Author + information
- Received February 28, 2017
- Accepted April 27, 2017
- Published online December 4, 2017.
- Björn Redfors, MD, PhDa,
- William A. Gray, MDb,
- Randall J. Lee, MDc,
- Kenneth A. Ellenbogen, MDd,
- Machaon Bonafede, PhDe and
- Ori Ben-Yehuda, MDa,f,∗ ()
- aClinical Trials Center, Cardiovascular Research Foundation, New York, New York
- bLankenau Medical Center, Wynnewood, Pennsylvania
- cUniversity of California San Francisco, San Francisco, California
- dVCU Pauley Heart Center, Virginia Commonwealth University Medical Center, Richmond, Virginia
- eTruven Health Analytics, Cambridge, Massachusetts
- fNewYork-Presbyterian Hospital/Columbia University Medical Center, New York, New York
- ↵∗Address for correspondence:
Dr Ori Ben-Yehuda, Cardiovascular Research Foundation, 1700 Broadway, 9th Floor, New York, New York 10019.
Objectives This study sought to determine how many patients with atrial fibrillation (AF) are not treated with oral anticoagulants (OAC) due to a high risk of bleeding and to characterize their risk of ischemic stroke in a real-world setting.
Background AF is associated with a 5-fold increased risk of ischemic stroke. OAC reduce the risk of stroke in patients with AF who do not have an increased bleeding risk, but no comparably effective treatment exists for patients with contraindications to OAC.
Methods We analyzed administrative claims data from individuals with commercial and Medicare supplemental health insurance in the United States. We selected patients with AF and a documented contraindication to OAC who were not treated with OAC. The primary endpoint was the occurrence of ischemic stroke, as calculated per patient year.
Results We identified 1,300,643 patients with AF claims, of which 43,248 had a contraindication event and remained OAC naive for at least 1 year or died in hospital. More than 80% of the patients had a CHA2DS2-VASc score of >1. The incidence of ischemic stroke was 4.1% in the overall OAC naive cohort and was more common with increasing CHA2DS2/CHA2DS2-VASc score. Hemorrhagic stroke occurred in 3.6%. For patients with previous intracranial bleeding, the incidences of ischemic and hemorrhagic stroke were 12.2% and 20.3%, respectively.
Conclusions Patients with AF who are not treated with OAC due to increased bleeding risk are common. These patients are at considerable risk of both ischemic and hemorrhagic events. A novel approach to stroke prophylaxis in this population is needed.
Atrial fibrillation (AF) affects 1 in 25 adults over the age of 60 and almost 1 in 10 over the age of 80 (1). The lifetime risk of developing AF for a 40-year-old person is between 20% and 25%, and is expected to increase over the next decades (2). AF is associated with a 5-fold increased risk of ischemic stroke and accounts for up to 15% of all strokes and as many as 30% of strokes in patients >80 years of age (3).
Oral anticoagulants (OAC) effectively reduce the risk of ischemic stroke in patients with AF (4). However, this benefit comes with an increased risk of bleeding, including disabling intracranial and major life-threatening bleeding. For patients who do not have an increased baseline risk of bleeding, several large multicenter randomized clinical trials have conclusively established a favorable risk:benefit ratio with both the traditional vitamin K antagonists OAC and the new factor Xa inhibitors apixaban, edoxaban, and rivaroxaban, and the thrombin inhibitor dabigatran (4–9). Although the results of these trials are directly applicable to many patients with AF, important subsets of patients were excluded from the trials.
Most important, these trials excluded all patients who had recently suffered a bleeding event or who were considered to have an increased bleeding risk (5–7). These patients often have a high risk of thrombotic events, including ischemic stroke, and would benefit substantially from an effective preventive treatment (10,11).
The aims of this study are to determine in a real-world population how many patients with AF have previously had sufficiently severe bleeding events or have sufficiently high bleeding risk that OAC are considered contraindicated by their treating physician, to characterize the risk of ischemic stroke and major bleeding in these patients, and to identify subgroups in which these risks are increased.
To study the rates of stroke and systemic embolism in patients with AF who were contraindicated for OAC, we analyzed administrative claims data from a large population of individuals with commercial and Medicare supplemental health insurance in the United States. Health plan enrollment data and detailed patient-level administrative claims for inpatient, outpatient, and pharmacy services from January 1, 2009, to December 31, 2013, were obtained from the multipayer Truven Health MarketScan Commercial and the Medicare Supplemental Research databases. As of 2011, the commercial database included records of nearly 40 million individuals with employer-sponsored insurance in the United States. The Medicare Supplemental Database included records of 3.4 million retirees with employer-sponsored supplemental Medicare coverage.
