Prior Antithrombotic Use Is Associated With Favorable Mortality and Functional Outcomes in Acute Ischemic Stroke
Background and Purpose—Antithrombotics are the mainstay of treatment in primary and secondary prevention of stroke, and their use before an acute event may be associated with better outcomes.
Methods—Using data from Get With The Guidelines-Stroke with over half a million acute ischemic strokes recorded between October 2011 and March 2014 (n=540 993) from 1661 hospitals across the United States, we examined the unadjusted and adjusted associations between previous antithrombotic use and clinical outcomes.
Results—There were 250 104 (46%) stroke patients not receiving any antithrombotic before stroke; of whom approximately one third had a documented previous vascular indication. After controlling for clinical and hospital factors, patients who were receiving antithrombotics before stroke had better outcomes than those who did not, regardless of whether a previous vascular indication was present or not: adjusted odds ratio (95% confidence intervals) were 0.82 (0.80–0.84) for in-hospital mortality, 1.18 (1.16–1.19) for home as the discharge destination, 1.15 (1.13–1.16) for independent ambulatory status at discharge, and 1.15 (1.12–1.17) for discharge modified Rankin Scale score of 0 or 1.
Conclusions—Previous antithrombotic therapy was independently associated with improved clinical outcomes after acute ischemic stroke. Ensuring the use of antithrombotics in appropriate patient populations may be associated with benefits beyond stroke prevention.
Even though the use of antithrombotic medications for primary and secondary prevention of cardiovascular disease is increasing, many patients with indications are still not receiving antithrombotic medications and have an acute ischemic stroke.1 It also remains unclear whether previous therapy can also improve outcomes from those still having an acute ischemic stroke. Possible mechanisms for such a benefit include: attenuating the volume of the initial thrombus, preventing clot propagation, and reducing the risk of early recurrent thrombosis or embolism.
Despite theoretical mechanisms for benefit, existing evidence on the topic is conflicting. Kwok et al2 found that previous antithrombotic use was not associated with reduced mortality up to 1 year after stroke presentation. In contrast, a large registry from Canada reported a beneficial association between previous use of antithrombotics and improved functional outcome.3,4 Indeed, recent studies found a reduction in initial stroke severity in previous antiplatelet users in ischemic stroke,5,6 suggesting that previous antithrombotic therapy may moderate ischemic stroke evolution from the earliest moments of onset. To date, however, all these studies were small or moderate in size, and some of these conflicts may be because of unstable estimates.
Using data from the American Heart Association/American Stroke Association Get With The Guidelines-Stroke database, our study aims were to (1) describe characteristics of ischemic stroke patients by receipt or nonreceipt of antithrombotic medication before stroke; (2) determine whether prestroke antithrombotic use is related to outcomes at discharge, and whether this relationship varies with indication for antithrombotic use; and (3) determine whether previous warfarin use is associated with outcomes at discharge among patients with atrial fibrillation (AF) or flutter, taking international normalized ratio (INR) control into account.
The American Heart Association/American Stroke Association Get With The Guidelines-Stroke database data collection methods have been previously described.7–10 In brief, 1661 hospitals used an Internet-based Patient Management Tool (Quintiles, Cambridge, MA) to enter data, receive decision support, and obtain feedback via on demand reports of performance on quality measures and recorded data from consecutive admissions for acute ischemic stroke. There were a total of 624 883 patients with ischemic stroke at 1705 participating centers between October 1, 2011, and March 31, 2014. Of them, 19 381 were transferred to another acute facility, left against medical advice, or had no data on discharge status; 63 738 had missing data on previous antithrombotic use, and further 771 patients were excluded because of data on vascular indication for the use of antithrombotics was missing.
Trained hospital personnel abstracted data using the Internet-based Patient Management Tool with standardized data definitions and detailed coding instructions. The Internet-based system performs checks to ensure that the reported data are complete and internally consistent. In addition, data quality is monitored for both completeness and accuracy. Hospitals that participate must receive approval through their local institutional review boards or a waiver of individual consent under the common rule. Quintiles is the data collection coordination center for the American Heart Association/American Stroke Association Get With The Guidelines programs. The Duke Clinical Research Institute (Durham, NC) serves as the data analysis center. Hospital characteristics (ie, academic teaching status, bed size) were based on American Hospital Association data.10 Past medical history was defined on the basis of preexisting conditions, with the exclusion of conditions that were newly diagnosed during the hospital stay.
