Comparison of the Incidence of Major Bleeding With Rivaroxaban Use Among Nonvalvular Atrial Fibrillation Patients With Versus Without Diabetes Mellitus

Diabetes mellitus (DM) is a common co-morbidity in those with nonvalvular atrial fibrillation (NVAF). Most patients with DM and NVAF have a CHA2DS2-VASc score of ≥1 and should be considered for oral anticoagulation therapy for stroke prevention per treatment guidelines. The most important risk associated with anticoagulation is bleeding, which may be higher in those with NVAF plus DM. Our objective was to evaluate the incidence and characteristics of major bleeding (MB) in rivaroxaban users diagnosed with NVAF, further comparing those with DM versus those without DM, in a real-world clinical setting. Electronic medical records of >10 million patients from the Department of Defense Military Health System were queried to identify rivaroxaban users with NVAF over a 2.5-year period. Major bleeding–related hospitalization was identified by a validated case-finding algorithm. Patient characteristics, incidence and management of MB, and fatal outcomes were assessed by DM status. Of 44,793 rivaroxaban users with NVAF, 12,039 (26.9%) had DM, who were more likely men, younger, with more co-morbidity and higher CHA2DS2-VASc scores. Major bleeding incidence was higher among those with DM compared with those without, 3.68 (95% confidence interval CI 3.37 to 4.03) versus 2.51 (95% CI 2.34 to 2.69) per 100 person-years, and intracranial bleeding incidence was 0.19 (95% CI 0.13 to 0.28) versus 0.25 (95% CI 0.20 to 0.31) per 100 person-years. Fatal outcomes were rare for both cohorts, 0.09 per 100 person-years. In conclusion, in this post-marketing study of 44,793 rivaroxaban users with NVAF, patients with DM had more co-morbidities and higher incidence of MB compared with those without DM. Diabetes mellitus (DM) is a common co-morbidity in those with nonvalvular atrial fibrillation (NVAF). Most patients with DM and NVAF have a CHA2DS2-VASc score of ≥1 and should be considered for oral anticoagulation therapy for stroke prevention per treatment guidelines. The most important risk associated with anticoagulation is bleeding, which may be higher in those with NVAF plus DM. Our objective was to evaluate the incidence and characteristics of major bleeding (MB) in rivaroxaban users diagnosed with NVAF, further comparing those with DM versus those without DM, in a real-world clinical setting. Electronic medical records of >10 million patients from the Department of Defense Military Health System were queried to identify rivaroxaban users with NVAF over a 2.5-year period. Major bleeding–related hospitalization was identified by a validated case-finding algorithm. Patient characteristics, incidence and management of MB, and fatal outcomes were assessed by DM status. Of 44,793 rivaroxaban users with NVAF, 12,039 (26.9%) had DM, who were more likely men, younger, with more co-morbidity and higher CHA2DS2-VASc scores. Major bleeding incidence was higher among those with DM compared with those without, 3.68 (95% confidence interval CI 3.37 to 4.03) versus 2.51 (95% CI 2.34 to 2.69) per 100 person-years, and intracranial bleeding incidence was 0.19 (95% CI 0.13 to 0.28) versus 0.25 (95% CI 0.20 to 0.31) per 100 person-years. Fatal outcomes were rare for both cohorts, 0.09 per 100 person-years. In conclusion, in this post-marketing study of 44,793 rivaroxaban users with NVAF, patients with DM had more co-morbidities and higher incidence of MB compared with those without DM. Rivaroxaban is a direct oral anticoagulant with a rapid onset of administration and has been Food and Drug Administration approved for multiple indications, including for reduction of risk of stroke and systemic embolism in patients with NVAF. The present study objective was to evaluate the incidence, management, and outcomes of major bleeding (MB) in patients with NVAF receiving rivaroxaban, comparing patients with versus without diabetes mellitus (DM). This retrospective observational study used electronic medical records (EMRs) from the US Department of Defense (DoD) health care database. The DoD Military Health System (MHS) covers active and retired military service members and their families and is one of the largest electronic cradle-to-grave health care systems in the United States with nearly 10 million active beneficiaries.12012 MHS stakeholders' report.http://www.health.mil/Reference-Center/Reports?query=MHS+2012Google Scholar The MHS contains longitudinal EMRs that are continually updated and comprised inhospital services, outpatient visits, and clinical/medical and pharmacy data. The MHS is not linked with data from the Veterans Affairs (VA). The MHS and the VA are separate entities and provide care through health care systems predominantly