ACCF/HRS/AHA/ASE/HFSA/SCAI/SCCT/SCMR 2013 Appropriate Use Criteria for Implantable Cardioverter-Defibrillators and Cardiac Resynchronization Therapy: A Report of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, Heart Rhythm Society, American Heart Association, American Society of Echocardiography, Heart Failure Society of America, Society for Cardiovascular Angiography and Interventions, Society of Cardiovascular Computed Tomography, and Society for Cardiovascular Magnetic Resonance FREE

Steven R. Bailey, MD, FACC, FSCAI, FAHA, Moderator

Andrea M. Russo, MD, FACC, FHRS, Writing Group Liaison⁎

Suraj Kapa, MD, Writing Group Liaison

Michael B. Alexander, MD, FACC§Health Plan Representative

Steven R. Bailey, MD, FACC, FSCAI, FAHA∥American College of Cardiology Foundation Representative

Ulrika Birgersdotter-Green, MD, FHRS∥

Alan S. Brown, MD, FACC, FAHA, FNLA∥

Richard A. Grimm, DO, FACC, FASE¶American Society of Echocardiography Representative

Paul J. Hauptman, MD#Heart Failure Society of America Representative

Sharon A. Hunt, MD, FACC#

Rachel Lampert, MD, FACC, FHRS⁎

JoAnn Lindenfeld, MD, FACC⁎⁎American Heart Association Representative

David J. Malenka, MD, FACC∥

Kartik Mani, MD††Society for Cardiovascular Angiography and Interventions Representative

Joseph E. Marine, MD, FACC, FHRS⁎

Edward T. Martin, MD, FACC, FACP, FAHA‡‡Society for Cardiovascular Magnetic Resonance Representative

Richard L. Page, MD, FACC, FHRS, FAHA∥

Michael W. Rich, MD, FACC§§American Geriatrics Society Representative

Paul D. Varosy, MD, FACC, FHRS⁎

Mary Norine Walsh, MD, FACC∥

Michael J. Wolk, MD, MACC, Chair

Steven R. Bailey, MD, FACC, FSCAI, FAHA

John U. Doherty, MD, FACC

Pamela S. Douglas, MD, MACC, FAHA

Robert C. Hendel, MD, FACC, FAHA, FASNC

Christopher M. Kramer, MD, FACC

James K. Min, MD, FACC

Manesh R. Patel, MD, FACC

Leslee Shaw, PhD, FACC, FASNC

Raymond F. Stainback, MD, FACC, FASE

Joseph M. Allen, MA

• Abstract…...1320

• Preface…...1321

• 1Introduction…...1321
• 2Methods…...1322

  • Indication Development…...1322

  • Rating Process and Scoring…...1322

• 3Assumptions…...1323

  • General Clinical Assumptions…...1323

  • Practice Parameters/Standard of Care…...1324

  • Cost/Value…...1325

  • Guidance Specifically for AUC Users…...1325

• 4Definitions…...1325
• 5Abbreviations…...1328
• 6Results of Ratings…...1328
• 7Appropriate Use Criteria for ICD/CRT Indications…...1329

  • Section 1: Secondary Prevention ICD…...1329

    • (Table 1). CAD: VF or Hemodynamically Unstable VT Associated With Acute (<48 h) MI (Newly Diagnosed, No Prior Assessment of LVEF)…...1329

    • (Table 2). CAD: VF or Hemodynamically Unstable VT <48 h (Acute) Post-Elective Revascularization…...1329

    • (Table 3). CAD: VF or Hemodynamically Unstable VT (No Recent MI [≤40 Days] Before VF/VT and/or No Recent Revascularization [≤3 Months] Before VF/VT)…...1329

    • (Table 4). CAD: VF or Hemodynamically Unstable VT During Exercise Testing Associated With Significant CAD…...1330

    • (Table 5). No CAD: VF or Hemodynamically Unstable VT…...1330

    • (Table 6). Genetic Diseases with Sustained VT/VF…...1330

    • (Table 7). No Structural Heart Disease (LVEF ≥50%) or Known Genetic Causes of Sustained VT/VF…...1331

    • (Table 8). Syncope in Patients Without Structural Heart Disease…...1331

    • (Table 9). Syncope in Patients With Coronary Artery Disease…...1332

    • (Table 10). Syncope in Patients With Nonischemic Structural Heart Disease…...1332

    • (Table 11). Sustained Hemodynamically Stable Monomorphic VT Associated With Structural Heart Disease…...1333

  • Section 2: Primary Prevention ICD…...1333

    • (Table 12). Post-Acute Myocardial Infarction (≤40 Days) LVEF ≤30%…...1333

    • (Table 13). Post-Acute Myocardial Infarction (≤40 Days) LVEF 31% to 40%…...1334

    • (Table 14). Post-Acute Myocardial Infarction (≤40 Days) and Pre-Existing Chronic Cardiomyopathy (≥3 Months)…...1334

    • (Table 15). Post-Myocardial Infarction (≤40 Days) and Need for Guideline-Directed Pacemaker Therapy Post-MI (e.g., SSS, CHB, or Other Indications for Permanent Pacemaker)…...1334

    • (Table 16). Post-Myocardial Infarction (>40 Days) With Ischemic Cardiomyopathy…...1335

    • (Table 17). Duration of Guideline-Directed Medical Therapy for Ischemic Cardiomyopathy Without Recent MI (Revascularization Not Indicated)…...1335

    • (Table 18). Nonischemic Cardiomyopathy…...1336

    • (Table 19). Genetic Conditions (Excludes Syncope and Sustained VT, Covered in Section 1)…...1336