Previous antithrombotic use was defined as any anticoagulant or antiplatelet use before the index stroke. Patients were considered to have a vascular indication for antithrombotic use if their medical history included coronary artery disease, previous stroke or transient ischemic attack, or AF or flutter. Our study examined patient-relevant outcomes of in-hospital mortality, discharge to home, ability to ambulate independently at discharge, disability at the time of discharged defined using modified Rankin Scale score of ≥2 and acute hospital length of stay (LOS).
Analyses were performed using SAS version 9.3 or higher (SAS Institute, Cary NC). We compared the baseline characteristics for patients by (1) previous use of antithrombotic (yes versus no) and (2) previous use of antithrombotic (yes versus no) and indication (any versus none; 4 groups). Differences are compared with Pearson χ2 tests for categorical and Wilcoxon rank-sum and Kruskal–Wallis tests for continuous variables, and data are presented descriptively.
To evaluate associations, multivariable logistic regression was used for binary outcomes. Multiple regression was used for LOS, which was transformed using the natural log to achieve approximate normality. In addition to the term for previous antithrombotic use, each model also contained a term for antithrombotic indication and a term for the interaction between them. Models were adjusted for covariates at admission including age, sex, race, body mass index, medical history, on-hours arrival, and site characteristics. Missing values of covariates were imputed using multiple imputation (25 imputations). Generalized estimating equations were used to account for clustering within hospitals.
We present previous antithrombotic odds ratios within indication subgroups if the interaction between antithrombotic use and indication is significant, and if not, present separate antithrombotic and indication odds ratios (ie, main effects). For LOS outcome, the interaction between previous antithrombotic use and indication was significant, and thus the ratio of expected LOS for the 2 groups is reported, ie, the data are presented as the ratio of expected LOS in the first group compared with the second, that is, (expected days in group 1)/(expected days in group 2). Therefore, for LOS outcome, risk relationships are shown for each variable within levels of the other variable.
Because National Institutes of Health Stroke Scale (NIHSS) is missing in a proportion of patients (23%), a sensitivity analysis was performed in which these models were repeated, including NIHSS as a covariate, in the subset of patients with available NIHSS data.
Persistent or paroxysmal AF/flutter during the index admission and previous medical history of AF/flutter were used to define AF/flutter in this study. To determine whether previous warfarin use is associated with outcomes among patients with AF/flutter, taking INR control into account, patients were grouped as (1) previous use of anticoagulants with INR >1.4 at admission, (2) previous use of anticoagulants with INR ≤1.4 at admission, and (3) no anticoagulants before admission. Patients who were on an anticoagulant but who do not have INR data were excluded from this analysis. Models used were the same as for the main analysis, except that the 3 warfarin groups were included in the model in place of the antithrombotic and indication terms, an additional term was added for nonanticoagulant antithrombotics (eg, aspirin and other antiplatelet drugs), and patients taking new oral anticoagulants were excluded. Each warfarin group was compared with the no-warfarin group (reference group).
A total of 540 993 patients at 1661 sites admitted with an ischemic stroke during the study period were included in the current study. Over half (53.8%) were receiving an antithrombotic agent (either an antiplatelet/combination or an anticoagulant); 253 552 (46.9%) were taking antiplatelet drugs and 57 543 (10.6%) were taking an anticoagulant (the sum total is >53.8% because some patients were taking both; inclusion and exclusion is available in the Figure).
The characteristics of all patients included in the analyses and then separately for those who received antithrombotics before the index ischemic stroke and those who did not are shown in Table 1. With large numbers, the P values are highly significant between the 2 populations. People who did not take any antithrombotic before stroke were younger, more likely to be women, less likely to be white, more likely to have abnormal lipid profile, more likely to be a current smoker, but with lower prevalence of comorbid medical conditions including previous history of stroke and AF/flutter, and were more likely to be ambulatory independently before stroke. Although many of the acute biochemistry and hematological parameters, and site characteristics, were statistically significantly different between these groups, the magnitudes of differences were negligible.
Sample characteristics comparison by previous antithrombotic use and indication for its use shows prestroke antithrombotic users were older and more likely to be white. Patients without prestroke antithrombotic use despite an indication show similar characteristics to users who did not have previous cardiovascular diseases. The total number of the subgroups analyzed by the presence or absence of previous indications for antithrombotic use and exclusions are presented in the Figure. Approximately 30% of patients had mismatch between vascular indication and usage in this cohort. Table I in the online-only Data Supplement shows the types of antithrombotic use among the users of antithrombotics before the index stroke.
