Risk Stratification for Arrhythmic Sudden Cardiac Death
2011; Lippincott Williams & Wilkins; Volume: 123; Issue: 21 Linguagem: Inglês
10.1161/circulationaha.110.959734
ISSN1524-4539
AutoresJeffrey J. Goldberger, Alfred E. Buxton, Michael Cain, Otto Costantini, Derek V. Exner, Bradley P. Knight, Donald M. Lloyd‐Jones, Alan H. Kadish, Byron Lee, Arthur J. Moss, Robert J. Myerburg, Jeffrey E. Olgin, Rod Passman, David Rosenbaum, William Stevenson, Wojciech Zaręba, Douglas P. Zipes,
Tópico(s)Atrial Fibrillation Management and Outcomes
ResumoHomeCirculationVol. 123, No. 21Risk Stratification for Arrhythmic Sudden Cardiac Death Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessResearch ArticlePDF/EPUBRisk Stratification for Arrhythmic Sudden Cardiac DeathIdentifying the Roadblocks Jeffrey J. Goldberger, MD, Alfred E. Buxton, MD, Michael Cain, MD, Otto Costantini, MD, Derek V. Exner, MD, Bradley P. Knight, MD, Donald Lloyd-Jones, MD, Alan H. Kadish, MD, Byron Lee, MD, Arthur Moss, MD, Robert Myerburg, MD, Jeffrey Olgin, MD, Rod Passman, MD, David Rosenbaum, MD, William Stevenson, MD, Wojciech Zareba, MD and Douglas P. Zipes, MD Jeffrey J. GoldbergerJeffrey J. Goldberger From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , Alfred E. BuxtonAlfred E. Buxton From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , Michael CainMichael Cain From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , Otto CostantiniOtto Costantini From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , Derek V. ExnerDerek V. Exner From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , Bradley P. KnightBradley P. Knight From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , Donald Lloyd-JonesDonald Lloyd-Jones From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , Alan H. KadishAlan H. Kadish From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , Byron LeeByron Lee From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , Arthur MossArthur Moss From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , Robert MyerburgRobert Myerburg From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , Jeffrey OlginJeffrey Olgin From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , Rod PassmanRod Passman From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , David RosenbaumDavid Rosenbaum From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , William StevensonWilliam Stevenson From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). , Wojciech ZarebaWojciech Zareba From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). and Douglas P. ZipesDouglas P. Zipes From the Path to Improved Risk Stratification/Northwestern University, Chicago, IL (J.J.G.); Brown University, Providence, RI (A.E.B.); State University of New York at Buffalo (M.C.); MetroHealth Campus, Case Western Reserve University, Cleveland, OH (O.C., D.R.); Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada (D.V.E.); Northwestern University, Chicago, IL (B.P.K., D.L.-J., A.H.K., R.P.); University of California at San Francisco (B.L., J.O.); University of Rochester Medical Center, Rochester, NY (A.M., W.Z.); University of Miami School of Medicine, Miami, FL (R.M.); Brigham and Women's Hospital, Boston, MA (W.S.); and Indiana University School of Medicine, Indianapolis (D.P.Z.). Originally published31 May 2011https://doi.org/10.1161/CIRCULATIONAHA.110.959734Circulation. 2011;123:2423–2430Athough it is difficult to determine the precise number, the range for the number of sudden cardiac deaths (SCDs) per year in the United States alone has been reported from 184 000 to 462 000,1 with estimates that 50% to 70% are due to tachyarrhythmic mechanisms. Regardless of where within this range the true number lies, this represents a large epidemiological problem that warrants serious attention and attempts to identify solutions. There are many obstacles to achieving this laudable goal. First and foremost, although the vast majority of SCD victims have underlying structural heart disease (in particular, coronary artery disease), a significant percentage of SCD victims have previously unrecognized cardiac disease2; on autopsy, advanced coronary artery disease with or without evidence of unstable plaques and acute or healed myocardial infarctions (often clinically silent) are commonly detected.2,3 The American Heart Association estimates that 195 000 first silent myocardial infarctions occur per year.4 Strategies to reduce SCD among individuals without known cardiac disease must therefore focus on better screening and identification of risk factors for coronary disease, with either known risk factors or heretofore unknown or unidentified risk factors. In patients with known cardiac disease, there may be diverse pathogeneses for sudden death, including primary ventricular tachyarrhythmias and acute myocardial ischemia/infarction, among