  • Section 3: Comorbidities…...1337

    • (Table 20). Special Conditions/Comorbidities in Patients for Primary Prevention (Meeting Indications of ICD Implant Related to HF Diagnosis With LVEF ≤30% on Guideline-Directed Medical Therapy >3 Months)…...1337

  • Section 4: ICD Generator Replacement at Elective Replacement Indicator (ERI)…...1338

    • (Table 21). Primary Prevention ICD at Initial Implant…...1338

    • (Table 22). Secondary Prevention ICD at Initial Implant…...1338

    • (Table 23). Primary Prevention at Initial Implant: Replacement of CRT-ICD for ERI…...1339

    • (Table 24). Secondary Prevention at Initial Implant: Replacement of CRT-ICD for ERI…...1339

  • Section 5: Dual-Chamber ICD (As Opposed to Single-Chamber ICD for Patients Who Meet Criteria for ICD Implantation)…...1339

    • (Table 25). Conduction System Abnormalities…...1340

    • (Table 26). No Conduction Abnormalities…...1341

    • (Table 27). Tachyarrhythmias…...1341

    • (Table 28). Other Disorders…...1341

  • Section 6: CRT—No Prior Implant…...1342

    • (Table 29). Ischemic Cardiomyopathy…...1342

    • (Table 30). Nonischemic Cardiomyopathy…...1343

    • (Table 31). LVEF >35% of Any Etiology (ICD Indicated)…...1343

    • (Table 32). LVEF ≤35% of Any Etiology…...1344

    • (Table 33). Pre-Existing or Anticipated RV Pacing With a Clinical Indication for ICD or Pacemaker Implantation…...1344

    • (Table 34). Refractory Class III/IV HF <3 Months Post-Revascularization and/or ≤40 Days Post-MI…...1344

• 8Figures…...1345

  • • (Figure 17_gr1). Secondary Prevention: CAD—VF or Hemodynamically Unstable VT Associated With Acute (<48 h) MI…...1345

    • (Figure 17_gr2). Secondary Prevention: VF or Hemodynamically Unstable VT With No Recent MI and/or No Recent Revascularization…...1345

    • (Figure 17_gr3). Secondary Prevention: VF or Hemodynamically Unstable VT—No CAD With Structural Heart Disease or Genetic Disorders…...1346

    • (Figure 17_gr4). Secondary Prevention: No Structural Heart Disease (LVEF ≥50%) or Known Genetic Causes of Sustained VT/VF…...1346

    • (Figure 17_gr5). Secondary Prevention: Syncope in Patients Without Structural Heart Disease…...1347

    • (Figure 17_gr6). Secondary Prevention: Syncope in Patients With Coronary Artery Disease…...1347

    • (Figure 17_gr7). Secondary Prevention: Syncope in Patients With Nonischemic Structural Heart Disease…...1348

    • (Figure 17_gr8). Secondary Prevention: Sustained Hemodynamically Stable Monomorphic VT Associated With Structural Heart Disease…...1348

    • (Figure 17_gr9). Primary Prevention: Coronary Artery Disease, Post-Acute MI (≤40 Days), LVEF ≤40%…...1349

    • (Figure 17_gr10). Primary Prevention: Coronary Artery Disease, Prior MI (>40 Days) With Ischemic Cardiomyopathy…...1349

    • (Figure 17_gr11). Primary Prevention: Nonischemic Cardiomyopathy…...1350

    • (Figure 17_gr12). Primary Prevention: Nonischemic Cardiomyopathy, Specific Etiologies…...1350

    • (Figure 17_gr13). Primary Prevention: Genetic Conditions (Excludes Syncope and Sustained VT)…...1351

    • (Figure 17_gr14). Primary Prevention: Comorbidities (Rarely Appropriate Indications)…...1351

    • (Figure 17_gr15). CRT: No Prior Implant—Ischemic Cardiomyopathy…...1352

    • (Figure 17_gr16). CRT: No Prior Implant—Nonischemic Cardiomyopathy…...1352

    • (Figure 17_gr17). CRT: No Prior Implant—LVEF >35% of Any Etiology (ICD Indicated)…...1353

    • (Figure 17_gr18). CRT: No Prior Implant—LVEF ≤35% of Any Etiology (NYHA Functional Class IV on IV Inotropic Support)…...1353

    • (Figure 17_gr19). CRT: No Prior Implant—Pre-Existing or Anticipating RV Pacing With a Clinical Indication for ICD or Pacemaker Implantation…...1354

    • (Figure 17_gr20). Refractory Class III/IV Heart Failure <3 Months Post-Revascularization and/or ≤40 Days Post-MI…...1354

• 9Discussion…...1355
  • 9.1ICDs: Initial Implantation…...1355
  • 9.2CRT Devices…...1358
  • 9.3Generator Replacement…...1358
  • 9.4Dual-Chamber Versus Single-Chamber ICDs…...1359
  • 9.5Application of Pre-Specified Cutoffs…...1359
  • 9.6Clinical Judgment and the Understanding of AUC Ratings…...1360
  • 9.7Reimbursement and Disclaimer…...1360
  • 9.8Application of Criteria…...1360

• Appendix A: Additional Methods…...1361

  • Relationships With Industry and Other Entities…...1361

  • Literature Review…...1361

• Appendix: B: ACCF/HRS/AHA/ASE/HFSA/SCAI/SCCT/SCMR 2013 Appropriate Use Criteria for ICD/CRT Participants…...1361

  Appendix C: ACCF/HRS/AHA/ASE/HFSA/SCAI/SCCT/SCMR 2013 Appropriate Use Criteria for ICD/CRT Participants—Relationships With Industry and Other Entities (Relevant)…...1364

The American College of Cardiology Foundation in collaboration with the Heart Rhythm Society and key specialty and subspecialty societies conducted a review of common clinical scenarios where implantable cardioverter-defibrillators (ICDs) and cardiac resynchronization therapy (CRT) are frequently considered. The clinical scenarios covered in this document address secondary prevention, primary prevention, comorbidities, generator replacement at elective replacement indicator, dual-chamber ICD, and CRT.

