubMedGoogle ScholarCrossref
Lotze  U, Liebetrau  J, Malsch  I,  et al.  Medical treatment of patients with atrial fibrillation aged over 80 years in daily clinical practice: influence of age and CHADS2 score.  Arch Gerontol Geriatr. 2010;50(1):36-41.PubMedGoogle ScholarCrossref
Monette  J, Gurwitz  JH, Rochon  PA, Avorn  J.  Physician attitudes concerning warfarin for stroke prevention in atrial fibrillation: results of a survey of long-term care practitioners.  J Am Geriatr Soc. 1997;45(9):1060-1065.PubMedGoogle ScholarCrossref
Dharmarajan  TS, Varma  S, Akkaladevi  S, Lebelt  AS, Norkus  EP.  To anticoagulate or not to anticoagulate? a common dilemma for the provider: physicians’ opinion poll based on a case study of an older long-term care facility resident with dementia and atrial fibrillation.  J Am Med Dir Assoc. 2006;7(1):23-28.PubMedGoogle ScholarCrossref
Gage  BF, Birman-Deych  E, Kerzner  R, Radford  MJ, Nilasena  DS, Rich  MW.  Incidence of intracranial hemorrhage in patients with atrial fibrillation who are prone to fall.  Am J Med. 2005;118(6):612-617.PubMedGoogle ScholarCrossref
Man-Son-Hing  M, Nichol  G, Lau  A, Laupacis  A.  Choosing antithrombotic therapy for elderly patients with atrial fibrillation who are at risk for falls.  Arch Intern Med. 1999;159(7):677-685.PubMedGoogle ScholarCrossref
Butler  D, Kowall  NW, Lawler  E, Michael Gaziano  J, Driver  JA.  Underuse of diagnostic codes for specific dementias in the Veterans Affairs New England healthcare system.  J Am Geriatr Soc. 2012;60(5):910-915.PubMedGoogle ScholarCrossref
Choma  NN, Griffin  MR, Kaltenbach  LA, Greevy  RA, Roumie  CL.  Blood pressure control among patients with hypertension and newly diagnosed diabetes.  Diabet Med. 2012;29(9):1126-1133.PubMedGoogle ScholarCrossref
Miller  DR, Safford  MM, Pogach  LM.  Who has diabetes? best estimates of diabetes prevalence in the Department of Veterans Affairs based on computerized patient data.  Diabetes Care. 2004;27(suppl 2):B10-B21.PubMedGoogle ScholarCrossref
Rosen  AK, Rivard  P, Zhao  S,  et al.  Evaluating the patient safety indicators: how well do they perform on Veterans Health Administration data?  Med Care. 2005;43(9):873-884.PubMedGoogle ScholarCrossref
Siegel  D, Lopez  J, Meier  J.  Antihypertensive medication adherence in the Department of Veterans Affairs.  Am J Med. 2007;120(1):26-32.PubMedGoogle ScholarCrossref
Shen  Y, Hendricks  A, Zhang  S, Kazis  LE.  VHA enrollees’ health care coverage and use of care.  Med Care Res Rev. 2003;60(2):253-267.PubMedGoogle ScholarCrossref
Hylek  EM, Evans-Molina  C, Shea  C, Henault  LE, Regan  S.  Major hemorrhage and tolerability of warfarin in the first year of therapy among elderly patients with atrial fibrillation.  Circulation. 2007;115(21):2689-2696.PubMedGoogle ScholarCrossref
Tinetti  ME, Baker  DI, King  M,  et al.  Effect of dissemination of evidence in reducing injuries from falls.  N Engl J Med. 2008;359(3):252-261.PubMedGoogle ScholarCrossref