others. Although therapies exist for treatment of life-threatening ventricular tachyarrhythmias and prevention of myocardial infarction/coronary artery plaque rupture, significant challenges exist in identifying the individual patient within population subgroups who is at substantial personal risk of these events, and in whom the most intensive therapies could and should be applied. Although the incidence of out-of-hospital cardiac arrest due to ventricular tachycardia/fibrillation appears to be declining over time,4 this pathogenesis for SCD still occurs commonly. This article will therefore focus on the challenges and roadblocks that are present in the identification of patients at high risk for SCD due to ventricular tachyarrhythmias (referred to simply as SCD in the remainder of the article).Much effort has been focused on the problem of risk stratification for SCD over a period of decades. In contrast to the era when risk stratification was first attempted, at the present time, the availability of therapies that have been shown to reduce SCD in various at-risk groups, including medications such as β-blockers, angiotensin-converting enzyme inhibitors, statins, and aldosterone blockers and devices such as the implantable cardioverter-defibrillator (ICD), makes risk stratification for SCD a very relevant exercise. Because of the isolated effects of the ICD on sudden arrhythmic death resulting from ventricular tachyarrhythmias, clinical trials related to the ICD have provided important information on the utility of various risk stratification approaches for the prevention of SCD. The concepts formed in the development of this information and its incorporation into risk stratification guidelines should be transferable to other therapies, both current and future.The high prevalence of coronary artery disease, with its attendant risks for SCD, makes it a very important disease process to study, and potentially a great source of information regarding risk stratification. Thus, it will serve as the focus of this report. However, many of the fundamental benefits and limitations of risk stratification may be relevant to other cardiac diseases. For example, it seems likely that the pathophysiology of arrhythmias in congestive heart failure may share common elements regardless of whether the heart failure resulted from coronary heart disease or nonischemic cardiomyopathy.Multiple invasive and noninvasive tests have been evaluated, but currently no optimal strategy for risk stratification exists. An ideal risk stratification strategy would identify those patients who will experience SCD due to a reversible ventricular tachyarrhythmia within some specified time period (ie, 2 to 5 years) and exclude those who will not experience SCD. The current widely used strategy of stratifying risk on the basis of ejection fraction falls far short of these goals.1,5 The majority of patients who will experience SCD do not have a low ejection fraction,6,7 and many patients with a low ejection fraction may nevertheless be at low risk for SCD.8,9 Given the importance to public health, the increasing importance of appropriate deployment of precious resources such as ICDs in a contracting economic environment, and the continuing challenges of risk stratification despite the plethora of research time, effort, and resources expended on risk stratification, it is appropriate to chart a course that will most efficiently identify better strategies for risk stratification for SCD. In October 2005, the first of several annual meetings of experts in the areas of cardiovascular disease, cardiac electrophysiology, health policy, and outcomes research, as well as representatives from government and industry (Boston Scientific, Inc, Medtronic, Inc, and St Jude Medical, Inc, who provided unrestricted educational grants that were used to support the meetings), was held to identify approaches that will address the current limitations of risk stratification for SCD (a full list of attendees appears in Acknowledgments). The first step in charting a course for the future is to assess the roadblocks or limitations of the past approaches. Many issues related to risk stratification form significant challenges that must be considered. These include heuristic, statistical, and financial issues.Dichotomization of Risk and Risk StratificationRisk is a challenging concept for both physicians and patients. On a practical level, physicians incorporate risk stratification into their practice by implementing therapies based on the categorization of high or low risk. Thus, an ICD is currently recommended for primary prevention in patients considered to be at high risk for SCD.10 Because indications for therapy, as in an ICD prescription, are dichotomous, the underlying risk is often perceived to be dichotomous as well. However, in complex disease