Table 2 shows the regression model results (adjusted for variables described as in Methods section of this article), respectively, by previous antithrombotic use and indication. Although crude rates show poor outcome in those with previous use, regression models (with full adjustment) demonstrate that patients taking an antithrombotic before the index stroke were less likely to die during the hospitalization and more likely to be discharged to home, able to ambulate independently, and better functional outcomes at discharge compared with patients who were not on an antithrombotic. Table II in the online-only Data Supplement shows that patients with a vascular indication for antithrombotic use (medical history of coronary artery disease, previous stroke or transient ischemic attack, or AF or flutter) were more likely to have unfavorable/negative outcomes than patients without an indication. There were no significant interactions between antithrombotic use and previous indication for use, indicating that the lower rate of in-hospital death with antithrombotic use was similar for patients with and without a vascular indication for use. Discharge outcome to home among patients discharged alive, and ambulatory status among patients able to ambulate independently before the event, showed results that were consistent with the models in the larger set of patients.
LOS outcome analysis shows that patients taking an antithrombotic before stroke had a shorter expected hospital stay than patients not on an antithrombotic before the event. The significant interaction indicates that the expected relative LOS depends on whether the patient had an indication—there is more of an associated reduction in LOS, with compared with without antithrombotic, where there is an indication (Table 2).
Tables III and IV in the online-only Data Supplement show the results in a subset of patients in whom NIHSS was available with additional adjustment for NIHSS score. Results are generally consistent with those in the full cohort, except for LOS, which here does not have a significant interaction between antithrombotic and the presence or absence of an indication for its use.
Of 118 635 (22.1%) identified as AF/flutter, 29 425/118 636 (24.8%) had no previous medical history of AF/flutter but had either persistent or paroxysmal AF/flutter during the index admission (Table V in the online-only Data Supplement). Table 3 demonstrates that patients on warfarin with INR>1.4 had better outcomes than patients not on warfarin. Patients on warfarin with INR ≤1.4 had a higher risk of in-hospital death, lower likelihood of being discharged home, and a longer length of hospital stay, compared with patients not on warfarin before stroke, but a similar probability of being able to ambulate independently and having a low modified Rankin Scale score at the time of discharge.
To the best of our knowledge, this study is the largest to examine the association between previous antithrombotic use and important and relevant outcomes in patients admitted with an acute ischemic stroke. We found that the previous use of antithrombotics was associated with a favorable outcome for all outcomes assessed highlighting an important point that antithrombotics not only have effect on vascular outcome but also may reduce the severity of vascular outcomes. Associations between previous antithrombotic medication use and better poststroke outcomes were seen across patient subgroups. No significant interaction between antithrombotic use and indication was observed except for LOS outcome.
To date, the literature around the impact of previous antithrombotic use on stroke outcome has been inconsistent and shown conflicting results. However, they were limited by relatively small sample size (Sanossian et al,11 n=260; Vibo et al,12 n=433), or focused on mortality alone,2,12 or in a particular patient population or of certain age,13 or examined the stroke severity only.14 Relatively larger studies again showed conflicting results.2–6,15 The key advantages of our study are much larger sample size and analysis of diverse relevant and important outcomes for patients.
These findings may have important clinical implications. It appears that in US setting, a substantial proportion of patients who sustained ischemic stroke (78 465/540 993=14.5%) would have potentially benefited from antithrombotic use because a secondary preventive measure was not receiving antithrombotic agents at the time of index stroke onset. Similarly, we found antithrombotic use before stroke without a documented vascular indication in a substantial proportion of patients (Table II in the online-only Data Supplement). This use may be because of other appropriate indications but could also reflect self-medication, which would be concerning because of the potential harm from drug side effects.
White et al1 highlighted similar mismatches and the fact that although half of strokes cases were on antithrombotics and yet developed stroke, half of stroke patients might have been identified as high risk and been prescribed an antithrombotic medication that would have prevented a substantial number of stroke events. This, in combination with our current study findings, further strengthens the argument to base antithrombotic medication use on improved risk prediction scores. Indeed, Loke et al16 recently highlighted the lack of sensitivity of existing cardiovascular risk prediction tools in reliably identifying those groups of patients who are most likely to subsequently develop cardiovascular adverse events.17 Most recent guidelines from United States,18 Europe,19 United Kingdom,20 and ATP III21 mainly focused on cardiovascular risk factors and the use of antiplatelets in primary prevention was less well focused perhaps because of presumed adherence to established guidance.
