2014 ACC/AHA Key Data Elements and Definitions for Cardiovascular Endpoint Events in Clinical Trials
A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Data Standards (Writing Committee to Develop Cardiovascular Endpoints Data Standards)
- AHA Scientific Statements
- clinical trials
- cardiovascular endpoints
- clinical events
- myocardial infarction
- percutaneous coronary intervention
- peripheral vascular intervention
- heart failure
- unstable angina
Table of Contents
2.1. Writing Committee Composition 305
2.2. Relationships With Industry and Other Entities 305
2.3. Review of Literature and Existing Data Definitions 305
2.4. Development of Terminology Concepts 305
2.5. Consensus Development 306
2.6. Relation to Other Standards 306
2.7. Peer Review, Public Review, and Board Approval 306
Data Elements and Definitions 306
3.1. Death Attribution 306
3.1.1. Cardiovascular Cause of Death 307
3.1.2. Noncardiovascular Cause of Death 307
3.1.3. Undetermined Cause of Death 307
3.2. Myocardial Infarction 307
3.3. Hospitalization for Unstable Angina 308
3.4. Transient Ischemic Attack and Stroke 308
3.5. Heart Failure Event 308
3.6. Percutaneous Coronary Intervention 309
3.7. Peripheral Vascular Intervention 309
3.8. Stent Thrombosis 309
Informatics of Controlled Vocabularies 310
Appendix 1. Author Relationships With Industry and Other Entities (Relevant)—2014 ACC/AHA Key Data Elements and Definitions for Cardiovascular Endpoint Events in Clinical Trials 313
Appendix 2. Reviewer Relationships With Industry and Other Entities—2014 ACC/AHA Key Data Elements and Definitions for Cardiovascular Endpoint Events in Clinical Trials 314
Appendix 3. Death Attribution 315
Appendix 4. Myocardial Infarction 318
Appendix 5. Hospitalization for Unstable Angina 324
Appendix 6. Transient Ischemic Attack and Stroke 327
Appendix 7. Heart Failure Event 329
Appendix 8. Heart Failure Event: Additional Details 334
Appendix 9. Percutaneous Coronary Intervention 336
Appendix 10. Cardiovascular Anatomy 350
Appendix 11. Peripheral Vascular Intervention 356
The American College of Cardiology (ACC) and the American Heart Association (AHA) support their members’ goal to improve the care of patients with cardiovascular disease through professional education, research, and development of guidelines and standards and by fostering policies that support optimal patient outcomes. The ACC and AHA recognize the importance of the use of clinical data standards for patient management, assessment of outcomes, and conduct of research, and the importance of defining the processes and outcomes of clinical care, whether in randomized trials, observational studies, registries, or quality-improvement initiatives.
Clinical data standards strive to define and standardize data relevant to clinical concepts, with the primary goal of facilitating uniform data collection by providing a platform of clinical terms with corresponding definitions and data elements. Broad agreement on a common vocabulary with reliable definitions used by all is vital to pool and/or compare data across clinical trials to promote interoperability with electronic health records (EHRs) and to assess the applicability of research to clinical practice. The ultimate purpose of clinical data standards is to contribute to the infrastructure necessary to accomplish the ACC’s mission of fostering optimal cardiovascular care and disease prevention and the AHA’s mission of building healthier lives, free of cardiovascular diseases and stroke.
The specific goals of clinical data standards are:
To establish a consistent, interoperable, and universal clinical vocabulary as a foundation for both clinical care and clinical research, including clinical trials
To promote the ubiquitous use of EHRs and facilitate the exchange of data across systems through harmonized, standardized definitions of key data elements
To facilitate the further development of clinical registries, quality- and performance-improvement programs, outcomes evaluations, and clinical research, including the comparison of results within and across these initiatives
The key elements and definitions are intended to facilitate the consistent, accurate, and reproducible capture of clinical concepts; standardize the terminology used to describe cardiovascular diseases and procedures; create a data environment conducive to the assessment of patient management and outcomes for quality and performance improvement and clinical and translational research; and increase opportunities for sharing data across disparate data sources. The ACC/AHA Task Force on Clinical Data Standards selects cardiovascular conditions and procedures that will benefit from creation of a standard dataset. Subject matter experts are selected to examine/consider existing standards and develop a comprehensive, yet not exhaustive, standard dataset. When a data collection effort is undertaken, only a subset of the elements contained in a clinical data standards listing may be needed, or conversely, users may want to consider whether it may be necessary to collect some elements not listed. For example, in the setting of a randomized clinical trial of a new drug, additional information would likely be required regarding study procedures and drug therapies.
The ACC and AHA recognize that there are other national efforts to establish clinical data standards, and every attempt is made to harmonize newly published standards with existing standards. Writing Committees are instructed to consider adopting or adapting existing nationally and internationally recognized data standards if the definitions and characteristics are useful and applicable to the set under development. In addition, the ACC and AHA are committed to continually expanding their portfolio of data standards and will create new standards and update existing standards as needed to maintain their currency and promote harmonization with other standards as health information technology and clinical practice evolve.
The Writing Committee for the current effort was intentionally expanded to include a regulatory perspective. This reflects the key role of clinical event concepts in evaluating therapeutic safety and effectiveness in clinical trials. This unique collaboration between the regulatory sector and the ACC and AHA acknowledges the need to align key clinical concepts for regulatory reporting and clinical care.
The Health Insurance Portability and Accountability Act privacy regulations, which went into effect in April 2003, have heightened all practitioners’ awareness of our professional commitment to safeguard our patients’ privacy. The Health Insurance Portability and Accountability Act privacy regulations1 specify which information elements are considered “protected health information.” These elements may not be disclosed to third parties (including registries and research studies) without the patient’s written permission. Protected health information may be included in databases used for healthcare operations under a data use agreement. Research studies using protected health information must be reviewed by an institutional review board or a privacy board.