exclusive of one another; therefore, this study does not contain data from the VA patient population. The patients in this study are insured through the MHS, although they are not required to use military medical facilities for care. Many patients use TRICARE, the insurance arm of the DoD, to obtain care in non-military (civilian) facilities. Regardless of whether care takes place in a military or civilian facility, all claims and related clinical information for each encounter are captured in the DoD MHS databases.2Tamayo S. Peacock W.F. Patel M. Sicignano N. Hopf K.P. Fields L.E. Sarich T. Wu S. Yannicelli D. Yuan Z. Characterizing major bleeding in patients with nonvalvular atrial fibrillation: a pharmacovigilance study of 27 467 patients taking rivaroxaban.Clin Cardiol. 2015; 38: 63-68Crossref PubMed Scopus (144) Google Scholar The 2.5-year observational period for this study was January 1, 2013, to June 30, 2015, and the study population comprised patients with a diagnosis of NVAF who were taking rivaroxaban. Study subjects were identified using relevant International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis and procedure codes and Common Procedure Terminology/Healthcare Common Procedure Coding System procedure codes listed in any available procedure field within any medical encounter record. Eligible subjects were identified by searching care that was delivered in both military and civilian facilities, across all settings (inpatient, outpatient, emergency departments, etc.) within the MHS. All patients meeting the definition for NVAF were included in the study, regardless of incident or prevalent rivaroxaban use, as long as the patient was identified as having NVAF before or concurrent with rivaroxaban use. Data were collected on NVAF rivaroxaban users with and without DM. The diagnosis of DM was defined as the presence of diagnosis code 250.xx within 6 months before the bleeding date for MB cases and end of study participation for patients without MB. Data regarding rivaroxaban exposure were also collected, including dose at time of bleeding and duration of rivaroxaban exposure. Medication data were collected using prescription dispensing information from the date of medication initiation. Rivaroxaban use was compared between patients with and without DM. Demographic and baseline characteristic data were collected on all participants. Baseline characteristics included age, gender, DM status, and co-morbid conditions of interest, including cardiovascular conditions and history of bleeding. All co-morbid conditions were collected by querying the EMRs for the relevant diagnosis codes. CHA2SD2-VASc and modified HAS-BLED (no INR values included) scores were calculated for each patient. Patient characteristics were compared between those with and without DM. The primary outcome of interest was MB, as defined by the Cunningham algorithm,3Cunningham A. Stein C.M. Chung C.P. Daugherty J.R. Smalley W.E. Ray W.A. An automated database case definition for serious bleeding related to oral anticoagulant use.Pharmacoepidemiol Drug Saf. 2011; 20: 560-566Crossref PubMed Scopus (249) Google Scholar a validated method that uses administrative data for identification of MB events that result in a hospitalization. The algorithm identifies MB events from a primary discharge diagnosis using ICD-9-CM diagnosis and ICD-9-CM/Common Procedure Terminology procedure codes. Identification of MB events was differentiated by bleeding site: gastrointestinal bleeding, intracranial bleeds, and bleeding at other sites. The use of bleeding diagnoses showed a positive predictive value of 89% to 99% in Cunningham's validation study, and this algorithm has been used in other clinical studies to identify serious bleeding events.4Graham D.J. Reichman M.E. Wernecke M. Zhang R. Southworth M.R. Levenson M. Sheu T.C. Mott K. Goulding M.R. Houstoun M. MaCurdy T.E. Worrall C. Kelman J.A. Cardiovascular, bleeding, and mortality risks in elderly Medicare patients treated with dabigatran or warfarin for nonvalvular atrial fibrillation.Circulation. 2015; 131: 157-164Crossref PubMed Scopus (560) Google Scholar, 5Staerk L. Lip G.Y. Olesen J.B. Fosbol E.L. Pallisgaard J.L. Bonde A.N. Gundlund A. Lindhardt T.B. Hansen M.L. Torp-Pedersen C. Gislason G.H. Stroke and recurrent haemorrhage associated with antithrombotic treatment after gastrointestinal bleeding in patients with atrial fibrillation: nationwide cohort study.BMJ. 2015; 351: h5876Crossref PubMed Scopus (111) Google Scholar, 6Kawai V.K. Cunningham A. Vear S.I. Van Driest S.L. Oginni A. Xu H. Jiang M. Li C. Denny J.C. Shaffer C. Bowton E. Gage B.F. Ray W.A. Roden D.M. Stein C.M. Genotype and risk of major bleeding during warfarin treatment.Pharmacogenomics. 