The indications (clinical scenarios) were derived from common applications or anticipated uses, as well as from current clinical practice guidelines and results of studies examining device implantation. The 369 indications in this document were developed by a multidisciplinary writing group and scored by a separate independent technical panel on a scale of 1 to 9 to designate care that is Appropriate (median 7 to 9), May Be Appropriate (median 4 to 6), and Rarely Appropriate (median 1 to 3). The final ratings reflect the median score of the 17 technical panel members: 45% of the indications were rated as Appropriate, 33% were rated May Be Appropriate and 22% were rated Rarely Appropriate.

In general, Appropriate designations were assigned to scenarios for which clinical trial evidence and/or clinical experience was available that supported device implantation. By contrast, scenarios for which clinical trial evidence was limited or device implantation seemed reasonable for extenuating reasons were categorized as May Be Appropriate. Scenarios for which there were data showing harm, or no data were available, and medical judgment deemed device therapy ill-advised were categorized as Rarely Appropriate. For example, comorbidities including life expectancy and cognitive function impacted appropriateness ratings.

The Appropriate Use Criteria for ICD/CRT have the potential to enhance physician decision making, healthcare delivery, and reimbursement policy. Furthermore, recognition of clinical scenarios rated as May Be Appropriate facilitates the identification of areas that would benefit from future research.

In an effort to respond to the need for the rational use of cardiovascular services including imaging and invasive procedures in the delivery of high-quality care, the American College of Cardiology Foundation (ACCF) in collaboration with the Heart Rhythm Society (HRS) has undertaken a process to describe the appropriate use of implantable cardioverter-defibrillators (ICDs) and cardiac resynchronization therapy (CRT) for selected patient populations.

Appropriate use criteria (AUC) publications reflect an ongoing effort by the ACCF to critically and systematically create, review, and categorize clinical situations where physicians utilize diagnostic tests and procedures to care for patients with cardiovascular disease. The process is based on current understanding of the benefits and risks of the procedures examined. Although impossible to be entirely comprehensive given the wide diversity of clinical disease, the indications are meant to identify common clinical scenarios encompassing the majority of situations encountered in contemporary practice. Given the breadth of information they convey, the indications do not directly correspond to the Ninth Revision of the International Classification of Diseases system, as these codes do not include clinical information such as disease severity or symptom status.

The ACCF and HRS believe that careful blending of a broad range of clinical experiences and available evidence-based information will help guide a more efficient and equitable allocation of healthcare resources in cardiovascular care and device implantation. The ultimate objective of AUC is to improve patient care and health outcomes in a cost-effective manner, but it is not intended to ignore uncertainty and nuance intrinsic to clinical decision making. Therefore, AUC should not be considered substitutes for sound clinical judgment and practice experience.

The AUC process itself is also evolving. The initial AUC documents were directed primarily towards noninvasive cardiovascular imaging tests. Revisions to several of these imaging documents have already been published ((1),(2),3). The goal for the AUC process is to provide contemporary reference documents that incorporate new research in a timely manner, including the results of AUC implementation studies. AUC for ICD and CRT is the third in a more recent series of AUC documents that examine the use of invasive procedures ((4),5). Because ICDs and CRT play a central role in the care of patients with cardiovascular disease, guidance around the rationale and evidence-based use of the procedure is the goal of the current document.

Andrea M. Russo, MD, FACC, FHRS Co-Chair, ICD/CRT Writing Group

Raymond F. Stainback, MD, FACC, FASE Co-Chair, ICD/CRT Writing Group

Michael J. Wolk, MD, MACC Chair, Appropriate Use Criteria Task Force

The American College of Cardiology Foundation (ACCF), in collaboration with the Heart Rhythm Society (HRS), developed common clinical scenarios where implantable cardioverter-defibrillators (ICDs) and cardiac resynchronization therapy (CRT), also known as biventricular pacing, are frequently considered. These implanted devices are also collectively referred to as cardiovascular implantable electronic devices (CIEDs). The indications, as presented in these clinical scenarios, were derived from common presentations or anticipated uses, as well as from current clinical practice guidelines. The 369 indications in this document were developed by a writing group with diverse clinical expertise and rated by a separate independent technical panel on a scale of 1 to 9, to designate care that is Appropriate (median 7 to 9), May Be Appropriate (median 4 to 6), and Rarely Appropriate (median 1 to 3). Members of the writing group and the technical and review panels were selected in large part because of their active involvement in the clinical practice of electrophysiology, heart failure, and other related areas of cardiovascular medicine.

Describing the appropriate use criteria (AUC) for ICD and CRT has the potential to enhance physician decision making, healthcare delivery, and reimbursement policy. Furthermore, recognition of clinical scenarios categorized as May Be Appropriate facilitates identification of areas that would benefit from future research.