We have included identifying information in all clinical data standards to facilitate uniform collection of these elements when appropriate. For example, a longitudinal clinic database may contain these elements because access is restricted to the patient’s caregivers. Conversely, registries may not contain protected health information unless specific permission is granted by each patient. These fields are indicated as protected health information in the data standards.
In clinical care, caregivers communicate with each other through a common vocabulary. In an analogous fashion, the integrity of clinical research depends on firm adherence to prespecified procedures for patient enrollment and follow-up; these procedures are guaranteed through careful attention to definitions enumerated in the study protocol, case report forms, and clinical event committee charters. When data elements and definitions are standardized across studies, comparison, pooled analysis, and meta-analysis are enabled, thus deepening our understanding of individual studies.
The recent development of quality-performance measurement initiatives, particularly those for which the comparison of providers is an implicit or explicit aim, has further raised awareness about the importance of data standards. Indeed, a wide audience, including nonmedical professionals such as payers, regulators, and consumers, may draw conclusions about care and outcomes. To understand and compare care patterns and outcomes, the data elements that characterize them must be clearly defined, consistently used, and properly interpreted.
William S. Weintraub, MD, MACC, FAHA
Chair, ACC/AHA Task Force on Clinical Data Standards
This publication, commissioned by the ACC/AHA, is the product of a novel collaboration between the ACC and AHA, the US Food and Drug Administration (FDA), and the Standardized Data Collection for Cardiovascular Trials Initiative (SCTI). The aim of the collaboration is to promulgate, for regulatory submissions, clinical data standards for key cardiovascular and stroke endpoint events in clinical trials. The Writing Committee for this particular data standard is unique in that it was convened to achieve the following goals: to address issues particular to regulatory submissions and to provide a framework of data standards that would simplify the design and conduct of clinical trials for those considering regulatory submissions. The intent of this Writing Committee is not to be overly prescriptive. For example, the stroke data elements are minimal by design to allow for flexibility needed to conduct global clinical trials for drugs and devices. In a trial focused on the treatment of stroke, investigators may define additional outcomes and data elements for a particular stroke treatment or indication. The Writing Committee recognizes that these standards may be used for other types of clinical trials and clinical care processes where appropriate. These data standards are a first step in developing a universal language for clinical trials and other types of health-related research.
Effective communication is a cornerstone of the healthcare enterprise. A prerequisite for providing seamless care is the universal and consistent use of medical vocabularies.2 Cardiovascular endpoints such as death, myocardial infarction, stroke, and revascularization are critical in assessing diagnostic and therapeutic approaches in the clinical care, research, and regulatory domains. With the adoption of EHR solutions has come the opportunity to manage health-related information as discrete data.3 Therefore, the ACC/AHA Task Force on Clinical Data Standards established this Writing Committee to identify and harmonize the common data elements involved in key cardiovascular endpoint events. Doing so will allow this vocabulary to be used to improve the assessment of process, performance, and outcomes across multiple dimensions of health care.
In this work, the term “vocabulary” includes the terminology concept, the concept definition, a suggested label for the corresponding data element, permissible values of the data element, and definitions for the permissible values.
The Writing Committee identified the ongoing work of the SCTI as the foundation for the development of this vocabulary. The Writing Committee’s task was to review, refine, and advance as a clinical standard the cardiovascular endpoint terminology set developed by the SCTI. This terminology set largely reflects endpoints related to the symptoms, manifestations, treatment, and consequences of coronary artery disease in both cardiovascular and noncardiovascular drug and device trials. Endpoint concepts related to carotid/cerebral revascularization, peripheral surgical revascularization, and treatment of diseases of the aorta are beyond the scope of this document.
First convened in 2009 by the FDA, the SCTI is a working group formed to improve the quality and efficiency of clinical trials. It includes representatives from academia, professional societies, the Clinical Data Interchange Standards Consortium, Health Level 7, the Clinical Trials Transformation Initiative, pharmaceutical and cardiovascular device manufacturers, and the FDA (which includes the Center for Drug Evaluation and Research and the Center for Devices and Radiological Health). The original goal of the working group was to improve the quality and efficiency of cardiovascular clinical trials (eg, acute coronary syndrome trials, percutaneous coronary intervention [PCI] trials). Subsequently, in recognition of the growing interest in cardiovascular events in noncardiovascular trials (eg, diabetes control trials, weight loss trials), this focus expanded to include noncardiovascular trials. To achieve its goal, the SCTI acknowledged the need for a consistent cardiovascular and stroke endpoint vocabulary comprising terms defined by objective criteria and reported uniformly.4 This framework of standardized key data elements in clinical trials (and potentially the clinical care domain), could also facilitate the conduct of meta-analyses to assess cardiovascular safety and compare the effectiveness of drug and device products.
The SCTI-developed working draft “Standardized Definitions for Cardiovascular and Stroke Endpoint Events in Clinical Trials” is the source document for the data standards in this publication.4 This source document identifies the cardiovascular and stroke endpoint terms relevant to clinical trials and regulatory submissions. The Writing Committee evaluated and harmonized these endpoint terms, categorized the attributes of each data element, developed permissible (ie, allowed) values, and tabulated the content to facilitate computational interoperability and the use of the terminology regulatory domains.
The objective is to use this controlled terminology, initially developed for clinical trials and regulatory submissions, to standardize event reporting and data collection when determinations must be made about cardiovascular and stroke endpoints in the context of clinical care processes, registries, EHRs, and longitudinal drug or device surveillance.
