SemOpenAlex |

SemOpenAlex

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2462861737> ?p ?o ?g. }

Showing items 1 to 100 of ±134 with 100 items per page.

W2462861737 endingPage "2670" @default.
W2462861737 startingPage "2662" @default.
W2462861737 abstract "HomeCirculationVol. 133, No. 25The Future of Clinical Trials in Cardiovascular Medicine Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyRedditDiggEmail Jump toFree AccessResearch ArticlePDF/EPUBThe Future of Clinical Trials in Cardiovascular Medicine Scott D. Solomon and MD Marc A. PfefferMD, PhD Scott D. SolomonScott D. Solomon From Cardiovascular Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA. Search for more papers by this author and Marc A. PfefferMarc A. Pfeffer From Cardiovascular Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA. Search for more papers by this author Originally published21 Jun 2016https://doi.org/10.1161/CIRCULATIONAHA.115.020723Circulation. 2016;133:2662–2670IntroductionFew medical disciplines have been as influenced by the results of randomized controlled trials as cardiovascular medicine. Whether for the treatment of hypertension, acute myocardial infarction, use of specific interventional techniques, devices, or primary prevention, the randomized trial has emerged as the principal method by which new therapies are evaluated. Results of clinical trials have transformed cardiovascular practice from one that was anecdote based to one that is evidence based. Before the first outcomes trial in cardiovascular disease, the VA Cooperative trial1,2 in the late 1960s, clinicians based their practice on tradition, often steeped in unjustified beliefs about untested therapies, or hopelessly confounded observational studies. The switch from empirical therapy to therapies based on the results of well-performed and informative clinical trials ushered in the evidenced-based era in cardiovascular medicine. Today, the introduction of a new therapy would be unheard of without the justifying results from a randomized clinical outcome trial.Yet clinical trials in cardiovascular medicine have grown in size, scope, and complexity. As therapeutic approaches in cardiovascular medicine have become more effective, the level of evidence required to support incremental new advances has increased substantially. This fact, in conjunction with real and perceived increases in regulatory requirements, and the mandate for better estimation of cardiovascular safety for noncardiovascular therapies, have resulted in bloated and prohibitively expensive trials. Thus, despite the success of clinical trials in cardiovascular medicine over the past 25 years, the ability to bring new cardiovascular therapies to patients will require new approaches to curtail cost and maintain quality of future trials.The Current Clinical Trial Landscape in Cardiovascular MedicineA clinical trial, defined as “a prospective study evaluating the effect and value of intervention(s) against a control in human beings,”3 is generally designed for 1 of 2 purposes: hypothesis generation or hypothesis testing. The former requires careful, detailed testing of novel therapies that have usually been established in preclinical models or have substantial biological plausibility but have not yet been determined to have efficacy in humans. The latter provides the evidence, within a degree of statistical certainty, that the therapy is both effective and safe. Development of pharmacological therapies generally follow an orderly and deliberate progression from early phase I studies designed to establish initial safety, to phase II trials, aimed at assessing biological efficacy, identifying an appropriate dose, ruling out major safety concerns, and providing sufficient confidence and enthusiasm to proceed to the much costlier “pivotal” phase III outcome trials required for regulatory approval. Phase IV postapproval studies are performed to provide additional postmarketing safety information, to generate mechanistic data to support an indication, to expand indications, or sometimes even to heighten the awareness of and familiarity with a new therapy. Devices follow a similar, although less rigorously defined course, and their approvals are influenced by a somewhat different regulatory pathway.4,5The level of evidence required for either regulatory approval or adoption by the cardiovascular community depends, to a large extent, on the disease and type of therapy. In some cases, new therapies are only required to demonstrate efficacy with respect to a surrogate biomarker for regulatory approval and adoption, as has been the case with new antihypertensive agents, lipid-lowering agents, and glucose-lowering drugs. Nevertheless, concerns about safety, which are nearly impossible to assuage in small efficacy trials, have resulted in the requirement for either pre- or postapproval assessments large enough to detect safety signals. For relatively rare diseases, such as primary pulmonary hypertension,6 hypertrophic cardiomyopathy, or amyloid heart disease, in which large outcomes trials would be