2014; 15: 1973-1983Crossref PubMed Scopus (42) Google Scholar, 7Alonso A. Bengtson L.G. MacLehose R.F. Lutsey P.L. Chen L.Y. Lakshminarayan K. Intracranial hemorrhage mortality in atrial fibrillation patients treated with dabigatran or warfarin.Stroke. 2014; 45: 2286-2291Crossref PubMed Scopus (60) Google Scholar, 8Lauffenburger J.C. Rhoney D.H. Farley J.F. Gehi A.K. Fang G. Predictors of gastrointestinal bleeding among patients with atrial fibrillation after initiating dabigatran therapy.Pharmacotherapy. 2015; 35: 560-568Crossref PubMed Scopus (26) Google Scholar, 9Lane M.A. Zeringue A. McDonald J.R. Serious bleeding events due to warfarin and antibiotic co-prescription in a cohort of veterans.Am J Med. 2014; 127: 657-663 e652Abstract Full Text Full Text PDF PubMed Scopus (48) Google Scholar Major bleeding events were included in our analyses if they occurred during rivaroxaban exposure plus 7 days post-discontinuation. Patients were evaluated for MB throughout the study period until censored at the earliest occurrence of any of the following events: an MB event, death, loss of MHS eligibility, or end of study. Fatal outcomes, defined as deaths occurring during MB-related hospitalizations, were also evaluated. For patients who experienced an MB event, additional data points were collected on inpatient MB management. Baseline patient characteristics were evaluated between patients whom developed an MB versus who did not and further stratified by DM status. Means and SD were reported for continuous variables, whereas frequencies were reported for categorical variables. Differences in categorical variables were tested with the chi-square tests, and differences in continuous variables were tested with t tests, although no a priori hypothesis testing was planned. Given the large sample size of the study, any differences observed based on statistical testing should be interpreted within the proper context (e.g., clinical importance). Major bleeding incidences, patient characteristics, medication use, MB management, and patient outcomes were evaluated by DM status. The incidences for MB and fatal outcomes were calculated using a person-time approach: the number of patients with a first episode of MB divided by the rivaroxaban exposure time at risk, presented per 100 person-years. Incidences are reported with 95% confidence intervals (CIs). All analyses were performed using SAS, v9.4 (SAS Institute Inc., Cary, North Carolina). This Post-marketing Safety Surveillance study was funded by Janssen Scientific Affairs, LLC, and Bayer HealthCare. Health ResearchTx conducted the analyses. The research data were derived from an approved Naval Medical Center, Portsmouth, VA IRB protocol, and the research was conducted in compliance with federal and state laws, including the Health Insurance Portability and Accountability Act of 1996. We identified 44,793 rivaroxaban users with NVAF, of which 12,039 (26.9%) had DM. Regardless of bleeding status, those with DM were more likely men, younger, and had more co-morbidities and higher CHA2DS2-VASc scores (Table 1). Patients with DM appeared to have more co-morbid conditions and more frequent use of baseline medications compared with those without DM (Table 1). Across the overall NVAF cohort, the most common co-morbidities were hypertension, coronary heart disease, and heart failure. Those with DM who experienced MB had the highest prevalence of these conditions, 94.5%, 59.5%, and 46.8%, respectively (Table 2). In fact, patients who experienced MB tended to have more co-morbidities regardless of DM status (Table 2).Table 1Baseline patient characteristics of atrial fibrillation rivaroxaban users, comparing those with and without diabetes mellitusCharacteristicAtrial Fibrillation withDiabetesN=12,039Atrial Fibrillationwithout DiabetesN=32,754p-valueAge ∗Age is at time of MB., mean ±SD, (years)75.5 ±8.576.6 ±10.5<.0001Men7301 (60.6%)17,794 (54.3%)<.0001 Hemophilia2 (0.0%)4 (0.0%)0.6629 †Fisher's Exact test. Bleeding History44 (0.4%)69 (0.2%)0.0038 Ulcer166 (1.4%)246 (0.8%)<.0001 History of Seizures236 (2.0%)453 (1.4%)<.0001 Diagnosed Dementia942 (7.8%)1930 (5.9%)<.0001 Hepatic Disease885 (7.4%)1323 (4.0%)<.0001 Prior Ischemic Stroke684 (5.7%)1188 (3.6%)<.0001 Heart Failure3774 (31.4%)5518 (16.9%)<.0001 Previous Cerebrovascular Event2089 (17.4%)3935 (12.0%)<.0001 Hypertension10,548 (87.6%)19,619 (59.9%)<.0001 Renal Disease ‡Renal disease as defined using ICD-9 codes, reflecting potential renal function impairment. Creatinine clearance could not be assessed, since lab results were generally unavailable.3325 (27.6%)3596 (11.0%)<.0001 Coronary Heart Disease5523 (45.9%)8564 (26.2%)<.0001 Venous Thromboembolism723 (6.0%)1495 (4.6%)<.0001 Malignancy2171 (18.0%)5018 (15.3%)<.0001CHADS2 Scores Scores for patients with that patients with nonvalvular rivaroxaban with the For patients with the