This report addresses the appropriate use of ICD and CRT. Determination of the criteria for the implantation of ICDs and/or CRT is based on the results of evidence derived from clinical trials. The same evidence has been incorporated into clinical practice guidelines. However, there is also recognition that in “real world” scenarios, expert opinion is of value in addressing patient populations that are either not represented in currently available randomized clinical trials or for treatment recommendations that are supported by lower levels of evidence. In addition, it is important to recognize that when patients are excluded from a clinical trial, the results of the trial should not be interpreted to mean that the treatment was proven to be ineffective for patients who were excluded. Physicians must use their best judgment in deciding whether a treatment might be beneficial to patients who would have been excluded from a clinical trial. Accordingly, the AUC were designed to include a broad spectrum of clinical scenarios representative of those encountered by physicians in their daily practice. The AUC are also intended to highlight areas of potential misapplication of technology (overutilization) in addition to areas of possible underutilization. For practical implementation, the document was not designed to be all encompassing, and therefore, the writing group focused on the more commonly encountered scenarios. As with other AUC documents, there is an implicit acknowledgment that important deficiencies may be revealed by subsequent clinical trials and AUC implementation studies, which will not only guide further research but also inform expedient updates in the AUC for ICD/CRT. As the field advances, the healthcare community needs to understand how to best incorporate this technology into daily clinical care. The ACCF and the HRS are dedicated to this effort.

A detailed description of the methods that were used for rating the selected clinical indications can be found in a previous publication, “ACCF Proposed Method for Evaluating the Appropriateness of Cardiovascular Imaging” (6). Briefly, this process combines evidence-based medicine and practice experience by engaging a technical panel in a prospective, modified Delphi exercise. The technical panel was created through nominations given by multiple relevant professional societies and provider-led organizations as well as from health policy and payer communities. To preserve objectivity, technical panels are created so as to not include a majority of individuals whose livelihood is tied to the technology under evaluation. During the development of this document, the AUC Task Force finalized a revision of the terminology and definitions to better clarify the appropriateness categories (7). As a result, the technical panel used the terminology described in the original methodology paper for all but the last round of rating. Further explanation of this change is provided in the following text.

In making its appropriateness determinations, the technical panel was provided with relevant evidence from the medical literature and practice guidelines. Technical panelists were asked to individually assess the benefits and risks of device implantation. Then, following a group discussion of the indications and related considerations, a second individual rating process was performed to determine the final ratings. After the rating process, the final appropriate use ratings were summarized using an established rigorous methodology (8).

The indications were constructed by a writing group with expertise in both the science and clinical practice of electrophysiology, heart failure, general cardiology, invasive cardiology, and noninvasive cardiac imaging. The writing group was tasked with developing a list of clinical scenarios covering the majority of patients that clinicians might consider referring for device implantation. The term “indication” is used interchangeably with “clinical scenario” in the document for brevity and does not imply that a procedure should necessarily be performed. Indication modifications were made through discussions with the ACCF AUC Task Force and feedback from reviewers that included additional experts in the areas noted in the preceding text, along with experts in the fields of geriatric medicine, internal medicine, and clinical outcomes research.

The indications included in this publication incorporate a wide range of cardiovascular signs, symptoms, disease states, and physiological assessments, including, but not limited to, measurement of the left ventricular ejection fraction (LVEF), duration of the QRS complex, monitoring data, and results of electrophysiological studies. Within each main disease category, a standardized approach was used to capture the majority of clinical scenarios with an attempt to avoid making the list of indications excessive. This document does not cover indications for implantation of devices in the pediatric population.

Wherever possible, indications were mapped to relevant clinical guidelines and key publications/references (see Guideline Mapping and References Online Appendix).

The technical panel first rated the indications independently. Then, the technical panel convened for a face-to-face meeting to discuss each indication. At this meeting, panelists were provided with their scores and a blinded summary of their peers' scores. After the meeting, panelists once again independently rated each indication to indicate their final scores. The technical panel completed an additional rating process to address a few areas that required further clarification and readdressed ratings following the introduction of revised terminology as outlined in the recently updated methods document (7).

The members of the technical panel completed the rating process using the old terminology and definitions (appropriate, uncertain, and inappropriate, as described in the original methods [6]), and were subsequently asked to participate in an additional round of rating to re-examine the indications using the new terminology and expanded definitions. The new terminology was finalized after the technical panel had completed the rating process. As a consequence, the additional round of rating was deemed necessary to minimize potential confusion related to differences in terminology between the old and updated methodology and to assess whether the change could impact the appropriateness classification of the different clinical scenarios. The final rating of the indications using the revised terminology and expanded definitions resulted in a change in ratings for only 1% of the total indications.

When rating each clinical scenario, the technical panel was asked to assess whether device implantation is Appropriate, May Be Appropriate, or Rarely Appropriate, according to the following definition of appropriate use: ICD and/or CRT implantation is appropriate in general when the expected value in terms of survival and/or other health benefits (symptoms, functional status, and/or quality of life) exceed the potential adverse health consequences relating to the acute procedural risk and the long-term consequences of living with an implanted device.

The technical panel scored each indication as follows:

Median Score 7 to 9: Appropriate care

An appropriate option for management of patients in this population due to benefits generally outweighing risks; effective option for individual care plans, although not always necessary, depending on physician judgment and patient-specific preferences (i.e., procedure is generally acceptable and is generally reasonable for the indication).

Median Score 4 to 6: May Be Appropriate care

At times an appropriate option for management of patients in this population due to variable evidence or agreement regarding the benefit/risk ratio, potential benefit based on practice experience in the absence of evidence, and/or variability in the population; effectiveness for individual care must be determined by a patient's physician in consultation with the patient based on additional clinical variables and judgment along with patient preferences (i.e., procedure may be acceptable and may be reasonable for the indication).