impractical or impossible, or for diseases where few effective therapies exist, the regulatory approval pathway can be somewhat less demanding than more common conditions such as heart failure.7,8The rational and deliberate march from phase I to II to III, with each step being dependent on the results of the former, has a number of benefits, and allows for investments to be made sequentially with the hope of maximizing the likelihood of success. This approach assumes that the surrogate end point assessed in phase II is indeed a predictor of success in an outcome trial. Nevertheless, it can be extremely costly and slow, and development times of up to a decade can diminish the patent life of the therapy, and enthusiasm to invest in the outcome trial.Outcomes TrialsPivotal trials required for regulatory approval in most cardiovascular therapies typically incorporate hard outcomes, events or a combination of events that have an impact on prognosis, reflect both the burden of the disease being studied and the potential response to a particular therapy. Over the past quarter century, the end points used in cardiovascular trials have become more directed and more reflective of the potential therapeutic targets, with a shift from using more nonspecific end points, such as all-cause mortality, to more disease-specific end points. Most cardiovascular trials currently use a combination of fatal (eg, death or cardiovascular death) and nonfatal end points (eg, myocardial infarction, stroke, heart failure hospitalization) in a composite often termed major adverse cardiovascular events, although the exact makeup of the components of major adverse cardiovascular events has been highly variable.9 From a regulatory perspective, pivotal trials are dependent on the success of the primary end point, and it is extremely rare for a therapy to be approved based on a trial in which the primary end point was not met.Cardiovascular Safety TrialsAlthough traditionally most trials are designed to determine whether a new therapy is efficacious, some large clinical trials in cardiovascular medicine are being performed for the express purpose of determining cardiovascular safety. The need for these trials arose from regulatory mandates stemming from high-profile identification of adverse cardiovascular outcomes with noncardiovascular therapies, including cyclooxygenase-2 inhibitors,10,11 thiazolidinediones,12 and erythropoietic-stimulating agents.13–15 Indeed, approval of new diabetes therapies for agents that have already been shown to be effective in lowering hemoglobin A1C currently requires very large and expensive (5000–15 000 patient) safety trials designed to rule out cardiovascular risk by excluding a specific hazard.16–18The Increased Cost of Clinical TrialsOne of the major drivers for innovation in the conduct of clinical trials is the rising cost of drug development, which has increased at a much higher rate than inflation during the past 25 years.19 The total development costs for a drug have risen nearly 13-fold ($100 million in 1975 to $1.3 billion in 2005 in constant dollars). For new diabetes agents and factor Xa inhibitors, the total cost of phase II trials has been estimated at ≈$35 000 per patient and for phase III trials, ≈$47 000 per patient,20 resulting in total costs for phase III outcomes trials of up to half a billion dollars. This rise has been a result of increases in the length of trials, the work required by study personnel, the number of trial visits and procedures, the type and amount of data collected, and site monitoring, all driven to some extent by real or perceived regulatory requirements, often involving lengthy and costly pretrial approvals.21 This trend has resulted in a greater proportion of drug development dollars being spent on trials, especially phase III clinical outcomes trials, and a greater proportion of trials being performed in less costly regions outside the United States.22 These considerations have resulted in shifts away from the cardiovascular arena by some pharmaceutical companies23 that have perceived a relatively greater potential return from investments in noncardiovascular areas.The Evolving Clinical Trial LandscapeAdvances in trial design and execution have been the direct result of efforts to improve the efficiency of trials. The drive to reduce the cost and complexity of trials is leading to a heightened awareness of the need for more targeted therapies for more selected populations. This evolution is being fueled by a better understanding of the underlying biology of heart disease leading to better phenotyping of patients and more specific therapeutic targeting, improved use of EMRs and other electronic technologies for identification of patients and ascertainment of events, consideration of novel approaches to study design and analysis that will allow for reducing the size of trials or performing large trials with reduced overall complexity, improvement in surrogate outcomes to allow for rapid identification of therapies likely to succeed in outcomes trials, and novel statistical