dose with the 10 atrial CHA2DS2-VASc heart hypertension, diabetes stroke or or systemic disease, MB major SD is at time of Exact Renal disease as defined using ICD-9 codes, reflecting potential renal function impairment. Creatinine clearance could not be assessed, since lab results were generally The for patients with that patients with nonvalvular rivaroxaban with the For patients with the dose with the in a patient characteristics of atrial fibrillation rivaroxaban users with and without diabetes comparing by major Fibrillation with Fibrillation without ∗Age is at time of MB., mean ±SD, (0.0%)4 Bleeding History of Diagnosed Hepatic (4.0%)<.0001 Prior Ischemic Heart Previous Cerebrovascular Renal Disease disease as defined using ICD-9 codes, reflecting potential renal function impairment. Creatinine clearance could not be assessed, since lab results were generally Coronary Heart Venous Scores Scores for patients with that patients with nonvalvular rivaroxaban with the For patients with the dose with the 10 CHA2DS2-VASc heart hypertension, diabetes stroke or or systemic disease, MB major SD is at time of Renal disease as defined using ICD-9 codes, reflecting potential renal function impairment. Creatinine clearance could not be assessed, since lab results were generally The for patients with that patients with nonvalvular rivaroxaban with the For patients with the dose with the in a atrial CHA2DS2-VASc heart hypertension, diabetes stroke or or systemic disease, MB major SD CHA2DS2-VASc heart hypertension, diabetes stroke or or systemic disease, MB major SD In the DM that experienced MB, a higher of subjects were the dose compared with the in the other The approved for rivaroxaban a dose of for patients with to that these subjects may have had of renal impairment. Major bleeding incidence was higher among those with DM compared with those without DM, 3.68 (95% CI 3.37 to 4.03) versus 2.51 per 100 (95% CI 2.34 to to bleeding were highest for bleeding of gastrointestinal regardless of DM status, (95% CI to per 100 for those with DM, and (95% CI to per 100 for those without DM 1). The incidence of intracranial bleeding was and the intracranial bleeding was in the DM compared with the 0.19 (95% CI 0.13 to 0.28) versus 0.25 (95% CI 0.20 to 0.31) per 100 person-years. of was between the days for the DM compared with days for the not Bleeding management were between the DM and with higher use of and in patients with versus without DM 2). Fatal outcomes during MB hospitalization were and the incidence was the in both and 0.09 per 100 person-years. Patients with DM were at the time of those without DM, mean of versus respectively (Table bleeding and fatal outcome data among rivaroxaban users with atrial by diabetes mellitus Patients MB incidence was calculated using person-time for the value time at for all first within the period study. per 100 (95% with Fatal (95% hospitalization for the MB between and at time of confidence MB major SD The MB incidence was calculated using person-time for the value time at for all first within the period hospitalization for the MB in a CI confidence MB major SD In this post-marketing observational study of 44,793 rivaroxaban users with NVAF, that the incidence of MB for those with DM was higher in those without DM, 3.68 versus 2.51 per 100 person-years, This between the is more compared with the MB reported in the Food and Drug Administration clinical for rivaroxaban for of Stroke and in Fibrillation which that the MB was in those with DM versus per 100 in those without S. Z. J.L. Patel M.R. G. and and of rivaroxaban in patients with diabetes and nonvalvular atrial fibrillation: the rivaroxaban direct compared with for prevention of stroke and embolism in atrial fibrillation Heart 2015; Full Text Full Text PDF PubMed Scopus Google Scholar our showed a of intracranial hemorrhage compared with the results from DM is a risk for a number of cardiovascular including and heart Zhang H. L. Clinical characteristics and of diabetes mellitus on outcomes in patients with nonvalvular atrial 2015; PubMed Scopus Google Scholar, a major risk for cardiovascular by the Diabetes the and the Diabetes the Institute of Diabetes and and and the Heart PubMed Scopus Google Scholar Diabetes is also a major risk for E.L. the and and PubMed Scopus Google Scholar, K. Diabetes mellitus as a risk for PubMed Scopus Google Scholar, E. Diabetes an risk for J PubMed Scopus Google Scholar In fact, heart disease and stroke are the of among with Institute of Diabetes and and Heart Disease and Scholar The of DM on the cardiovascular also to outcomes for those with cardiovascular Zhang H. L. Clinical characteristics and of diabetes mellitus on outcomes in patients with nonvalvular atrial 2015; PubMed Scopus Google Scholar Diabetes also the incidence of