Median Score 1 to 3: Rarely Appropriate care

Rarely an appropriate option for management of patients in this population due to the lack of a clear benefit/risk advantage; rarely an effective option for individual care plans; exceptions should have documentation of the clinical reasons for proceeding with this care option (i.e., procedure is not generally acceptable and is not generally reasonable for the indication).

The division of these scores into 3 levels of appropriateness should be viewed as a continuum. When there is diversity in opinion regarding the management of a particular clinical scenario such that scores fall in the intermediate level of appropriateness, they are labeled May Be Appropriate, as critical patient information or research data may be lacking or discordant. This must not be treated as either Appropriate or Rarely Appropriate, but rather as a distinct category of May Be Appropriate. It is anticipated that the AUC standards will continue to be revised as further data are generated and information from the implementation of the criteria is accumulated.

The level of agreement among panelists as defined by RAND (8) is analyzed based on the BIOMED rule. For each clinical scenario, the voting process produces a result in which there is either mathematical agreement or disagreement among panelists. Agreement exists when 4 or fewer panelists' ratings fell outside the 3-point region containing the median score.

Disagreement exists when 5 or more panelists' ratings fall in both the Appropriate and the Rarely Appropriate categories. Any indication having disagreement will be placed in the May Be Appropriate category regardless of the final median score. The final scores were obtained after the panel had the opportunity to discuss the clinical scenarios at a face-to-face meeting that was followed by a second-round rating to eliminate the possibility of misinterpretation of either the indication wording or the published clinical data.

To limit inconsistencies in interpretation, specific assumptions were considered by the writing group in development and were used by the technical panel in rating the clinical indications for the appropriate use of device implantation. Other assumptions also reviewed in the Discussion relate to the interpretation of AUC results and implementation.

• 1For each indication, the rating should reflect whether device implantation is reasonable for the patient according to the appropriate use definition. It should not be assumed that for each indication the decision to treat has already been made.
• 2A qualified clinician has completed a thorough clinical history and physical examination such that the clinical status of the patient can be assumed to be valid as stated in the indication. For example, a patient said to be asymptomatic is truly asymptomatic for the condition in question, and sufficient questioning of the patient has been undertaken.
• 3End-of-life discussion, advanced directive, and patient consent have been adequately addressed. Patients are assumed to be candidates for ICD/CRT only after a decision-making discussion has been undertaken between the patient, appropriate family and/or legal decision makers, and the physician. It is assumed that the patient and/or decision makers are educated sufficiently to understand whether or not ICD/CRT implantation is consistent with current care intentions or with prior advance directives.
• 4The clinical scenarios should be preferentially rated based on evidence from published literature and clinical practice guidelines regarding the risks and benefits of ICD/CRT. Selected specific patient groups not well represented in the literature or in clinical practice guidelines are presented in many of the current clinical scenarios because the writing group recognizes that decisions about device implantation in such patients are frequently required. Examples of such patients include those with end-stage renal disease or advanced age.
• 5All patients are receiving optimal care, also called “guideline-directed medical therapy” (GDMT) in ACC/AHA Clinical Practice Guidelines, including guideline-based risk factor modification for primary or secondary prevention for coronary artery disease (CAD) and heart failure in cardiovascular patients unless specifically noted (9).
• 6There are no unusual extenuating logistical or process-of-care circumstances such as inability to comply with follow-up due to any number of reasons (e.g., mental instability, lack of transportation) unless specifically noted.
• 7There are no technical limitations for device implantation or other comorbidities that are likely to substantially increase procedural risk, unless specifically noted.
• 8Coronary artery disease: for sections that reference revascularization, additional assumptions may apply, including but not limited to the following:
  • aFor scenarios in which no revascularization is planned, it should be assumed that revascularization is not indicated unless otherwise specified, for example, there are no major epicardial coronary lesions measuring ≥70% (non-left main) or ≥50% (left main) or no evidence of ischemia by fractional flow reserve or perfusion imaging.
  • bOther scenarios may include cases where patients are not candidates for revascularization for whatever reason, including but not limited to severe, diffuse CAD that is not amenable to revascularization.
  • cWhen revascularization is considered or performed, it is assumed that patients are also acceptable candidates for revascularization based on the absence of other noncardiac comorbidities that would be a contraindication for revascularization.
  • dIf patients are candidates for revascularization, and revascularization is planned, electrophysiology (EP) testing should not be performed until after the intended revascularization is performed.
  • eAn ICD should not be implanted before revascularization to circumvent the current Centers for Medicare & Medicaid Services 3-month waiting-period rule ((10),11).
• 9An assessment of the LVEF during hospitalization following acute infarction or revascularization generally prompts consideration of ICD/CRT implantation. When a subsequent waiting period is required (e.g., after guideline-directed medical therapy, myocardial infarction, or revascularization), it is assumed that the final decision to treat will be based on a follow-up LVEF assessment after expiration of the waiting period, and that the imaging facility understands that quantitative measurement of the LVEF is an important goal of the exam.