and analytic methods to improve the efficiency of trials.Improvements in Patient Selection, Phenotyping, and Biological TargetingPhenotyping, Patient Selection, and Personalized MedicineCardiovascular clinical trials have generally categorized patients broadly, and have applied their results to the average patient. With the exception of niche diseases with well-defined phenotypes, such as amyloid heart disease, hypertrophic cardiomyopathy, primary pulmonary hypertension, the main categories of cardiovascular trials have been encompassing: acute coronary syndrome, myocardial infarction, hypertension, and heart failure. Heart failure, typically grouped into reduced or preserved ejection fraction, chronic or acute decompensated, is rarely phenotyped further, despite the fact that the etiology of heart failure can vary substantially. The reasons for this reluctance to phenotype cardiovascular diseases in a more granular way are multifactorial. Specific etiologies can be difficult to identify on an individual patient basis, especially late in the course of disease. Thus, most therapies have targeted final common pathways, an approach that has been largely successful in treating the majority of cardiovascular diseases. Moreover, sponsors of new therapies typically wish to maintain the largest potential market for their products, and testing therapies in specific niches would necessarily limit the indicated market.In contrast, oncological trials use careful and elaborate phenotyping and genotyping to identify and select groups of patients with specific subtypes of tumor cells or genetic signatures.24 These personalized medicine approaches have allowed for the development of highly targeted therapies with faster regulatory approval based on trials with far fewer patients than in the average pivotal cardiovascular trial. This tactic is just beginning to gain traction in the cardiovascular arena, with several ongoing cardiovascular trials targeting diseases based on specific genotypes, including patients with cardiac amyloid attributable to specific transthyretin mutations,25,26 lamin mutations,27 long-QT syndrome,28 and hypertrophic cardiomyopathy.29 Moreover, identification of genes that can influence specific drug responses30 is a strategy that has already been tested in several trials,31 and trials testing genotype- or phenotype-based strategies are likely to become more common in the future.Biological TargetingAs understanding of the fundamental biology underlying diseases allows for better phenotyping of patients, it will also allow for better targeting of specific diseases with novel therapies that directly influence causal disease processes rather than simply the pathophysiologic end result. To date, most therapies in cardiovascular medicine have been directed at pathophysiologic alterations that are substantially downstream from the causation. Although patients with heart failure because of specific genetic mutations may respond to the same medications used to treat heart failure of other etiologies, there may be considerable advantages to more upstream targeted approaches, such as amyloid heart disease therapies that are directed at preventing misfolding of the transthyretin protein.32 Given that overall patient exposure will be relatively low with niche therapies, long-term safety concerns will likely require postapproval pharmacovigilance approaches.33Identification of More Meaningful Surrogate End PointsSurrogate end points are biomarkers used to determine whether a particular therapy has a presumed favorable action and can serve as a predictor of success in a pivotal trial. Lowering of blood pressure or low-density lipoprotein cholesterol, reduction in N-terminal pro-brain natriuretic peptide,7 and improvement in ventricular remodeling34 have all been used as surrogate markers in cardiovascular disease. Admittedly, the extent to which these are accurate intermediates in the pathway between the disease and the outcome, and more than just risk factors, has been questioned. Currently the number of accepted surrogates for assessment of new therapies and their utility in predicting response to therapy are relatively limited. Novel biochemical biomarkers35 and new imaging methods36 that have been shown to be mediators of risk in cardiovascular disease have also been proposed as measures of therapeutic efficacy. Identification of novel and accurate biomarkers that could serve as surrogate end points would facilitate rapid phase II development, including dose finding, and thereby reduce the time required to decide to proceed to the all-important phase III trial necessary to determine safety,37 or to halt a development program and thereby minimize expenditure on ineffectual therapies.Improvement in the Technology and Conduct of Clinical TrialsUse of Electronic Medical RecordsThe use of sophisticated and comprehensive electronic medical records (EMRs) systems in an ever increasing number of healthcare networks offers enormous opportunities to improve the conduct of clinical trials.38 Although not generally