Nonetheless, the Writing Committee recognizes that this terminology is not applicable to all dimensions of health care and that some of these definitions are not applicable to clinical care processes. For example, the “Hospitalization for Unstable Angina” definition may not be optimal for the Centers for Medicare and Medicaid Services because other clinical scenarios could be consistent with unstable angina but not fulfill all the criteria needed to define this endpoint in a clinical trial. Similar scenarios could be relevant to other nonfatal endpoint definitions.
To avoid ambiguity, we recommend maintaining the cardiovascular and stroke endpoint vocabulary described in this document as “regulatory specific” distinct from other vocabularies. Although endpoint definitions may evolve over time, a period in which definitions remain static is needed for terms to be used successfully to conduct a meta-analysis. Furthermore, the ACC/AHA Task Force on Clinical Data Standards should include regulatory data standards experts—in addition to members of this Writing Committee, SCTI representatives, and other clinical trial experts—as an integral part of any necessary definition revision process, because such experts understand the requirements for evaluating drug and device products.
2.1. Writing Committee Composition
The ACC/AHA Task Force on Clinical Data Standards selected the members of the Writing Committee. The committee consisted of 16 people with domain expertise in internal medicine, cardiovascular medicine, neurology, clinical research, epidemiology, invasive and interventional therapies, outcomes assessment, medical informatics, health information management, and healthcare services research and delivery.
2.2. Relationships With Industry and Other Entities
The ACC/AHA Task Force on Clinical Data Standards makes every effort to avoid actual or potential conflicts of interest that might arise as a result of an outside relationship or a personal, professional, or business interest of any member of the Writing Committee. Specifically, all members of the Writing Committee are required to complete and submit a disclosure form showing all such relationships that could be perceived as real or potential conflicts of interest. These statements are reviewed by the ACC/AHA Task Force on Clinical Data Standards and updated when changes occur. Authors’ and peer reviewers’ relationships with industry and other entities pertinent to this data standards document are disclosed in Appendixes 1 and 2, respectively. In addition, for complete transparency, the disclosure information of each Writing Committee member—including relationships not pertinent to this document—is available as an online supplement at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIR.0000000000000156/-/DC1. The work of the Writing Committee was supported exclusively by the ACC and AHA without commercial support. Writing Committee members volunteered their time for this effort. Meetings of the Writing Committee were confidential and attended only by committee members and staff.
2.3. Review of Literature and Existing Data Definitions
Cardiovascular endpoint concepts have long been used in clinical care and research to ascertain and assess outcomes of diagnostic and therapeutic approaches. A series of reference publications provided the foundation for the cardiovascular endpoint concepts identified by the SCTI and developed by the Writing Committee.5–14 What makes this work unique is that it reviews and refines the terms as developed by the SCTI explicitly for use in reporting clinical trial results and in regulatory submissions, and it delineates where these concepts could or should not be used as the foundational vocabulary in routine clinical care.
2.4. Development of Terminology Concepts
The terminology set addressed in this body of work includes cardiovascular endpoints of universal interest in clinical care, research, and regulatory review: death (specifically attribution of the cause of death), myocardial infarction, stroke, transient ischemic attack (TIA), coronary intervention (including stent thrombosis), peripheral vascular intervention, hospitalization for unstable angina, and acute heart failure (HF) events.
The Writing Committee aggregated, reviewed, harmonized, and extended these terms to develop a controlled, semantically interoperable, machine-computable terminology set that would be usable, as appropriate, in as many contexts as possible. As necessary, the Writing Committee identified the contexts where individual terms required differentiation depending on their proposed use (ie, research/regulatory versus clinical care contexts).
The Writing Committee tabulated the content and provided sufficient structure to build and model the informatics formalisms to achieve computational interoperability. The resulting appendices (Appendixes 3–7 and 9–11) list the “terminology concept” in the first column, followed by a clinical definition (“concept definition”) of the terminology concept in the second column. A data element label is suggested for forms-based approaches to data capture. The allowed responses (“permissible values”) for each terminology concept in the next column are the acceptable “answers” for capturing the information. For terminology concepts with multiple permissible values, a bulleted list of the permissible values is provided in the same row as the terminology concept, with successive rows listing each permissible value and corresponding permissible value definition. The process of converting the prose description of an endpoint into this tabular format can be seen by comparing the source text for a HF endpoint event (Appendix 8, an excerpt from the SCTI draft document) and the tabular representation of the same concept in Appendix 7. Where possible, clinical definitions of endpoints (and the corresponding permissible values) are repeated verbatim as defined by the SCTI or as previously published in reference documents.
2.5. Consensus Development
The ACC/AHA Task Force on Clinical Data Standards established the Writing Committee in November 2012, according to the processes described in the Task Force on Clinical Data Standards’ methodology statement.15 As described previously, the responsibility of the Writing Committee was to review and refine the list of candidate terms identified by the SCTI and to harmonize the attributes and other informatics formalisms required to attain interoperability of the terms. During the first 6 months of 2013, the work of the Writing Committee was accomplished through a series of teleconference and Web conference meetings, along with extensive correspondence by e-mail. The review work was distributed among subgroups of the Writing Committee on the basis of their interest and expertise in the components of the terminology set. The proceedings of the work groups were then assembled, resulting in the vocabulary and associated descriptive prose in Section 3. All members reviewed and approved the final vocabulary.
2.6. Relation to Other Standards
The Writing Committee reviewed the work of the SCTI along with available published data standards, specifically those developed for death, acute myocardial infarction, stroke, TIA, unstable angina/non–ST-elevation myocardial infarction, HF, PCI, and peripheral vascular intervention.4–13,16–21 Existing published definitions were adjusted to eliminate verbiage not relevant to an actual definition (eg, instructions such as the phrase “indicate whether the patient has …” have been eliminated).