The strengths of our study include the large sample size and prospective data collection. One of the key strengths of the article includes the robust statistical analysis with ability to control for potential confounders and ability to understand the confounding effect by indication through analysis of indication for antithrombotic versus their effect on the outcomes examined. We were able to examine the outcomes by previous antithrombotic use and by vascular indication. We were also able to examine the outcomes by previous warfarin use and INR among patients with AF/flutter. Furthermore, we were able to analyze the data taken into account of the stroke severity at onset (eg, NIHSS score) in this current report.
Our study has some limitations. As a hospital-based registry, some cases of stroke might not have been included, such as patients who died before admission. Patients and hospitals may not be entirely representative of the US population, but the sample population is comparable to all US patients hospitalized with stroke.22 Residual measured and unmeasured confounders may account for some of these findings such as factors that could have influenced previous use of antithrombotic therapy and adherence to prescribed therapy. For example, those who use antithrombotic agents may have a healthy user effect or greater use of health care. Although a substantial proportion of patients had missing data on NIHSS and that was because of nonrandom missing, repeating the analyses in those who had NIHSS data yielded the similar results. As an observational study, the causality cannot be assumed. Nonetheless, the observed associations have plausible explanations, as we alluded in the beginning of this article.
In summary, using the American Heart Association/American Stroke Association Get With The Guidelines-Stroke registry data with over half a million of patients with ischemic stroke, we present evidence that previous antithrombotic use was associated with favorable outcomes in ischemic stroke and thus highlights the importance of primary and secondary stroke preventions with antithrombotic medications when indicated. A substantial proportion of patients in our study sustained stroke despite being on antithrombotic agents and the fact that patients with INR <1.4 among people with AF/flutter had worse outcomes suggests the urgent need to address issues of medication compliance and adequate anticoagulation in stroke prevention. Ensuring appropriate use of antithrombotics at a population level may have substantial benefit to patients with stroke and health economy in a global scale.
Drs Myint and Smith conceived the idea and developed the analysis plan with critical input from coauthors. A.S. Hellkamp analyzed the data. Drs Myint and Smith drafted the article with input from all coauthors. All authors contributed in interpretation of results and in making an important intellectual contribution to the article.
Sources of Funding
The Get With The Guidelines (GWTG)-Stroke program is currently supported, in part, by a charitable contribution from Bristol-Myers Squibb/Sanofi Pharmaceutical Partnership and the American Heart Association Pharmaceutical Roundtable. GWTG-Stroke has been funded in the past through support from Boehringer-Ingelheim and Merck. These funding agencies did not participate in design or analysis, article preparation, or approval of this study.
Dr Bhatt discloses the following relationships: Advisory Board: Cardax, Elsevier Practice Update Cardiology, Medscape Cardiology, Regado Biosciences; Board of Directors: Boston VA Research Institute, Society of Cardiovascular Patient Care; Chair: American Heart Association Quality Oversight Committee; Data Monitoring Committees: Duke Clinical Research Institute, Harvard Clinical Research Institute, Mayo Clinic, Population Health Research Institute; Honoraria: American College of Cardiology (Senior Associate Editor, Clinical Trials and News, ACC.org), Belvoir Publications (Editor in Chief, Harvard Heart Letter), Duke Clinical Research Institute (clinical trial steering committees), Harvard Clinical Research Institute (clinical trial steering committee), HMP Communications (Editor in Chief, Journal of Invasive Cardiology), Journal of the American College of Cardiology (Guest Editor; Associate Editor), Population Health Research Institute (clinical trial steering committee), Slack Publications (Chief Medical Editor, Cardiology Today’s Intervention), Society of Cardiovascular Patient Care (Secretary/Treasurer), WebMD (CME steering committees); other: Clinical Cardiology (Deputy Editor), National Cardiovascular Data Registry (NCDR)-ACTION Registry Steering Committee (Vice-Chair), Veterans Affairs