designed for this purpose, EMRs based on structured databases can be queried for specific inclusion and exclusion criteria to identify and screen prospective patients for clinical trials, and can be used to ascertain potential end points in trials. Thus far, they have proven most adept at the former. Although these systems offer the promise of identifying large numbers of patients who meet specific eligibility criteria, approaches using electronic records will be most effective when a trial’s inclusion and exclusion criteria are broad enough that patients can be adequately screened by EMR queries without substantial time-consuming additional manual intervention. In practice, although EMR approaches can be a good means for initial screening, substantial human input will generally be required to identify truly eligible patients from a very broad list. Most importantly, electronic screening cannot, in most cases, obviate the need for the critical consent process, which often require one-on-one discussions between investigators and patients. However, identification of patients likely to fulfill a study’s entry criteria before more labor-intensive screening processes can substantially reduce the effort necessary and improve trial efficiency.In a completely electronically connected healthcare network, clinical trial outcomes could also be identified efficiently by an EMR. This approach requires complete or near-complete capture of the key end points typically used in clinical trials, including deaths and hospitalizations with admission diagnoses and pharmacy prescriptions. Integrated healthcare delivery systems such as Kaiser Permanente or the Veterans Administration have successfully laid the groundwork to include their EMR infrastructure in research.39 Although EMRs can help with further categorizing these end points, it has not yet been clear that they would be able to do so with the degree of granularity sufficient to avoid the need for traditional end point adjudication, or that the regulatory environment would be conducive to this approach. Moreover, even the most sophisticated networks have some degree of leakage, especially in locations where seasonal travel or out-of-network hospitalizations, not captured by the EMR, are common. Although the limitations of incomplete follow-up can be overcome to some extent by increasing the sample size, it is likely that, in many locations, including most of the United States, hybrid approaches combining EMR with more traditional ascertainment methods will continue to be necessary for the foreseeable future.Use of Novel Data Collection MethodsCardiovascular clinical trials have long used innovative and novel technologies for data acquisition. Automated blood pressure cuffs and ambulatory blood pressure monitors have replaced manual blood pressure assessments in clinical trials beyond just hypertension, and assessments of cardiac size and function from imaging modalities such as echocardiography or cardiac MRI have been used both as primary and secondary end points in trials.40 Implanted devices such as pacemakers or implantable cardioverter defibrillators can provide continuous data on cardiac rhythms, including the occurrence of hard-to-detect rhythms such as paroxysmal atrial fibrillation. Less invasive technologies that capture these type of continuous data, including bandage-sized patches that are applied for weeks at a time41 or tiny permanently implanted subcutaneous implantable loop recorders to monitor cardiac rhythms,42,43 or even more invasive approaches such as devices that monitor pulmonary artery pressures, are likely to be used more extensively in future trials.Higher-quality versions of consumer-based devices can collect and provide large amounts of accurate data directly to investigators through smart phones and home wireless networks. These include bathroom scales, home blood pressure devices, or sensor-based devices worn as wristbands, or even built into the fabric of clothing. Several all-in-one devices which capture nearly continuous data on ambient activity, oxygen saturation, heart rate, galvanic skin response are currently being tested.44 Tablets or even smart phone interfaces can be used directly by patients to capture patient-reported outcomes. All these technologies have the potential to provide more continuous data to investigators, rather than just that which can be collected at study visits, and simultaneously reduce overall patient burden, time spent by study personnel, and expense. Moreover, these approaches could potentially open up trials to patients who would not have previously been willing to enroll in trials that required substantial and time-consuming study visits, and thus broaden the spectrum of patients being studied.Use of Other e-Technologies and Social MediaIn addition to the incorporation of EMRs into clinical trials, other electronic technologies have the potential to increase the efficiency of trials.45 Although electronic tools have been used for study management and operations for several decades, the use of these technologies