Through the affirmation and refinement of existing data standards, the Writing Committee anticipates that the vocabulary will facilitate the uniform adoption of these terms, where appropriate, by the clinical care, clinical and translational research, regulatory, quality and outcomes, and EHR communities.
2.7. Peer Review, Public Review, and Board Approval
The “2014 ACC/AHA Key Data Elements and Definitions for Cardiovascular Endpoint Events in Clinical Trials” statement was reviewed by official reviewers nominated by the ACC and AHA. To increase its applicability further, the document was posted on the ACC Web site for a 30-day public comment period. This document was approved for publication by the ACC Board of Trustees on November 12, 2014, and by the AHA Science Advisory and Coordinating Committee on June 13, 2014. The Writing Committee anticipates that these data standards will require review and updating in the same manner as other published guidelines, performance measures, and appropriate use criteria. The Writing Committee will therefore review the set of data elements periodically, starting with the anniversary of publication of the standards, to ascertain whether modifications should be considered.
3. Data Elements and Definitions
As described above, the SCTI identified candidate cardiovascular and stroke endpoint event terms in the draft document “Standardized Definitions for Cardiovascular and Stroke Endpoint Events in Clinical Trials.” The document delineated the following cardiovascular and neurological endpoint events in the context of clinical trials and regulatory reporting:
Undetermined cause of death
Hospitalization for unstable angina
Transient ischemic attack and stroke
Heart failure event
Percutaneous coronary intervention
Peripheral vascular intervention
The SCTI envisioned that data collection and exchange standards for these endpoint events would allow individual investigators and clinical research organizations to collect and exchange research data consistently and efficiently. It also envisioned that adoption of the controlled terminology within the research and regulatory sectors could facilitate consistency across the clinical care domain.
3.1. Death Attribution
Death is classified into 1 of 3 categories: 1) cardiovascular death; 2) noncardiovascular death; and 3) undetermined cause of death. The intent of the classification schema is to identify one, and only one, of the categories as the underlying cause of death. The key priority is differentiating between cardiovascular and noncardiovascular causes of death (Appendix 3).
Collection of appropriate source documentation is critical for rigorous adjudication of the cause of death. Although death certificates establish that the patient died, reliance on information included in death certificates may be problematic; several studies have demonstrated inaccurate coding in the death certificate when death certificates were compared with adjudicated outcomes.22,23 In contrast, autopsy reports are often valuable in assessing the cause of death and should be used whenever possible.
For sudden deaths, even when witnessed, death attribution may be difficult if only limited information is available. Frequently, these deaths are attributed to either sudden cardiac death (cardiovascular death) or death due to an undetermined cause. Sensitivity analyses may be helpful in determining the effect of these events on the primary and major secondary endpoints in a particular clinical trial or development program.
3.1.1. Cardiovascular Cause of Death
Frequently, the cardiovascular death category is not divided further into subcategories such as death resulting from an acute myocardial infarction, sudden cardiac death, or HF, because the cause of death is so often unknown or ambiguous (eg, Does a death after a myocardial infarction count as a myocardial infarction, sudden death, arrhythmic death, or HF death?). Moreover, the underlying cause of death and the mode of death (ie, most proximate event associated with death) may overlap substantially. In contrast, precision is more achievable with respect to nonfatal events.
However, in cases where subclassification is desired, the Writing Committee recommends a uniform approach for categorizing the attributable cause (and not just the proximate event) for cardiovascular death. The suggested subcategories for attribution of death to a cardiovascular etiology are acute myocardial infarction, sudden cardiac death, HF, stroke, cardiovascular procedure, cardiovascular hemorrhage, and other cardiovascular causes. “Death due to other cardiovascular causes” refers to a cardiovascular death not included in the above categories but with a specific known cause, such as a pulmonary embolism or peripheral arterial disease. In addition, “death due to cardiovascular hemorrhage” refers to a death related to hemorrhage such as a nonstroke intracranial hemorrhage, nonprocedural or nontraumatic vascular rupture (eg, aortic aneurysm), or pulmonary hemorrhage from a pulmonary embolism. In contrast, if a pulmonary hemorrhage were a result of a contusion from a motor vehicle accident, the cause of death would be noncardiovascular (death due to trauma). Although these subcategories may not be applicable to all study populations, therapeutic areas, or drug and device development programs, the events in these subcategories occur relatively frequently in the general population and probably contribute to mortality in any observation. In some trials, subclassification of cardiovascular causes of death may prove helpful in understanding pathophysiology in the context of drug or device programs.
3.1.2. Noncardiovascular Cause of Death
Identifying noncardiovascular causes of death is important when assessing competing mortality risks in both cardiovascular and noncardiovascular trials. The proposed schema for noncardiovascular causes of death is more general than that for cardiovascular causes of death. This noncardiovascular schema could be expanded to capture other causes for specific trials in particular therapeutic areas or for specific drug or device development programs if a specific toxicity has been identified in nonclinical work or early clinical trials. When death is clearly due to a noncardiovascular cause, a cardiovascular cause of death is excluded. The proposed values represent commonly used noncardiovascular categories.
3.1.3. Undetermined Cause of Death
In general, this category of death should apply to few patients in well-run clinical trials. Attribution of causality may be limited or impossible if information available at the time of death is minimal or nonexistent. In such cases, the date of death may be the only data element captured.
The key priority is to prespecify how these deaths will be classified and to implement a uniform approach throughout the conduct of the trial. Occasionally, it may not be possible to determine exact causality when 2 lethal conditions contribute to death equally. In this circumstance, 1 condition should be chosen, with consideration of the issue being studied. For example, if cardiac safety is under consideration and the competing causes of death are cardiovascular and noncardiovascular, cardiovascular death should take precedence.