states, such as coronary atherosclerosis, the biological and statistical basis for risk estimation exceeds the limits of a dichotomous function, especially of a single risk stratification variable.11 Even in circumstances in which risk may logically be considered dichotomous, as in hereditary disorders that appear to be monogenic, the notion that those carrying the genetic variant are uniformly at risk is often an oversimplification. Variable expression, environmental interactions, and modifier genes impede dichotomizing risk in that carriers of the same genetic variant do not necessarily experience the same adverse outcomes, as observed in related carriers with long-QT syndrome or Brugada syndrome. Similarly, SCD (as well as coronary artery disease) is multifactorial, and is associated with a continuous risk function, as demonstrated in an analysis of the Multicenter Unsustained Tachycardia Trial (MUSTT) population.9 Although practice guidelines and reimbursement decisions tend to dichotomize patients into low- and high-risk groups to implement therapies such as the ICD, this is an artificial separation. The risk profile of patients with a left ventricular ejection fraction of 34% (below the threshold of 35% for ICD implantation) is not likely to be substantially different from the risk profile of patients whose ejection fraction is somewhat higher than this threshold. Thus, if a continuous risk function can be generated for a population, then the appropriate level of risk that justifies an intervention can be set on the basis of risk/benefit and cost/benefit considerations.Competing RisksAnother issue, which is less well understood and/or appreciated, is the presence of competing risks for nonsudden death that can modify the relationship between arrhythmia risk and mortality. This concept can best be understood in the context of risk estimation by noting that many risk factors for SCD are also significantly associated with death due to other cardiovascular causes (and even noncardiovascular causes).12 This was nicely demonstrated in an analysis from the MUSTT,9 in which a scoring system was generated for total mortality and arrhythmic death. Most of the factors, such as ejection fraction, history of heart failure, intraventricular conduction defect, and inducible ventricular tachycardia, were significant contributors to both risk scores. Other post hoc analyses of data from randomized clinical trials using clinical factors/risk scores have demonstrated differential ICD benefits on the basis of risk profile8,13,14; on the basis of clinical criteria, high-risk groups are identifiable in whom there may not be a survival benefit of the ICD, presumably because of competing risks of death from other causes. Current risk stratification strategies (which typically rely on standard Cox models) do not account for competing risks, which limits some of their discrimination and calibration utilities. Furthermore, the current focus on risk prediction for the short term (ie, the next 5 years) limits the identification of individuals who are at risk in the longer term or who have a life expectancy longer than 10 years. To appropriately select the optimal therapeutic strategy for a patient, physicians require tools that allow them to go beyond prognosticating all-cause mortality and identify various sources of life-threatening risk confronting a patient, with the associated absolute risk level for each source. With this information, physicians can determine whether the patient is more likely to succumb to SCD than other causes.Dynamic Risk ProfilingAlthough many clinical trials, by design, execution, and analysis, have treated risk characteristics as static variables, many risk functions are likely dynamic.15 The quantitative and qualitative durability of a risk marker, such as ejection fraction, measured at different times after a myocardial infarction remains only partially defined, as does the implication of repeated measures over time. For example, risk of sudden death after a myocardial infarction declines with time from the infarction. In the Valsartan in Acute Myocardial Infarction (VALIANT) study, the monthly risk of SCD declined from 1.4% the first month to 0.14% after 2 years.16 The timing of risk assessment could therefore be an important variable. Indeed, the Risk Estimation Following Infarction—Noninvasive Evaluation (REFINE) study17 demonstrated that risk stratification testing 2 to 4 weeks after a myocardial infarction did not predict risk of serious events after a myocardial infarction, whereas testing 10 to 14 weeks after myocardial infarction did. A similar result was found in the Cardiac Arrhythmias and Risk Stratification After Myocardial Infarction (CARISMA) study.18 Moreover, a lack of favorable remodeling in autonomic tone in the initial 3 months after myocardial infarction was