• 10Operators performing device implantation have appropriate clinical training (12) and experience consistent with established standards of care and have satisfactory outcomes as assessed by quality assurance monitoring, including national benchmark data from the National Cardiovascular Data Registry (NCDR) ICD registry (13). ICDs are implanted with transvenous electrodes. Although different means of delivering electrical therapy have only recently become available, specifically, a totally subcutaneous ICD system, a standard transvenous approach was assumed for the purpose of this document.
• 11Geographic/regional variability: issues of local availability of skill in performing the procedure should not be considered during the rating process, as it is assumed that skilled operators and appropriate implantation resources are locally available.
• 12Adjunctive cardiac imaging modalities are often required for appropriate patient selection. These may include coronary angiography or cardiac computed tomography for the determination of coronary anatomy in addition to other noninvasive cardiac imaging modalities, including echocardiography, cardiac magnetic resonance imaging, and radionuclide imaging for initial assessment of cardiac structure and function (LVEF) and when needed, a follow-up determination of the LVEF. It assumed that laboratories performing these services have appropriate clinical training and experience, perform these studies and interpret them according to national standards, and have satisfactory outcomes as assessed by quality improvement monitoring.
• 13It is recognized that there may be variability in the measurement of the LVEF at different points in time and utilizing different imaging modalities. The labs performing the LVEF assessments will have quality assurance measures in place to ensure accuracy of each individual method for determining and reporting left ventricular (LV) function.
• 14For all indications, it is assumed that the LVEF stated in the indications was measured within a timeframe relevant to making the decision about eligibility for ICD implantation. It is assumed that repeat evaluation of the LVEF will be performed after an appropriate duration of time following recovery from myocardial infarction or revascularization, or following GDMT in the setting of a new diagnosis of heart failure or cardiomyopathy, before determining ICD eligibility.
• 15All procedures presented are to be considered for clinical indications and not as part of a research protocol.
• 16With respect to CRT, atrial arrhythmias (including atrial fibrillation, atrial flutter, and atrial tachycardia) are not included in the indication tables. There are fewer data available for CRT in patients with persistent atrial arrhythmias, and the writing group elected to avoid additional scenarios for practical reasons, as the document already includes a large number of scenarios. However, it is assumed that the presence of intermittent or persistent atrial arrhythmias would not preclude CRT implantation, and the benefits of CRT would also apply to patients with persistent atrial arrhythmias, as long as CRT is maintained nearly 100% of the time.
• 17The potential adverse effects of right ventricular (RV) pacing in the setting of pre-existing LV systolic dysfunction are well described ((14),(15),16). Therefore, attempts should be made to reduce unnecessary RV pacing by appropriate programming of single- and dual-chamber ICDs, whenever possible.
• 18Single- versus dual-chamber ICD selection: It is assumed that most patients undergoing ICD implantation who have standard dual-chamber pacing indications will undergo attempted insertion of an atrial lead as described in the 2008 ACC/AHA/HRS device-based therapy guidelines document, and a separate consensus document pertaining to selection of dual- versus single-chamber devices for pacemaker patients, which was recently published (17). However, there is currently controversy regarding single- versus dual-chamber device selection in patients who do not meet strict pacing indications but are undergoing ICD implantation without CRT, and this is an area of ongoing investigation. For example, it has been hypothesized that the availability of dual-chamber discriminators might improve discrimination of ventricular arrhythmias from supraventricular arrhythmias and thus potentially reduce unnecessary ICD shocks. However, a recent meta-analysis demonstrated that the proportion of patients receiving inappropriate therapy was not different between single- and dual-chamber devices using technology available at the time (18). It is currently unknown whether recent advancements in technology utilizing the most current ICD systems will show any benefit of dual-chamber devices in the absence of standard pacing indications, and studies evaluating this controversial topic are currently ongoing. Because there is a difference in cost and a potential difference in longevity of single- versus dual-chamber devices, and dual-chamber systems may potentially improve discrimination between ventricular and supraventricular arrhythmias, but have a higher risk of dislodgment due to the addition of the atrial lead, these scenarios were felt to be important to address in this document.
• 19Decisions for ICD implantation should be based on a reasonable expectation of survival with a good functional status for at least 1 year. The clinical trial populations used to derive published predictive survival models may differ from the general heart failure population with regard to age and comorbidities. Therefore, consideration should be given to advanced age or other comorbidities that might reduce the likelihood of benefit or increase the risk of ICD therapy for individuals.

• 20From the standpoint of the practicing physician caring for an individual patient, potential clinical benefits of device implantation should be the highest priority, and this is weighed against potential risks of the procedure. As related to societal benefits, costs should also be considered in relationship to potential benefits in order to better understand comparative value. However, very little has been done to assess cost effectiveness of ICD or CRT treatment across a spectrum of conditions and comorbidities. Although cost and value are clearly important variables, which are also relevant to payers and policymakers, it is recognized that healthcare providers typically do not primarily base individual patient decisions about device implantation on these considerations. Therefore, it is anticipated that technical panel members rate the scenarios primarily based on risks/benefits, although cost/value considerations may also be taken into consideration if deemed Appropriate by panel members for particular scenarios.

• 21Reducing care that is Rarely Appropriate remains a valuable means to reduce costs and population risks of ICD and/or CRT implantation.
• 22The category of May Be Appropriate should be used when insufficient clinical data are available for a definitive categorization, or there are substantial differences in opinion regarding the appropriateness of that indication. The absence of definitive data supporting implantation in a particular subset of patients does not imply lack of benefit, and in such cases, careful investigation of the particulars of the clinical scenario is warranted. The designation of May Be Appropriate should not be used as the sole grounds for denial of reimbursement in an individual patient.

Definitions of terms used throughout the indication set are listed here.

The duration of heart failure symptoms is defined as the duration of symptoms since the initial diagnosis of heart failure to the date of the device implantation. Clinical trials and the NCDR ICD registry have utilized time frames of <3 months, 3 to 9 months, and >9 months. The writing group recognizes that 3 months may equate to more or less than 90 days, depending on the calendar months. The 3-month term was chosen because it was used in some randomized clinical trials related to timing for device implantation and is the basis of coverage in the 2005 National Coverage Determination of the Centers for Medicare & Medicaid Services for nonischemic dilated cardiomyopathy.