3.2. Myocardial Infarction
The categorization and definitions of the types of myocardial infarction are derived from the “Third Universal Definition of Myocardial Infarction,”13,14 the “2012 ACCF/AHA Guideline for the Management of Unstable Angina/Non-ST-Elevation Myocardial Infarction,”12 and the “2013 ACCF/AHA Guideline for the Management of ST-Elevation Myocardial Infarction.”18
The key recommendation is to base thresholds for biomarker detection of myocardial infarction on 99th percentile values (ie, the upper reference limit) rather than on “upper limit of normal” values. Multiple assays exist for cardiac troponin and the MB fraction of creatine kinase (CK-MB), and assay characteristics vary by manufacturer. Some assays reported by local laboratories provide the 99th percentile and a higher “decision limit” or upper limit of normal above which myocardial infarction should be considered. The “Third Universal Definition of Myocardial Infarction”13,14 recommends the use of the 99th percentile upper reference limit as the reference standard. The data elements developed in Appendix 4 allow both the 99th percentile and the upper limit of normal (or both) to be captured, depending on the reporting approach used in the central or local laboratory. Instead of listing every cardiac biomarker assay in Appendix 4, we have elected to represent all assays with the generic term [cardiac biomarker]. The actual biomarker assay used should replace the generic term [cardiac biomarker]. Collection of serial biomarker values to capture all measurements (and to reflect rise or fall of the biomarker) would recursively use the same data element construct as the approach for capturing a single value. Cardiac troponin is the preferred biomarker. If troponin values are not available, then CK-MB mass is used as an alternative.
The terminology set includes data elements for stent restenosis without occlusion as a type of acute myocardial infarction (type 4c) and asymptomatic postbaseline myocardial infarction detected during follow-up. The data elements that reflect old or prior myocardial infarction at baseline are not included here but can be found in the “2013 ACCF/AHA Key Data Elements and Definitions for Measuring the Clinical Management and Outcomes of Patients With Acute Coronary Syndromes and Coronary Artery Disease.”16
The Writing Committee acknowledges that there is disagreement about how to define a “clinically relevant myocardial infarction” after coronary revascularization (PCI or coronary artery bypass graft). An expert consensus group from the Society for Cardiovascular Angiography and Interventions24 proposes the use of CK-MB instead of troponin and different cut points from those included in the “Third Universal Definition of Myocardial Infarction.”13,14 A detailed discussion of these differences is beyond the scope of this publication but is provided by White.25 At this time, the Writing Committee continues to support the “Third Universal Definition of Myocardial Infarction”13 for harmonization purposes but recognizes that this matter requires further study. As long as cardiac biomarker values (both cardiac troponin and CK-MB) and 99th percentile upper reference limit values are recorded, virtually any definition of periprocedural myocardial infarction can be applied and examined with respect to outcome.
3.3. Hospitalization for Unstable Angina
Hospitalization for unstable angina is a commonly used endpoint in clinical trials evaluating the efficacy or safety of cardiovascular therapies such as lipid-modifying agents, antihypertensive drugs, antithrombotic therapies, and coronary interventions. Unlike traditional endpoints such as death, myocardial infarction, or stroke, hospitalization for unstable angina, by necessity, involves some degree of subjective assessment of the most likely etiology of symptoms resulting in hospital admission. The terminology set for unstable angina (Appendix 5) focuses on data elements needed for determining whether symptoms truly represent cardiovascular ischemia, including the character and duration of the presenting symptoms, the proximity of symptom onset to hospitalization, and the duration of hospitalization. Electrocardiographic abnormalities are pivotal to the diagnosis. Such abnormalities include the presence or absence of deviations in the ST segment, morphology of ST-segment changes (horizontal or downsloping versus upsloping), and the magnitude of the deviation. Many patients without high-risk features (ie, patients with low TIMI [Thrombolysis in Myocardial Infarction] or GRACE [Global Registry of Acute Coronary Events] risk scores) undergo provocative testing for inducible myocardial ischemia, requiring measurement of ST elevation or depression during electrocardiographic monitoring. Alternatively, exercise or pharmacological stress testing may involve assessment of wall motion abnormalities on echocardiography and/or reversible perfusion defects by nuclear scintigraphy or magnetic resonance imaging. Other important data elements include angiographic evidence of the severity of coronary stenosis or presence of coronary thrombus in a vessel believed to be responsible for the ischemic signs and symptoms. Additional data elements include the need for coronary revascularization by PCI or coronary bypass surgery of lesion(s) believed responsible for the hospitalization.
The need for escalation of pharmacological therapy (nitrates, beta-blockers, or other antianginal therapy) may provide supportive evidence for a diagnosis of unstable angina. Last, to fulfill the criteria for unstable angina, cardiac biomarkers must be negative and there can be no evidence of acute myocardial infarction.
3.4. Transient Ischemic Attack and Stroke
TIA and stroke endpoints (Appendix 6) are designed to capture the incidence of new TIA and stroke, type of stroke (ie, ischemic, hemorrhagic, or undetermined), and severity of stroke (ie, mortality and level of functional disability). The modified Rankin Scale26 is recommended as the measure of disability. Hemorrhagic stroke may be further subcategorized as intracerebral hemorrhage or subarachnoid hemorrhage if there is sufficient information to make this determination.
The Writing Committee proposes the following definition of TIA: “a transient episode of focal neurological dysfunction caused by brain, spinal cord, or retinal ischemia, without acute infarction.” This definition is identical to that adopted by the SCTI (FDA Stroke Team) and is based on one previously proposed in an AHA/American Stroke Association (ASA) scientific statement9 with one subtle but important difference: as defined by the scientific statement, TIA is “a transient episode of neurological dysfunction caused by focal brain, spinal cord, or retinal ischemia, without acute infarction.” The SCTI definition that the Writing Committee has adopted emphasizes the clinical presentation rather than the anatomic location of the TIA and may be more appropriate for clinical trial use because the availability of imaging modalities may vary greatly from one study center to the next.