for communication with study subjects is growing rapidly. Websites are increasingly being used to provide study information to investigators and patients. More recently, Internet and social media platforms, including social networks like Facebook, have been used to recruit patients,46–48 and emails and text messaging have been used to remind patients of study visits, improving retention of patients.49 As the demographics of social network users changes, these platforms will become even more viable as a method for recruiting and following trial subjects. Yet, despite the potential of e-technologies, substantial issues of patient privacy and Internet security, which have largely been addressed by EMR systems, remain to be addressed, and currently represent the greatest barriers to more widespread adoption of these technologies. In the case of some rare diseases, patient advocate groups have served as potential gateways to novel clinical trials. Indeed, direct patient engagement and participation in study design are seen as a potential necessity for some clinical trial–funding opportunities. Networks such as PCORnet50 are directly engaging patients, family members, and caregivers in prioritizing research questions and participating in trial leadership and governance.51Finally, open-source development environments from large technology manufacturers such as Apple52 and Google53 have resulted in a recent explosion of medical mobile applications, many of which are geared toward collecting data that can be used for research. Recently, >11 000 people signed up for a heart disease study based at Stanford University in 1 day by using an Apple ResearchKit–based mobile application.54 Smartphone apps will likely play a growing role in recruitment, retention, and communicating with patients in clinical trials.Improvements in Trial Concepts and DesignTrials of Therapeutic StrategiesCardiovascular clinical trials are increasingly testing therapeutic strategies rather than individual therapies. These strategies can involve either multiple therapies or therapeutic approaches, or a combination of the results of diagnostic testing in conjunction with a therapeutic approach. The ACCORD trial55 testing the strategy of intensive or less intensive blood pressure lowering and glucose control in patients with diabetes mellitus, and the SPRINT trial56 comparing blood pressure–lowering strategies are examples of strategy trials. Similarly, the PROMISE trial57 randomly assigned patients to a strategy of initial anatomic testing with the use of coronary computed tomographic angiography or to functional testing for the management of coronary disease. The ongoing GUIDE-IT trial will assess the strategy of lowering N-terminal pro-brain natriuretic peptide in patients with heart failure irrespective of the specific therapies used to achieve this goal.58 In cases where strategies involve the use of a device or diagnostic test, care needs to be taken to ensure that it is ultimately the strategy, rather than the device or diagnostic test, that is being tested. Trials that incorporate both diagnostic and therapeutic components test approaches to patient care that are more similar to the complex decisions that physicians make in practice, and are likely to become more commonplace in the future.Cluster Randomization TrialsTrials of specific strategies can be effectively tested on a group of patients rather than individual patients by randomly assigning groups of patients rather than individual patients. Entire clinics or even entire communities can be assigned to a specific interventional strategy. One such trial tested involvement of practice nurses and implementation of a chronic care model in patients with diabetes mellitus.59 This approach is especially useful when the primary intervention involves a method of practice or a complex group of interventions that are testing different healthcare delivery strategies, and may be more efficient, simpler to execute, and less expensive than traditional individual patient randomization approaches.Asking Multiple Questions SimultaneouslyAlthough factorial designs have been used as a way to test multiple therapies or strategies in a single patient population, they represent 1 potential approach to substantially reduce costs and leverage the clinical trial infrastructure to answer 2 questions for little more than the cost of answering one.60 When the likelihood of an interaction between the therapies is low, the increased sample size needed is substantially less than it would be to conduct 2 individual trials. There are many examples of successful factorial studies, including the Physician’s Health Study of aspirin and β-carotene,61,62 the Women’s Health Initiative of hormonal therapy and dietary intervention,63 the HOPE trial testing ramipril or vitamin E in patients at high risk for cardiovascular events,64 and the DREAM trial testing ramipril and rosiglitazone in patients with diabetes mellitus.65,66 Indeed, factorial designs could even potentially be used by >1 sponsor to ask 2 distinct questions with