Hartholt  KA, van der Velde  N, Looman  CW,  et al.  Trends in fall-related hospital admissions in older persons in the Netherlands.  Arch Intern Med. 2010;170(10):905-911.PubMedGoogle ScholarCrossref
Pieracci  FM, Eachempati  SR, Shou  J, Hydo  LJ, Barie  PS.  Use of long-term anticoagulation is associated with traumatic intracranial hemorrhage and subsequent mortality in elderly patients hospitalized after falls: analysis of the New York State Administrative Database.  J Trauma. 2007;63(3):519-524.PubMedGoogle ScholarCrossref
Tinetti  ME, Han  L, Lee  DS,  et al.  Antihypertensive medications and serious fall injuries in a nationally representative sample of older adults.  JAMA Intern Med. 2014;174(4):588-595.PubMedGoogle ScholarCrossref
Tinetti  ME, Speechley  M, Ginter  SF.  Risk factors for falls among elderly persons living in the community.  N Engl J Med. 1988;319(26):1701-1707.PubMedGoogle ScholarCrossref
Graafmans  WC, Ooms  ME, Hofstee  HM, Bezemer  PD, Bouter  LM, Lips  P.  Falls in the elderly: a prospective study of risk factors and risk profiles.  Am J Epidemiol. 1996;143(11):1129-1136.PubMedGoogle ScholarCrossref
Ganz  DA, Bao  Y, Shekelle  PG, Rubenstein  LZ.  Will my patient fall?  JAMA. 2007;297(1):77-86.PubMedGoogle ScholarCrossref
Penninx  BW, Pluijm  SM, Lips  P,  et al.  Late-life anemia is associated with increased risk of recurrent falls.  J Am Geriatr Soc. 2005;53(12):2106-2111.PubMedGoogle ScholarCrossref
Nevitt  MC, Cummings  SR, Kidd  S, Black  D.  Risk factors for recurrent nonsyncopal falls: a prospective study.  JAMA. 1989;261(18):2663-2668.PubMedGoogle ScholarCrossref
van Doorn  C, Gruber-Baldini  AL, Zimmerman  S,  et al; Epidemiology of Dementia in Nursing Homes Research Group.  Dementia as a risk factor for falls and fall injuries among nursing home residents.  J Am Geriatr Soc. 2003;51(9):1213-1218.PubMedGoogle ScholarCrossref
Lip  GY, Frison  L, Halperin  JL, Lane  DA.  Comparative validation of a novel risk score for predicting bleeding risk in anticoagulated patients with atrial fibrillation: the HAS-BLED (Hypertension, Abnormal Renal/Liver Function, Stroke, Bleeding History or Predisposition, Labile INR, Elderly, Drugs/Alcohol Concomitantly) score.  J Am Coll Cardiol. 2011;57(2):173-180.PubMedGoogle ScholarCrossref
Kokotailo  RA, Hill  MD.  Coding of stroke and stroke risk factors using International Classification of Diseases, revisions 9 and 10.  Stroke. 2005;36(8):1776-1781.PubMedGoogle ScholarCrossref
Buresly  K, Eisenberg  MJ, Zhang  X, Pilote  L.  Bleeding complications associated with combinations of aspirin, thienopyridine derivatives, and warfarin in elderly patients following acute myocardial infarction.  Arch Intern Med. 2005;165(7):784-789.PubMedGoogle ScholarCrossref
Jia  H, Damush  TM, Qin  H,  et al.  The impact of poststroke depression on healthcare use by veterans with acute stroke.  Stroke. 2006;37(11):2796-2801.PubMedGoogle ScholarCrossref
U.S. Department of Veterans Affairs. Pharmacy benefits management services. Accessed August 14, 2015.