In contrast to TIA, stroke is defined on the basis of the presence of acute infarction as demonstrated by imaging or based on the persistence of symptoms. The Writing Committee acknowledges that the categorization of TIA versus ischemic stroke depends partly on the sensitivity of the diagnostic assessments for brain infarction. For example, patients with symptoms of short duration (eg, <24 hours) but evidence of infarction on magnetic resonance imaging could be categorized as having had an ischemic stroke. In contrast, in the absence of highly sensitive magnetic resonance imaging evidence, the same patient could be categorized as having had a TIA. The presence of persisting symptoms should be considered sufficient evidence for stroke rather than TIA. Primarily on the basis of consensual practice (rather than objective evidence), the AHA/ASA has recommended the existence of symptoms for at least 24 hours as an operational definition of persisting symptoms to indicate stroke rather than TIA.19 However, the time cut point that best discriminates between infarction and the absence of infarction remains largely undefined. Accordingly, for any clinical trial that plans to use the duration of symptom persistence to operationally discriminate between TIA and stroke, the Writing Committee and SCTI recommend prespecifying the particular duration in the protocol, although both acknowledge that a duration ≥24 hours is frequently used.
Depending on the trial objectives, additional optional information could be recorded for analysis, including physical examination findings (eg, using the AHA/ASA guideline–recommended National Institutes of Health Stroke Scale), presumed mechanism of ischemic stroke, and impact on additional patient-centered outcomes such as basic or instrumental activities of daily living. Investigators interested in collecting more detailed information on stroke outcomes should consider using the National Institute of Neurological Disorders and Stroke Common Data Elements, available online at www.commondataelements.ninds.nih.gov.
3.5. Heart Failure Event
HF is a common outcome of many different etiologies and may be associated with cardiovascular and noncardiovascular treatment modalities. Accurate recognition of HF events is important because of the poor outcomes associated with them and because of their increasing prevalence and societal burden. In clinical trials, when a specific uniform definition is lacking, the concurrence between the initial and adjudicated assessment of HF is lower than is the case with adjudications of myocardial infarction and/or stroke.27 This lack of concurrence illustrates the challenges investigators face in classifying HF events and underlines the importance of a standardized definition of event. A consistent definition will ensure that all HF events are accurately reported in all clinical trials and registries.
The proposed HF endpoint event (Appendix 7) has been constructed independent of whether the exacerbation of HF results in hospitalization, recognizing that exacerbation of HF can often be managed on an outpatient basis such as with an urgent or unscheduled outpatient office/practice or emergency room visit. Instead, the key characteristic of a HF event is the need for a resource-intensive response to failure of the primary therapeutic management strategy. The terminology set for a HF event requires both subjective and objective findings, including worsening symptoms and signs, as well as laboratory evidence supporting the diagnosis of worsening HF. Also incorporated into the definition is the requirement for a substantive intensification in HF therapies, whether pharmacological, mechanical, or both. For additional details, see Appendix 8.28–40
3.6. Percutaneous Coronary Intervention
The vast majority of catheter-based interventional cardiology procedures are performed to treat atherosclerotic coronary artery lesions.
For coronary revascularization procedures, it is important to determine whether the procedure was performed to treat symptoms of myocardial ischemia or based solely on coronary anatomic characteristics. It is also important to document whether the Heart Team considered the patient to be inappropriate for surgical revascularization due to prohibitive comorbidities. Medical records that include a description of symptoms and objective assessments of ischemia should be reviewed to determine whether the revascularization procedure was clinically indicated. Imaging reviews by independent core laboratories (eg, angiography, intravascular ultrasonography, optical coherence tomography) are particularly useful for reducing potential bias.
The terminology set for PCI (Appendix 9) concentrates on PCI status, procedural success, target lesion failure, target lesion revascularization, and both intraprocedural and vascular complications. Of specific note, the Writing Committee identified limitations in the nomenclature of the coronary arteries as described for the Coronary Artery Surgery Study (CASS),41 which has subsequently been updated by the Bypass Angioplasty Revascularization Investigators (BARI)42 and is currently used by the ACC National Cardiovascular Data Registry. For example, the nomenclature does not address the concepts of ostial or bifurcation disease and does not follow a consistent convention for naming coronary segments. The Writing Committee therefore proposes an update to the CASS/BARI/National Cardiovascular Data Registry coronary artery nomenclature (Appendix 10) to better capture data evaluating treatment approaches to ostial and bifurcation disease, improve the consistency and completeness of coronary artery nomenclature, include the Medina Classification43,44 as a standard, and reflect universal conventions and terminology currently used by angiography core laboratories. Finally, as new classes of intracoronary therapy (eg, drug-coated balloons, bioresorbable drug-eluting stents/scaffolds) are developed, novel mechanisms of failure may be identified that will require modification and addition to this controlled vocabulary.
3.7. Peripheral Vascular Intervention
Peripheral artery disease (PAD) is widespread. Of all the atherosclerotic syndromes, the clinical relevance of PAD is poorly appreciated by primary care physicians, cardiovascular specialists, and patients alike. Not only does PAD reduce the physical functioning of affected patients, but it is associated with a marked increase in all-cause and cardiovascular mortality.