demonstrated to be a strong, consistent risk factor for sudden death in a combined analysis of the CARISMA and REFINE studies.19In addition, temporal variations in risk occur as functions of time of day,20–22 day of the week,20–22 and season of the year.20,21,23,24 Risk of sudden death is also known to be dramatically increased during exertion,25,26 but it is unknown whether it is better to assess risk under these conditions versus at rest. Finally, the frequency with which risk should be assessed is unknown because the duration of the predictive value of a test is rarely studied. These are all important variables that must be considered for incorporation into the framework of risk assessment.Risk SubjectivityPrevious studies have indicated that subjective estimation of cardiovascular disease risk by physicians in the absence of scientifically based risk estimation equations is inaccurate, with both systematic underestimation and overestimation observed.27,28 As challenging as these concepts may be for caregivers to incorporate into their clinical approach to SCD prevention, it is even more of an issue for patients. Furthermore, patients have difficulty understanding and perceiving differences in both relative and absolute risk estimates.29 This is even more challenging when the topic is the risks for a frightening terminal event such as SCD.Improving Risk StratificationGiven the desirability of accurate risk stratification and the long history of research in this area, it is important to understand why the field is not further advanced. It is important to acknowledge that risk stratification is difficult. The goal of a clinical risk evaluation is to be able to predict whether an otherwise stable patient will suffer a fatal ventricular tachyarrhythmia. Because this process is multifactorial and is not yet completely understood, this represents a huge challenge.One possibility for the lack of sufficient progress in this area is that better risk stratification for SCD is unachievable. SCD, even when limited to ventricular tachyarrhythmias, can be the end result of a variety of different pathophysiological events, such as an acute ischemic event caused by plaque rupture or a sudden change in repolarization caused by electrolyte shifts or autonomic inputs; the presence or absence of substrate for scar-related ventricular tachycardia may also interact with these acute factors, resulting in either stable monomorphic ventricular tachycardia or ventricular fibrillation. Because each of these upstream events may be associated with different risk factors, it can be challenging to identify single risk factors for the end event. Because of the multifactorial nature of the problem and the unpredictability of its occurrence, it is possible that the clinician may never have enough information to better stratify risk than is currently available. Fortunately, compelling data exist that show that better risk stratification can be achieved. In a post hoc report from the Multicenter Automatic Defibrillator Implantation Trial II (MADIT-II),8 5 easily identifiable clinical factors (New York Heart Association class >II, atrial fibrillation, QRS duration >120 ms, age >70 years, blood urea nitrogen >26 mg/dL) were each identified to be associated with a modestly increased mortality risk. The absence of all of these factors, which was noted in almost one third of the population, identified a group with very low mortality risk; no ICD benefit could be detected in this group. This provides sobering support for the concept that there might be an easily identifiable subpopulation of patients who currently receive ICDs who will not benefit because their risk for SCD is too low, even though left ventricular ejection fraction is markedly reduced. Although it will be challenging to define and establish this population, this represents an opportunity to better apply risk stratification techniques into the clinical decision-making paradigm.On the other end of the spectrum, it is also important to consider the challenge of applying risk stratification techniques to the broad population currently considered to be at low risk for SCD to identify subgroups of patients at sufficiently high risk to warrant therapy. An example of such an approach is the theoretical 3-step tiered risk stratification strategy, in which Bailey et al30 stratified patients who had a myocardial infarction (no a priori ejection fraction criteria) into low-risk (2.9% 2-year risk of a major arrhythmic event) and high-risk (41.4% 2-year risk of a major arrhythmic event) subgroups; only 8.2% could not be stratified into 1 of these groups (they had an intermediate risk of 8.9%). Given the plethora of techniques and approaches that exist for risk stratification, coordinated, thoughtful efforts are necessary to develop an efficient and effective risk
Referência(s)