Dyssynchrony refers to “ventricular electromechanical delay,” which may be identified by multiple imaging techniques, including echocardiography. Prolongation of the QRS complex is seen in approximately one-third of patients with advanced heart failure, and this prolongation may be associated with varying degrees of ventricular electromechanical delay or “dyssynchrony.” Modifications in this delay are often seen with CRT pacing or “resynchronization therapy.” Studies utilizing CRT have also been performed in patients with narrow QRS complexes in the presence of dyssynchrony. However, no proven benefit has been demonstrated in this cohort with a QRS duration <120 ms (19). Additionally, recent meta-analyses question the utility of CRT in patients with QRS durations of 120 to 149 ms ((20),21). Enrollment criteria for CRT trials have typically been based on a QRS duration ≥120 ms, regardless of imaging techniques to evaluate the presence or absence of dyssynchrony. There is current controversy on the role of dyssynchrony in assessing the likelihood of response to CRT, and this argues that dyssynchrony assessments should not be included in consideration for CRT implantation. Therefore, due to the enrollment criteria used in clinical trials and the absence of consensus at this time regarding its assessment, measurement of dyssynchrony prior to implantation is not included in the AUC scenarios listed in this document.

When tolerated, GDMT (sometimes referred to as “optimal medical therapy”) should include aspirin (or a thienoypyridine if aspirin is not tolerated), statin therapy, angiotensin-converting enzyme inhibition (or an angiotensin receptor blocker) and the use of beta-blockers after myocardial infarction. Therapy for angina/ischemia should include at least 1 of the following medications: beta-blockers, calcium channel antagonists, or nitrates. Therapy should also be directed at optimizing the treatment of associated conditions such as diabetes and uncontrolled hypertension.

GDMT for heart failure in the setting of LV systolic dysfunction requires individualization but typically should include the combination of an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker and beta-blocker therapy adjusted to target doses as tolerated, with diuretics adjusted if/as needed to control fluid retention. In selected patients, the addition of aldosterone antagonists and hydralazine plus nitrate combinations should be considered. Patients who are going to receive substantial benefit from medical treatment alone usually show some clinical improvement during the first 3 to 6 months. Medical therapy is also assumed to include adequate rate control for tachyarrhythmias, including atrial fibrillation. Therefore, it is recommended that GDMT be provided for at least 3 months before planned reassessment of LV function to consider device implantation. If LV function improves to the point where primary prevention indications no longer apply, then device implantation is not indicated.

Heart failure is defined as a clinical syndrome characterized by specific symptoms described in the medical history and signs on the physical examination. The clinical symptoms of heart failure may include dyspnea on exertion, orthopnea, fatigue, or fluid retention. The clinical signs may include jugular venous pressure elevation, râles, an S3 gallop, or lower extremity edema. A low LVEF or diagnosis of cardiomyopathy alone, or peripheral edema without other clinical signs of heart failure, does not qualify as heart failure (22).

Patients may experience periods of clinical instability with hypotension, heart failure symptoms, pre-syncope or syncope, angina, or dyspnea. These symptoms are presumed to result from hypo-perfusion, with a cardiac output and/or rhythm that is inadequate to support normal organ function.

Inducibility is defined as the induction of sustained ventricular tachycardia (VT) or ventricular fibrillation (VF) at EP testing with an arrhythmia duration ≥30 s and/or resulting in hemodynamic compromise using standardized stimulation protocols.

The “Universal Definition of Myocardial Infarction” was developed by Thygesen and colleagues in 2007 and updated in 2012. The multifaceted clinical criteria include timing, mechanism (infarct type), biomarker status, and size. An elevated troponin is not necessarily indicative of an acute MI ((23),(24),25).

Not infrequently, a low-level troponin elevation is detected when blood is drawn routinely or as a consequence of protocol laboratory testing. If upon further evaluation the troponin levels do not exhibit a typical rise and fall pattern, or there is an alternative explanation for the troponin leak (e.g., cardiac arrest or external defibrillation) that can be explained by a diagnosis other than myocardial ischemia, this should not be misconstrued as a myocardial infarction (as defined by Thygesen et al. [(23),25]) based on the laboratory test alone ((23),(25),26). In addition, a nonspecific transient biomarker elevation may also occur in some situations of cardiac arrest in which there is a low-level rise in troponin with subsequent fall, in the absence of coronary artery disease or thrombosis. This also should not be considered a myocardial infarction without underlying coronary obstruction, as this leak of troponin is likely related to the arrest itself. These low-level rises in biomarkers should not preclude ICD implantation, if criteria for implantation are otherwise met.

The definitions are included in the table in the following text. The patient's NYHA functional classification at the time of the decision to implant the device should be used for this classification. If the patient has left ventricular dysfunction, but no symptoms of heart failure, this should be coded as “class I.” If the patient is hospitalized for heart failure at the time the decision is made to implant the device, the NYHA functional class on optimized GDMT should be utilized.

Ambulatory class IV is defined as class IV heart failure with: 1) no active acute coronary syndrome; 2) no inotropes; and 3) on GDMT.

A normal LVEF is defined as ≥50%.

Secondary Prevention (Section 1 Indications):

Secondary prevention refers to an indication for an ICD exclusively for patients who have survived 1 or more cardiac arrests or sustained ventricular tachycardia. Patients with cardiac conditions associated with a high risk of sudden death who have unexplained syncope that is likely to be due to self-terminating ventricular arrhythmias are also considered to have a secondary prevention indication.