Although vascular disease is defined as “all diseases of the arteries, veins, and lymphatic vessels,”10 for simplicity, this vocabulary for peripheral vascular intervention endpoints focuses on data elements that describe revascularization interventions involving the peripheral arterial circulation. These data standards concentrate on PAD involving the infrarenal aorta, iliac, and infrainguinal arteries and carotid, renal, mesenteric, and aortic interventions. Of note, upper extremity or intracranial vascular diseases are beyond the scope of this publication.
Appendix 11 lists the vocabulary to facilitate uniform reporting of endovascular and surgical interventions for patients with PAD, thereby allowing comparisons of drug, device, and surgical treatments for PAD. Included are harmonized definitions of success and failure that are derived from the coronary revascularization terminology, including concepts of target lesion and target vessel revascularization. Although somewhat arbitrary, the proposed construct includes the division of the lower extremity arterial circulation into the 3 “vessel” territories, or levels (aorto-iliac, femoral-popliteal, and tibioperoneal) analogous to the division of the 3 coronary vessel territories.
As new classes of endovascular therapy (eg, drug-coated balloons, bioresorbable drug-eluting stents/scaffolds) are developed, novel mechanisms of failure may be identified that will require modification and addition to this controlled vocabulary.
3.8. Stent Thrombosis
According to the classification proposed by the Academic Research Consortium, stent thrombosis is defined as definite, probable, or possible.7 Definite stent thrombosis is defined as occurring when clinical presentation is consistent with acute coronary syndrome and angiography or autopsy examination confirm stent occlusion or thrombus. Probable stent thrombosis is defined as death occurring within 30 days that cannot be attributed to another cause or when myocardial infarction occurs at any time point and is attributable to the target vessel in the absence of angiography confirming another culprit lesion. Finally, possible stent thrombosis is defined as occurring when the patient dies after >30 days and death is not explained by another cause. The terminology set (Appendix 9) focuses on data elements required for confirmation of stent thrombosis. To classify these events accordingly, the following information is required: clinical details surrounding the acute event; dates and procedural information for all prior stent procedures; serial electrocardiograms at the time of the event and for appropriate duration of follow-up; serial cardiac biomarkers; results of coronary angiography with review by an independent angiographic core laboratory or independent clinical events committee; and clinical details surrounding all deaths, including death certificate and autopsy report if applicable. When available data support >1 classification, the highest level of certainty should be reported.
4. Informatics of Controlled Vocabularies
Variability in the definitions, formatting, and encoding of clinical concepts hinders the use, exchange, and analysis of information in health care. Efficient use of healthcare information requires both syntactic interoperability (ie, standards and protocols for formatting, packaging, and transmission required for computer-to-computer data transfer) and semantic interoperability (ie, the capacity of computer systems to transmit data with unambiguous, shared meaning, enabling machine-computable logic, data federation, inferential processing, and knowledge discovery).45,46 To achieve these forms of interoperability, the Writing Committee specified the attributes of the endpoint concepts relevant to the informatics of controlled vocabularies. These attributes (terminology concept, concept definition, permissible values, permissible value definitions) are only a subset of those needed to characterize data elements. Other attributes are still needed to fully qualify a terminology set as a controlled vocabulary; these include preferred abbreviation, concept unique identifier, data type, data format, relationships to other terms, use of case context describing where and when a concept is assessed, and concept steward. The need to be explicit is particularly relevant because the class of endpoint events represents summative concepts more useful for assessing responses and outcomes to therapeutic approaches and treatments. The use of summative concepts contrasts with the emphasis in EHR solutions on diagnoses as classified by taxonomies such as the International Classification of Disease and the Systemized Nomenclature of Medicine—Clinical Terms.
Under FDA grant 1R24FD004411-01, this terminology set has been developed as a controlled vocabulary in the cardiovascular domain of the Clinical Data Acquisition and Standards Harmonization, and the tabular representation of this work is available for download at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIR.0000000000000156/-/DC2 and http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIR.0000000000000156/-/DC3. The terminology set will also be developed in an International Organization for Standardization/International Electrotechnical Commission 11179 standard metadata repository; this specification provides a standardized grammar and syntax for describing data elements and associated metadata, resulting in unambiguous representation and interpretation of data.47,48 Specifically, the endpoint concepts will be represented in the National Institutes of Health/National Cancer Institute Data Standards Registry and Repository to facilitate the use of the terminology set across the clinical care, research, and regulatory domains. Finally, it is intended for this terminology set to be developed and balloted through the HL7 EHR System Functional Model process to further foster adoption in EHR systems.
The Writing Committee acknowledges that cardiovascular and stroke endpoint event concepts are a subset of a larger set of cardiovascular endpoints. In particular, additional concepts such as those describing carotid/cerebral revascularization, peripheral surgical revascularization, aortic dissection, abdominal aortic aneurysm, aortic surgery, and valvular heart disease remain to be developed.
American College of Cardiology
Patrick T. O’Gara, MD, FACC, President
Shalom Jacobovitz, Chief Executive Officer
Lara E. Slattery, MHS, Team Leader, ACC Scientific Reporting
Amelia Scholtz, PhD, Publications Manager, Clinical Policy and Pathways
American College of Cardiology/American Heart Association
Maria Lizza D. Isler, BSMT, Specialist, Clinical Data Standards
American Heart Association
Elliott Antman, MD, FAHA, President
Nancy Brown, Chief Executive Officer
Rose Marie Robertson, MD, FACC, FAHA, Chief Science Officer
Gayle R. Whitman, PhD, RN, FAHA, FAAN, Senior Vice President, Office of Science Operations
Melanie B. Turner, MPH, Science and Medicine Advisor, Office of Science Operations
Jody Hundley, Production Manager, Scientific Publications, Office of Science Operations
WRITING COMMITTEE MEMBERS*
↵* The findings and conclusions in this report are those of the authors and do not necessarily represent the official positions of the US Food and Drug Administration.