Clinical Assessment, Reporting, and Tracking (VA CART) Research and Publications Committee (Chair); Research Funding: Amarin, AstraZeneca, Bristol-Myers Squibb, Eisai, Ethicon, Forest Laboratories, Ischemix, Medtronic, Pfizer, Roche, Sanofi Aventis, The Medicines Company; Royalties: Elsevier (Editor, Cardiovascular Intervention: A Companion to Braunwald’s Heart Disease); Site Co-Investigator: Biotronik, Boston Scientific, St. Jude Medical; Trustee: American College of Cardiology; Unfunded Research: FlowCo, PLx Pharma, Takeda. Dr Fonarow is a member of the Get With The Guidelines (GWTG) Executive Committee, has served as a consultant to Janssen (modest), receives research support from Patient-Centered Outcomes Research Institute (PCORI) (significant), and is an employee of UCLA, which holds a patent on retriever devices for stroke. Dr Myint has received support from Viforpharma to attend an advisory meeting. Dr Peterson has received research funding from Janssen Pharmaceuticals, Sanofi-Aventis, Genentech, Daiichi Sankyo, Eli Lilly and Astra Zeneca. He has served as a consultant for Boehringer Ingelheim, Astra Zeneca, and Janssen. Dr Peterson is also the principal investigator of the American Heart Association (AHA) Data Coordinating Centre at Duke Clinical Research Institute. Dr Reeves has received salary support from the Michigan Stroke GWTG Registry and is a member of the AHA’s GWTG Quality Improvement Subcommittee and Stroke Science Subcommittee. Dr Saver is a member of the GWTG-Stroke Science Subcommittee; is an employee of the University of California, which received funding for his scientific consulting services for trial design and conduct to Medtronic, Stryker, Neuravia, Boehringer Ingelheim (prevention only), Mitsubishi, and St. Jude Medical; and is an employee of the University of California, which holds a patent on retriever devices for stroke. Dr Schwamm is Chair of the GWTG Clinical Workgroup; serves as a consultant to the Massachusetts Department of Public Health, and as a member of the steering committee for the VICTORY AF (Evaluation of the Phased Radio Frequency Ablation System) and REACT AF (Real Evidence of Anticoagulation Treatment in AF) trials of stroke prevention in AF (Medtronic). Dr Smith is a member of the American Heart Association Get With The Guidelines Steering Committee; has received research support from the National Institutes of Health (National Institute of Neurological Disorders and Stroke R01NS062028), the Canadian Stroke Network, and the Canadian Institute for Health Research; has received speaking or writing fees from QuantiaMD, the BMJ Group, and the Canadian Conference on Dementia; and has served on an advisory board to Genentech. Dr Suter is used as Vice-President of Quality and Health IT by the American Heart Association/American Stroke Association, which has received grant support from numerous entities including Genentech, Boehringer Ingelheim, Janssen Pharmaceuticals, Daiichi Sankyo, Medtronic, Bristol-Myers Squibb, and Astra Zeneca; he has previously served as a consultant for Genentech and Daiichi Sankyo. The other authors report no conflicts.
Current address for P.J.S. Mayo Clinic.
Guest Editor for this article was Gregory W. Albers, MD.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.115.012414/-/DC1.
- Received December 11, 2015.
- Revision received May 17, 2016.
- Accepted May 23, 2016.
- © 2016 American Heart Association, Inc.
- White JR,
- Bettencourt-Silva JH,
- Potter JF,
- Loke YK,
- Myint PK
- Kwok CS,
- Skinner J,
- Metcalf AK,
- Potter JF,
- Myint PK
- Dowlatshahi D,
- Hakim A,
- Fang J,
- Sharma M
- Saposnik G,
- Hill MD,
- O’Donnell M,
- Fang J,
- Hachinski V,
- Kapral MK
- Kim WJ,
- Ko Y,
- Yang MH,
- Im SH,
- Park JH,
- Lee J,
- et al
- Schwamm LH,
- Fonarow GC,
- Reeves MJ,
- Pan W,
- Frankel MR,
- Smith EE,
- et al
- Smith EE,
- Shobha N,
- Dai D,
- Olson DM,
- Reeves MJ,
- Saver JL,
- et al
- 10.↵American Hospital Association. American Hospital Association Guide Book. Chicago, IL: American Hospital Association; 2008.
- Ricci S,
- Lewis S,
- Sandercock P
- Loke YK,
- White JR,
- Bettencourt-Silva JH,
- Potter JF,
- Myint PK
- Cooney MT,
- Dudina A,
- D’Agostino R,
- Graham IM
- Goff DC Jr,
- Lloyd-Jones DM,
- Bennett G,
- Coady S,
- D’Agostino RB,
- Gibbons R,
- et al
- Reiner Z,
- Catapano AL,
- De Backer G,
- Graham I,
- Taskinen MR,
- Wiklund O,
- et al
- 20.↵JBS3 Board. Joint British Societies’ consensus recommendations for the prevention of cardiovascular disease (JBS3). Heart. 2014;100(suppl. 2):ii1–ii67.
- 21.↵Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA. 2001;285:2486–2497.
- Reeves MJ,
- Fonarow GC,
- Smith EE,
- Pan W,
- Olson D,
- Hernandez AF,
- et al