separate therapies in a single patient population. A 2×2 factorial study testing both a pharmacological and device-based approach in heart failure, for example, would potentially reduce the development costs for 2 sponsors with very different goals. Although the logistics of multiple sponsorship and separate therapies can be challenging, the cost-savings potential is enormous, and several approaches to foster collaboration among multiple industry sponsors in clinical trials may help facilitate multisponsored trials.67The Large Simple TrialThe extremely large sample sizes required to demonstrate improvements in outcomes in populations with relatively good prognoses have made traditional trials in cardiovascular medicine prohibitively expensive. With very specific inclusion and exclusion criteria, these trials typically collect extensive amounts of data in a very rigorous, but time-consuming and resource-intensive manner. An alternative to these types of trials is the large simple or pragmatic trial. In general, pragmatic trials have relatively simplified inclusion and exclusion criteria, allowing for easy identification of patients, a straightforward study design, relative ease of administering the required intervention, and non–resource-intensive assessment of outcomes.68 Although these designs would not generally be practical for labor-intensive phase II trials requiring careful measurements of surrogate outcomes, they have already proven useful in studies where the number of patients enrolled is more important than the comprehensiveness of the data collected. Such a design could be applicable to a large mortality trial where mortality was the primary end point, because deaths are generally readily available from public records. The design may be less applicable for trials using end points that are more difficult to ascertain or require extensive adjudication. Yet the scope of large simple trials will likely evolve as data collection methods improve or become more automated, and healthcare networks can be more effectively used to identify events of interest. Past examples of large simple trials include GISSI69 or ISIS-370, both relatively short-term studies to test simple interventions in the post–mycardial infarction patient. Trials in primary prevention, where the number of patients required would be very high and follow-up exceptionally long because of low event rates, would be ideally suited to this design. Moreover, these trials work best when interventions are simple, such as in a vaccine trial, where drug delivery or adherence considerations would be negligible, and when the data being collected would be minimal.Just how simple the large simple trial can become without compromising quality or the ability to answer the experimental question remains to be seen. For example, typical approaches to simplify trials include using nonblinded designs, including open-label blinded end point designs, substantially reducing complexity, although open-label studies may increase the likelihood of dropout and bias. Another cornerstone of the large, simple trial is the ability to easily ascertain end points. In an ideal scenario, end points could be ascertained electronically, through sophisticated EMRs.71 However, in practice, even the most extensive and sophisticated networks may lose patients who receive out-of-network care. Increasing sample size, a common way to mitigate this problem, negates some of the cost-savings benefits of the design. Nevertheless, combining simple streamlined trial designs with electronic approaches to end point ascertainment offers the potential to randomly assign large numbers of patients in trials requiring minimal long-term follow-up activities by study personnel.The Randomized Registry TrialObservational registries have long been used to collect data about large numbers of subjects with broad or narrow inclusion criteria, often collecting data on demographics, risk factors, and outcomes similar to what would be collected in a clinical trial. The concept of the registry trial evolved out of the desire to leverage the data collected in registries to identify suitable patients for clinical trials and to better integrate the worlds of clinical research and clinical care.72 Moreover, it is not uncommon for trials, with very specific inclusion and exclusion criteria, to be criticized for not being generalizable to the broader population of patients with a particular disease and thus not representative of real-world patients,73 as evidenced by the fact tha" @default.
W2462861737 created "2016-07-22" @default.
W2462861737 creator A5031560506 @default.
W2462861737 creator A5054830306 @default.
W2462861737 date "2016-06-21" @default.
W2462861737 modified "2023-10-02" @default.
W2462861737 title "The Future of Clinical Trials in Cardiovascular Medicine" @default.
W2462861737 cites W1115877934 @default.
W2462861737 cites W1507364611 @default.
W2462861737 cites W1593442063 @default.
W2462861737 cites W1594438765 @default.
W2462861737 cites W1673665560 @default.
W2462861737 cites W1785062055 @default.