ACC/AHA TASK FORCE ON CLINICAL DATA STANDARDS MEMBERS
William S. Weintraub, MD, MACC, FAHA, Chair; Biykem Bozkurt, MD, PhD, FACC, FAHA; Gregg C. Fonarow, MD, FACC, FAHA; Robert C. Hendel, MD, FACC, FAHA†; Jeffrey P. Jacobs, MD, FACC; Hani H. Jneid, MD, FACC, FAHA; Michael A. Kutcher, MD, FACC; Judith H. Lichtman, MPH, PhD; Eric E. Smith, MD, MPH, FAHA‡; James E. Tcheng, MD, FACC; Tracy Y. Wang, MD, FACC, FAHA
†Former Task Force Chair; current chair during the writing effort.
‡Former Task Force member; current member during the writing effort.
This document was approved by the American Heart Association Science Advisory and Coordinating Committee on June 13, 2014, and by the American College of Cardiology Board of Trustees on November 12, 2014.
The online-only Comprehensive RWI Data Supplement table is available with this article at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIR.0000000000000156/-/DC1.
Two online-only Data Supplements are available with this article at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIR.0000000000000156/-/DC2 and http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIR.0000000000000156/-/DC3.
The American Heart Association requests that this document be cited as follows: Hicks KA, Tcheng JE, Bozkurt B, Chaitman BR, Cutlip DE, Farb A, Fonarow GC, Jacobs JP, Jaff MR, Lichtman JH, Limacher MC, Mahaffey KW, Mehran R, Nissen SE, Smith EE, Targum SL. 2014 ACC/AHA key data elements and definitions for cardiovascular endpoint events in clinical trials: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Data Standards (Writing Committee to Develop Cardiovascular Endpoints Data Standards). Circulation. 2015;132:302–361.
This article has been copublished in the Journal of the American College of Cardiology.
Copies: This document is available on the World Wide Web sites of the American College of Cardiology (www.cardiosource.org) and the American Heart Association (my.americanheart.org). A copy of the document is available at http://my.americanheart.org/statements by selecting either the “By Topic” link or the “By Publication Date” link. To purchase additional reprints, call 843-216-2533 or e-mail .
Expert peer review of AHA Scientific Statements is conducted by the AHA Office of Science Operations. For more on AHA statements and guidelines development, visit http://my.americanheart.org/statements and select the “Policies and Development” link.
Permissions: Multiple copies, modification, alteration, enhancement, and/or distribution of this document are not permitted without the express permission of the American Heart Association. Instructions for obtaining permission are located at http://www.heart.org/HEARTORG/General/Copyright-Permission-Guidelines_UCM_300404_Article.jsp. A link to the “Copyright Permissions Request Form” appears on the right side of the page.
- © 2015 American Heart Association, Inc.
- 1.↵Health Insurance Portability and Accountability Act of 1996. Public Law 104-191. 1996.
- Hicks KA,
- Hung HM,
- Mahaffey KW,
- et al
- Anderson JL,
- Adams CD,
- Antman EM,
- et al
- Cutlip DE,
- Windecker S,
- Mehran R,
- et al
- Dickstein K,
- Cohen-Solal A,
- Filippatos G,
- et al
- Easton JD,
- Saver JL,
- Albers GW,
- et al
- Hiatt WR,
- Goldstone J,
- Smith SC Jr.,
- et al
- Hunt SA,
- Abraham WT,
- Chin MH,
- et al
- Jneid H,
- Anderson JL,
- Wright RS,
- et al
- Thygesen K,
- Alpert JS,
- Jaffe AS,
- et al
- Hendel RC,
- Bozkurt B,
- Fonarow GC,
- et al
- Cannon CP,
- Brindis RG,
- Chaitman BR,
- et al
- Creager MA,
- Belkin M,
- Bluth EI,
- et al
- O’Gara PT,
- Kushner FG,
- Ascheim DD,
- et al
- Sacco RL,
- Kasner SE,
- Broderick JP,
- et al
- Weintraub WS,
- Karlsberg RP,
- Tcheng JE,
- et al
- Moussa ID,
- Klein LW,
- Shah B,
- et al
- Rankin J
- Drazner MH,
- Hellkamp AS,
- Leier CV,
- et al
- Moe GW,
- Howlett J,
- Januzzi JL,
- et al
- Nagueh SF,
- Bhatt R,
- Vivo RP,
- et al
- Alderman EL,
- Stadius M
- 49.CDISC Glossary. Available at: http://cdisc.org/therapeutic. Accessed December 12, 2014.
- 50.LOINC Test. Available at: http://www.loinc.org. Accessed December 12, 2014.
- van Swieten JC,
- Koudstaal PJ,
- Visser MC,
- et al
- Radford MJ,
- Arnold JM,
- Bennett SJ,
- et al
- 53.↵NCDR CathPCI Registry v4.4 Coder’s Data Dictionary. Available at: https://www.ncdr.com/WebNCDR/docs/public-data-collection-documents/cathpci_v4_codersdictionary_4-4.pdf?sfvrsn=2. Accessed December 12, 2014.
- Levine GN,
- Bates ER,
- Blankenship JC,
- et al
- Butman SM
- 56.National Heart, Lung, and Blood Institute. Coronary artery angiographic changes after percutaneous transluminal coronary angioplasty. In: Manual of Operations: NHLBI PTCA Registry. Bethesda, MD: National Heart, Lung, and Blood Institute; 1985:6–9.
- 59.STS/ACC TVT Registry v2.0 Coder’s Data Dictionary. Available at: https://www.ncdr.com/TVT/Libraries/TVT_Library/2_0_CoderDataDictionary.sflb.ashx. Accessed December 12, 2014.