Rosendaal  FR, Cannegieter  SC, van der Meer  FJ, Briët  E.  A method to determine the optimal intensity of oral anticoagulant therapy.  Thromb Haemost. 1993;69(3):236-239.PubMedGoogle Scholar
White  HD, Gruber  M, Feyzi  J,  et al.  Comparison of outcomes among patients randomized to warfarin therapy according to anticoagulant control: results from SPORTIF III and V.  Arch Intern Med. 2007;167(3):239-245.PubMedGoogle ScholarCrossref
Pisters  R, Lane  DA, Nieuwlaat  R, de Vos  CB, Crijns  HJ, Lip  GY.  A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation: the Euro Heart Survey.  Chest. 2010;138(5):1093-1100.PubMedGoogle ScholarCrossref
Lip  GY, Nieuwlaat  R, Pisters  R, Lane  DA, Crijns  HJ.  Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based approach: the Euro Heart Survey on atrial fibrillation.  Chest. 2010;137(2):263-272.PubMedGoogle ScholarCrossref
Gulløv  AL, Koefoed  BG, Petersen  P,  et al.  Fixed minidose warfarin and aspirin alone and in combination vs adjusted-dose warfarin for stroke prevention in atrial fibrillation: Second Copenhagen Atrial Fibrillation, Aspirin, and Anticoagulation Study.  Arch Intern Med. 1998;158(14):1513-1521.PubMedGoogle ScholarCrossref
Stroke Prevention in Atrial Fibrillation Investigators.  Adjusted-dose warfarin versus low-intensity, fixed-dose warfarin plus aspirin for high-risk patients with atrial fibrillation: Stroke Prevention in Atrial Fibrillation III randomised clinical trial.  Lancet. 1996;348(9028):633-638.PubMedGoogle ScholarCrossref
Sandhu  RK, Bakal  JA, Ezekowitz  JA, McAlister  FA.  Risk stratification schemes, anticoagulation use and outcomes: the risk–treatment paradox in patients with newly diagnosed non-valvular atrial fibrillation.  Heart. 2011;97(24):2046-2050.PubMedGoogle ScholarCrossref
Abdel-Latif  AK, Peng  X, Messinger-Rapport  BJ.  Predictors of anticoagulation prescription in nursing home residents with atrial fibrillation.  J Am Med Dir Assoc. 2005;6(2):128-131.PubMedGoogle ScholarCrossref
Yamada  M.  Cerebral amyloid angiopathy: emerging concepts.  J Stroke. 2015;17(1):17-30.PubMedGoogle ScholarCrossref
Shepherd  AM, Hewick  DS, Moreland  TA, Stevenson  IH.  Age as a determinant of sensitivity to warfarin.  Br J Clin Pharmacol. 1977;4(3):315-320.PubMedGoogle ScholarCrossref
Hylek  EM, Regan  S, Go  AS, Hughes  RA, Singer  DE, Skates  SJ.  Clinical predictors of prolonged delay in return of the international normalized ratio to within the therapeutic range after excessive anticoagulation with warfarin.  Ann Intern Med. 2001;135(6):393-400.PubMedGoogle ScholarCrossref
Rose  AJ, Hylek  EM, Ozonoff  A, Ash  AS, Reisman  JI, Berlowitz  DR.  Risk-adjusted percent time in therapeutic range as a quality indicator for outpatient oral anticoagulation: results of the Veterans Affairs Study to Improve Anticoagulation (VARIA).  Circ Cardiovasc Qual Outcomes. 2011;4(1):22-29.PubMedGoogle ScholarCrossref
Wieloch  M, Själander  A, Frykman  V, Rosenqvist  M, Eriksson  N, Svensson  PJ.  Anticoagulation control in Sweden: reports of time in therapeutic range, major bleeding, and thrombo-embolic complications from the national quality registry AuriculA.  Eur Heart J. 2011;32(18):2282-2289.PubMedGoogle ScholarCrossref
Iaboni  A, Flint  AJ.  The complex interplay of depression and falls in older adults: a clinical review.  Am J Geriatr Psychiatry. 2013;21(5):484-492.PubMedGoogle ScholarCrossref
Andrade  C, Sandarsh  S, Chethan  KB, Nagesh  KS.  Serotonin reuptake inhibitor antidepressants and abnormal bleeding: a review for clinicians and a reconsideration of mechanisms.  J Clin Psychiatry. 2010;71(12):1565-1575.PubMedGoogle ScholarCrossref
Dauerman  HL, Lessard  D, Yarzebski  J, Gore  JM, Goldberg  RJ.  Bleeding complications in patients with anemia and acute myocardial infarction.  Am J Cardiol. 2005;96(10):1379-1383.PubMedGoogle ScholarCrossref
Carbone  LD, Johnson  KC, Robbins  J,  et al.  Antiepileptic drug use, falls, fractures, and BMD in postmenopausal women: findings from the Women’s Health Initiative (WHI).  J Bone Miner Res. 2010;25(4):873-881.PubMedGoogle Scholar
Perucca  E.  Clinically relevant drug interactions with antiepileptic drugs.  Br J Clin Pharmacol. 2006;61(3):246-255.PubMedGoogle ScholarCrossref
Hart  RG, Diener  HC, Yang  S,  et al.  Intracranial hemorrhage in atrial fibrillation patients during anticoagulation with warfarin or dabigatran: the RE-LY trial.  Stroke. 2012;43(6):1511-1517.PubMedGoogle ScholarCrossref