We selected patients that had at least 1 non diagnostic medical claim with an International Classification of Diseases-9th edition-Clinical Modification (ICD-9-CM) diagnosis code for AF (ICD-9-CM diagnosis code 427.31) and a registered OAC contraindication. The most recently registered of the two served as the study index event. The following contraindications to OAC were included: hemorrhagic stroke; intracranial bleeding; intraocular bleeding; significant bleeding from the gastrointestinal, genitourinary, or respiratory tract; any bleeding requiring surgical intervention; and aortic dissection or cerebral aneurysm with high risk of bleeding that was not repaired within 3 months.
Patients were required to have had ≥12 months of continuous enrollment after the index event, with the exception of patients who died in hospital. We excluded from the main analysis patients who had post-operative AF, patients who had a heart transplant procedure within 1 year before the index date, and patients who were under 18 years of age. We also excluded from the main analysis patients who resumed OAC or low-molecular-weight heparin treatment within 1 year (Figure 1). We used a pre-index period of up to 12 months to identify the baseline characteristics of the patients and a 12-month post-index period to measure study outcomes.
We compared patients who experienced an event (defined as any of the endpoints described) with patients who did not. We present event rates by CHA2DS2 and CHA2DS2-VASc scores and compare stroke risk in this OAC contraindicated patient cohort to the adjusted stroke risk for a general population of AF patients as reported in the medical literature (3,12–14). We performed separate analyses on the subgroup of patients for whom the contraindication for OAC was a previous intracerebral or intracranial hemorrhage. We also compared the bleeding risk in this OAC contraindicated patient cohort to the bleeding risk for a general population of OAC-treated patients as reported in the medical literature (15).
The primary study endpoint was the occurrence of ischemic stroke. Secondary endpoints were any stroke, hemorrhagic stroke, death from any cause, cardiovascular death, stroke-related death, and any major nonintracranial bleeding (defined as any bleeding that required transfusion or surgical intervention). We also present the incidence of the combined endpoints any stroke or systemic embolization and any stroke, systemic embolization, or death.
We present binary and categorical baseline variables as percentages and continuous variables as mean ± SD and compare groups of patients by unpaired, 2-sided Student t tests or chi-square test. We present incidence rates for each outcome as the number of patients with an event divided by total years of person-time follow-up for all patients.
Study population and patient characteristics
We identified 1,300,643 patients with ≥1 medical claim containing an ICD-9-CM diagnosis code for AF and without any evidence of an open heart surgery procedure within 30 days before or on the date of the AF diagnosis. Of these patients, 161,606 had at ≥1 OAC contraindication event and out of these, 79,843 patients had ≥12 months post-index continuous enrollment or died in hospital. Eight of these patients had a code for a heart transplant procedure and 24 were under the age of 18. Out of the remaining 79,809 patients, 43,251 patients had no evidence of OAC or low-molecular-weight heparin use during the 12-month post-index period and were included in the study (Figure 1).
Patient demographics and type of OAC contraindication are presented in Table 1. Among the patients with OAC contraindications, >4 of 5 had a CHA2DS2-VASc score of >1 and 42.9% had a CHA2DS2-VASc-score of ≥4 (Table 2). Patients who experienced an event were older than patients who did not and were more likely to previously have had an ischemic stroke, transient ischemic attack, or myocardial infarction. They were also more likely to be treated with beta-blockers, angiotensin-converting enzyme inhibitors, or angiotensin receptor blockers, or statins (Table 1). Patients with events were more likely to have high CHA2DS2 and CHA2DS2-VASc scores (Table 2).
Incidence of ischemic and hemorrhagic events
The incidence of ischemic stroke was 4.1% in the overall study cohort and was 12.2% among patients with a previous cerebral or intracranial hemorrhage. Hemorrhagic stroke was almost as common as ischemic stroke in this population, and occurred in 1 in 5 patients with previous intracranial or intracerebral hemorrhage (Table 3). The incidence of ischemic stroke increased with increasing CHA2DS2 or CHA2DS2-VASc scores in the overall study population (p < 0.001) and was consistent with current reference rates for a general population of AF patients (Figure 2). The risk of hemorrhagic stroke also increased with increasing CHA2DS2/CHA2DS2-VASc, but the relationship was less steep than for ischemic stroke (Table 4). For patients with previous intracranial or intracerebral hemorrhage, the incidence of stroke seemed to be unrelated to CHA2DS2 or CHA2DS2-VASc score (Figure 3, Table 5). The risk of major bleeding also increased with increasing CHA2DS2-VASc and was considerably higher than in an OAC-treated general population of AF patients (Figure 4).
The risk of dying within 1 year was 12.7% for the overall patient cohort and 23.0% for patients with history of intracerebral or intracranial hemorrhage (Table 3).