W2462861737 cites W180946569 @default.
W2462861737 cites W1870916828 @default.
W2462861737 cites W1968271334 @default.
W2462861737 cites W1973962628 @default.
W2462861737 cites W1976196376 @default.
W2462861737 cites W1977578239 @default.
W2462861737 cites W1978284123 @default.
W2462861737 cites W1986705681 @default.
W2462861737 cites W1987579102 @default.
W2462861737 cites W1994079702 @default.
W2462861737 cites W1995227999 @default.
W2462861737 cites W1998943227 @default.
W2462861737 cites W2011284201 @default.
W2462861737 cites W2014415076 @default.
W2462861737 cites W2018276694 @default.
W2462861737 cites W2019068261 @default.
W2462861737 cites W2019234100 @default.
W2462861737 cites W2030333356 @default.
W2462861737 cites W2033980725 @default.
W2462861737 cites W2040029744 @default.
W2462861737 cites W2043696829 @default.
W2462861737 cites W2046224330 @default.
W2462861737 cites W2049154969 @default.
W2462861737 cites W2061596709 @default.
W2462861737 cites W2063374808 @default.
W2462861737 cites W2064959038 @default.
W2462861737 cites W2067235700 @default.
W2462861737 cites W2077033046 @default.
W2462861737 cites W2080718344 @default.
W2462861737 cites W2087260158 @default.
W2462861737 cites W2089985096 @default.
W2462861737 cites W2092075924 @default.
W2462861737 cites W2092608032 @default.
W2462861737 cites W2096355643 @default.
W2462861737 cites W2104895625 @default.
W2462861737 cites W2105400632 @default.
W2462861737 cites W2115607452 @default.
W2462861737 cites W2117185858 @default.
W2462861737 cites W2120427666 @default.
W2462861737 cites W2121175754 @default.
W2462861737 cites W2121283374 @default.
W2462861737 cites W2121908294 @default.
W2462861737 cites W2127125241 @default.
W2462861737 cites W2133776525 @default.
W2462861737 cites W2135589624 @default.
W2462861737 cites W2136145230 @default.
W2462861737 cites W2138839431 @default.
W2462861737 cites W2143094666 @default.
W2462861737 cites W2152949852 @default.
W2462861737 cites W2157065952 @default.
W2462861737 cites W2158131645 @default.
W2462861737 cites W2158968701 @default.
W2462861737 cites W2160133249 @default.
W2462861737 cites W2161392015 @default.
W2462861737 cites W2166155176 @default.
W2462861737 cites W2166650056 @default.
W2462861737 cites W2169490447 @default.
W2462861737 cites W2193508485 @default.
W2462861737 cites W2197071721 @default.
W2462861737 cites W2207992373 @default.
W2462861737 cites W2272130855 @default.
W2462861737 cites W2314701246 @default.
W2462861737 cites W2315403385 @default.
W2462861737 cites W2333929093 @default.
W2462861737 cites W2335312575 @default.
W2462861737 cites W2340149484 @default.
W2462861737 cites W2419595939 @default.
W2462861737 cites W4231673371 @default.
W2462861737 cites W4244918166 @default.
W2462861737 cites W4246475368 @default.
W2462861737 cites W4297446959 @default.
W2462861737 cites W53719006 @default.
W2462861737 doi "https://doi.org/10.1161/circulationaha.115.020723" @default.
W2462861737 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/27324361" @default.
W2462861737 hasPublicationYear "2016" @default.
W2462861737 type Work @default.
W2462861737 sameAs 2462861737 @default.
W2462861737 citedByCount "28" @default.
W2462861737 countsByYear W24628617372017 @default.
W2462861737 countsByYear W24628617372018 @default.
W2462861737 countsByYear W24628617372019 @default.
W2462861737 countsByYear W24628617372020 @default.
W2462861737 countsByYear W24628617372021 @default.
W2462861737 countsByYear W24628617372022 @default.
W2462861737 countsByYear W24628617372023 @default.