Primary Prevention (Section 2 Indications):

Primary prevention is an indication for an ICD to prevent sudden cardiac death. It refers to use of ICDs in individuals who are at risk for, but have not yet had, an episode of sustained VT, VF, or cardiac arrest.

A “narrow” QRS duration is <120 ms. A wide QRS is ≥120 ms and may have a left bundle branch block (LBBB), right bundle branch block (RBBB), or nonspecific intraventricular conduction delay morphology. For the purpose of this AUC document and for consistency with the focused update of the device-based therapy guidelines (29), “non-LBBB” morphology is used to refer to both RBBB and intraventricular conduction delay morphologies. For the purpose of CRT implantation, it is assumed that the wide QRS is present consistently, and does not represent an intermittent bundle branch block or intermittent QRS widening, thereby excluding QRS widening that is transient or rate-related. If there is discrepancy in the measurement of QRS duration on various electrocardiograms (ECGs), the most representative ECG obtained proximate to the final clinical decision-making process will be utilized to determine candidacy for CRT implantation.

Sudden cardiac arrest is defined as the sudden cessation of effective cardiac mechanical activity resulting in unresponsiveness, without normal breathing or signs of circulation. If corrective measures are not rapidly taken, this progresses to sudden death. Cardiac arrest should be used to signify an event that is reversed, usually by cardiopulmonary resuscitation and/or defibrillation, cardioversion, or cardiac pacing. The mechanism for a tachyarrhythmic arrest may be due to VT or VF, or VT degenerating into VF.

Syncope is defined as a sudden loss of consciousness with the inability to maintain postural tone, not related to anesthesia or a seizure disorder, with spontaneous recovery reported by the patient or an observer. This excludes cardiac arrest, which requires resuscitation.

For the purpose of this AUC document:

• •“Acute MI” is defined as ≤48 h after the onset of symptoms;
• •“Recent post-infarction” is defined as ≤40 days after the onset of acute MI symptoms ((30),31).

As part of ICD follow-up care, decisions must be made regarding the need for generator replacement at the time of battery depletion. In the absence of contraindications or the development of new comorbidities that may significantly limit life expectancy, generator replacement is now typically recommended for patients who had initial devices implanted for primary prevention indications when elective replacement is reached. However, it is recognized that there are few long-term data to support this standard of care. Nonetheless, in addition to assessing for pacemaker dependency, the presence or absence of ICD therapy for ventricular arrhythmias might be taken into account when considering the need for replacement, particularly if new comorbidities have developed that may otherwise have an impact on life expectancy.

Clinically relevant ventricular arrhythmias in an ICD recipient refer to:

• aVT leading to antitachycardia pacing, or VT/VF leading to shock therapy, or
• bVT duration ≥30 s in a monitor-only zone (or <30 s associated with hemodynamically significant symptoms), or
• cVT lasting ≥30 s at a rate near the tachycardia-detection threshold but not receiving therapy due to only intermittent detection.

In the case of antitachycardia pacing therapy for VT, it is recognized that many of these episodes might terminate spontaneously if detection is delayed. “Nonsustained VT” is VT of <30 s that terminates spontaneously before delivery of device therapy (including either antitachycardia pacing or shock therapy). It is recognized that implanting physicians will have a variety of different programming preferences, and some of these may include a monitor zone or prolonged detection duration in an attempt to minimize appropriate or inappropriate therapy for arrhythmias that may terminate spontaneously (32).

Ventricular fibrillation is a cardiac arrhythmia arising from the ventricles that occurs when the heart's electrical activity becomes disordered and rapid. VF is not synonymous with device-defined VF, as the device defines VT and VF solely based on the programmed heart rate and does not take into account the morphology of the arrhythmia.

VT is a cardiac tachyarrhythmia of 3 or more consecutive complexes in duration emanating from 1 of the ventricles with a rate of ≥100 beats/min. It can be “sustained” or “nonsustained.”

Ventricular Tachycardia, Sustained:

Sustained VT is defined as VT lasting ≥30 s or terminated by cardioversion or pacing before that time.

Ventricular Tachycardia, Hemodynamically Significant:

Hemodynamically significant VT is defined as VT that results in hypotension or hemodynamically significant symptoms such as angina, dyspnea, lightheadedness, pre-syncope, or syncope.

Ventricular Tachycardia, Nonsustained:

Nonsustained VT is defined as 3 or more consecutive premature ventricular complexes but lasting <30 s and terminating spontaneously, without associated hemodynamically significant symptoms, and rate ≥100 beats/min.

CAD = coronary artery disease CIED = cardiovascular implantable electronic device CRT = cardiac resynchronization therapy ECG = electrocardiogram GDMT = guideline-directed medical therapy HF = heart failure ICD = implantable cardioverter-defibrillator LBBB = left bundle branch block LV = left ventricular LVEF = left ventricular ejection fraction MI = myocardial infarction NYHA = New York Heart Association VAD = ventricular assist device VF = ventricular fibrillation VT = ventricular tachycardia

The final ratings for ICDs and CRT therapy are listed by indication in Tables (Table 1).(Table 2) to (Table 3).(Tables 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34) (indications listed by ratings are provided as an online appendix). The final score reflects the median score of the 17 technical panel members and has been labeled according to the categories of Appropriate (median 7 to 9), May Be Appropriate (median 4 to 6), and Rarely Appropriate (median 1 to 3). Of the ratings, 45% were rated as Appropriate, 33% were rated May Be Appropriate, and 22% were rated Rarely Appropriate (see Online Ratings Spreadsheet for more details).