D. risk for atrial fibrillation in a The Heart PubMed Scopus Google Scholar Clinical associated with DM an in atrial H. D. J.B. of and on and differences in the Heart PubMed Scopus Google Scholar and both of which are related to risk of S. T. Diabetes and risk of atrial Med. PubMed Scopus Google Scholar, M. M. A. M.R. R. of and with of and PubMed Scopus Google Scholar is with an risk of and the risk is further in those with E. for stroke prevention in atrial fibrillation among J Diabetes Scopus Google Scholar patients with both conditions should be considered for anticoagulation The risk associated with oral anticoagulation therapy in patients is L. for stroke prevention in nonvalvular atrial fibrillation: and an updated Heart Google Scholar Our showed that the MB of 3.68 per 100 among patients with DM treated with rivaroxaban in this real-world population is to the of per 100 experienced by rivaroxaban patients with DM in the S. Z. J.L. Patel M.R. G. and and of rivaroxaban in patients with diabetes and nonvalvular atrial fibrillation: the rivaroxaban direct compared with for prevention of stroke and embolism in atrial fibrillation Heart 2015; Full Text Full Text PDF PubMed Scopus Google Scholar, M.R. G. G. J.L. J.F. Rivaroxaban versus warfarin in nonvalvular atrial J Med. 2011; PubMed Scopus Google Scholar is not to the results across these the from the provide of clinical the of rivaroxaban therapy in the real-world NVAF population. the prevalence of both and DM are Diabetes Data from the Diabetes Scholar, for Disease and of with diagnosed United Scholar more patients are likely to be anticoagulant therapy for stroke and is to the risks of bleeding in with more that nearly of our NVAF study population also had DM, the prevalence of this of co-morbidity is diabetes is a for outcomes, and an important predictive in both the and CHA2DS2-VASc is important to the overall risks of outcomes with diabetes versus the risks of In the population reported the risk of a fatal MB was 0.09 per 100 patient This mortality risk should be compared the and stroke risk associated with a CHA2DS2-VASc This that although the risk of stroke in those with a CHA2DS2-VASc of is per 100 L. M. Lip G.Y. of risk for stroke and bleeding in patients with atrial fibrillation: the Fibrillation Heart PubMed Scopus Google Scholar the risk of fatal MB anticoagulant therapy is 0.09 per 100 person-years. Our study had The analyses included the largest cohort to date of patients with NVAF to rivaroxaban therapy in clinical data from this study provide information rivaroxaban therapy to patients and who those The MHS database used for this a large and cohort, which the US population at large for and clinical characteristics, the results of this study the of patients within the MHS is S. N. H. the military health as a for health care Med. PubMed Scopus Google Scholar compared with of the US and in the United Scholar of elderly patients in the MHS likely more study of conditions prevalent in The study has This retrospective was based on data points that were collected for and pharmacy records the dispensing information the administration of the and the use of rivaroxaban to a bleeding a The definition of DM was based on diagnosis codes, not clinical as and their to MB, were not evaluated. the of management and potential on MB is within this population. is linked with risk of both DM and Institute of Diabetes and and Heart Disease and Scholar, and atrial fibrillation: a of the and Cardiol. 2015; Full Text Full Text PDF PubMed Scopus Google Scholar and the between and DM is S. The of to diabetes and of data from J PubMed Scopus Google Scholar In our study values were available for of therefore, the of was not available for analyses. The definition for MB in this study is not an with the clinical definition as the algorithm is and on the information available in the a the clinical definition of MB was defined as bleeding associated with a in of a of of or MB in a or a fatal M.R. G. G. J.L. J.F. Rivaroxaban versus warfarin in nonvalvular atrial J Med. 2011; PubMed Scopus Google Scholar MB be of of the is also the of for or relevant that required medical S. Peacock W.F. Patel M. Sicignano N. Hopf K.P. Fields L.E. Sarich T. Wu S. Yannicelli D. Yuan Z. Characterizing major bleeding in patients with nonvalvular atrial fibrillation: a pharmacovigilance study of 27 467 patients taking rivaroxaban.Clin Cardiol. 2015; 38: 63-68Crossref PubMed Scopus (144) Google Scholar that a of patients with bleeding were without receiving any The to the and Health for during the of this study. is the of the Safety Surveillance study. The was and by conducted the data and data analyses. the and and to each and to the All and to each and approved the Yuan is a of Janssen LLC, who of The other have no of interest to