To our knowledge, this study represents the largest cohort of OAC-naive patients with AF who are not prescribed an OAC in the context of a previous history of major bleeding or risk of increased bleeding. The most important findings are that a considerable number of patients with AF have a contraindication to OAC treatment and that these patients are at a substantial risk of ischemic stroke.
Almost 1 in 8 patients with medical claims for AF had an accepted contraindication for OAC treatment. All these patients fall outside the inclusion criteria of the OAC randomized multicenter clinical trials, with resulting uncertainty as to the optimal therapeutic approach for stroke and bleeding prevention (3–8,16). Our study shows that a considerable number of patients with AF do not receive any anticoagulation therapy, although the majority of them have CHA2DS2 and CHA2DS2-VASc scores consistent with a high risk of ischemic stroke (3,16). We show that these risk scores are predictive of the stroke risk in this study population and that the risk of thromboembolic events is considerable. Bleeding risk was considerably higher in these OAC-naive patients than in the general OAC-treated population. These patients are at considerable risk of life-threatening or disabling bleeding, even in the absence of OAC, and this bleeding risk increases with increasing CHA2DS2/CHA2DS2-VASc scores. Our findings are consistent with previous reports that stroke risk and bleeding risk are closely related, and point to an unmet need for effective stroke-reducing therapeutic options in this population (10,17). Carefully conducted studies of treatment alternatives for these patients are needed.
The risk of both bleeding complications and thromboembolic events were particularly high in patients who had previously had a hemorrhagic stroke. In contrast with the overall study cohort, these patients’ risk does not seem to be predicted by CHA2DS2/CHA2DS2-VASc scores, which further complicates clinical decision making. It is possible that this is a distinct group of AF patients for whom the relationship between the traditional predictors of thromboembolic events and serious bleeding differ from the general AF population (18). Regardless, our data show that this subset of patients would benefit greatly from alternative methods of stroke prevention that do not impose an increased bleeding risk.
Our study is retrospective and based on claims data. Although the endpoints we assessed are likely to be captured in claims forms due to their seriousness, we cannot exclude the possibility that some endpoints were not captured, and we do not have specific details about the nonfatal clinical consequences of each event. Second, we did not have detailed data on previous OAC use or on indications and/or contraindications for concomitant therapies such as statins, angiotensin-converting enzyme inhibitors, and beta-blockers. Whether the observed differences in concomitant medications between patients who had versus did not have an event constitute an important finding or an epiphenomenon would need to be explored in other databases. Third, we did not have data for patients who suffered an out-of-hospital death. Last, we did not have data on the patients’ HAS-BLED score. The HAS-BLED score has been shown to superior to the CHADS and CHADS-VASc scores at predicting bleeding risk among patients with AF (15).
A considerable number of patients with AF are not treated with OAC because of a bleeding-related contraindication. These patients are at considerable risk of both ischemic and hemorrhagic events, as well as death and these risks are particularly high in patients with previous intracranial hemorrhage. Although both the CHA2DS2 and CHA2DS2-VASc risk scores are predictive for the overall population, patients with a history of intracranial hemorrhage were at uniformly high risk. Novel approaches to stroke prophylaxis in this population are needed.
COMPETENCY IN MEDICAL KNOWLEDGE: Patients with AF who are not treated with an OAC due to a bleeding-related contraindication are common and are at considerable risk of both ischemic and hemorrhagic events. Patients with previous intracranial hemorrhage are at particularly high risk.
COMPETENCY IN PATIENT CARE: Alternative methods for stroke prevention that do not increase bleeding risk should be considered for these patients, particularly if they have previously suffered an intracranial hemorrhage.
TRANSLATIONAL OUTLOOK: Carefully conducted studies of treatment alternatives for patients with AF and bleeding-related contraindications for OAC are needed.
This study was funded by Sentreheart, Inc. Dr. Lee is a consultant and equity holder for Sentreheart, Inc. Dr. Ellenbogen has received honoraria and a research grant, and acts as a consultant for Boston Science. Dr. Bonafede is an employee of Truven Health Analytics, an IBM Company. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
All authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.
- Abbreviations and Acronyms
- atrial fibrillation
- International Classification of Diseases-9th edition-Clinical Modification
- oral anticoagulants
- Received February 28, 2017.
- Accepted April 27, 2017.
- 2017 American College of Cardiology Foundation
- Lloyd-Jones D.M.,
- Wang T.J.,
- Leip E.P.,
- et al.
- January C.T.,
- Wann L.S.,
- Alpert J.S.,
- et al.
- Lip G.Y.,
- Frison L.,
- Halperin J.L.,
- Lane D.A.
- Roldan V.,
- Marin F.,
- Manzano-Fernandez S.,
- et al.
- Camm A.J.,
- Lip G.Y.,
- De Caterina R.,
- et al.
- Hylek E.M.,
- Evans-Molina C.,
- Shea C.,
- Henault L.E.,
- Regan S.