Author Affiliations: Division of Cardiovascular Medicine (Dr Goldberger), Department of Internal Medicine (Drs Goldberger and Fagerlin), Robert Wood Johnson Foundation Clinical Scholars Program (Dr Goldberger), VA Ann Arbor Center for Clinical Management Research (Dr Fagerlin), and Center for Bioethics and Social Sciences in Medicine (Dr Fagerlin), University of Michigan Health System, Ann Arbor.
The clear benefits of implantable cardioverter-defibrillators (ICDs) in selected populations have been consistently demonstrated through landmark investigations for more than 2 decades.1 After initially demonstrating efficacy in secondary prophylaxis of sudden cardiac death (SCD),2 ICDs are now first-line therapy for primary prophylaxis of SCD in many patients with ischemic or nonischemic cardiomyopathy.3,4
The dialogue between physician and patient about ICD implantation often does not address the complex question of whether patients prefer to live longer with progressive heart failure with an ICD—with the risk of adverse events that can accompany implantation (eg, infection, vascular complications, lead recalls, inappropriate shocks)—or risk dying from SCD without one. In this issue of the Archives, Caverly et al5 explore this issue by comparing cardiologists' attitudes toward the mortality benefit of ICDs to patients' preferences of whether to receive one for primary prophylaxis. Through a cross-sectional survey of approximately 1200 US physician members of the American College of Cardiology (ACC) who reported treating patients who have ICDs, they found that nearly 90% of the respondents indicated that the mortality benefit of the ICD played a considerable role in recommending an ICD, while only 38% felt the same way about patient preferences. In addition, 12% believed that patient preferences mattered “very little” or “not at all.”
The study reveals intriguing findings about a seemingly hidden element in the decision-making process: how physicians' attitudes shape patients' beliefs, attitudes and decision-making. If physicians fail to elicit and incorporate their patients' preferences in the decision-making process, patients may (1) have unrealistic expectations of the benefits and harms of the ICD, (2) be passive bystanders in the decision-making, and (3) experience decisional conflict stemming from feeling uninformed.6
ICD implantation does not fit neatly within the traditional bounds of shared decision-making (SDM), which is typically used for medical decisions in which there is not one obvious best course of action. As an example, elective percutaneous coronary intervention (PCI) vs optimal medical therapy for patients with stable coronary disease is an appropriate context for SDM, because while PCI carries the potential advantage of improving symptomatic angina over medical therapy,7 it also confers added risk. Whether SDM is appropriate regarding ICD implantation is debatable given the direct tie-in to mortality. ICDs are widely considered to be the sole effective therapy for both treatment and prevention of potentially lethal ventricular tachyarrhythmias. Indeed, the results in the present study clearly demonstrate that some health care providers feel that SDM may not apply to ICDs.
However, in many instances, better health outcomes (eg, adherence8) result when the patient takes an active role in the decision-making process. When ICDs are offered, physicians need to be especially aware of the substantial implications for patients' end of life, entailing a trade-off between dying from progressive heart failure or a fatal ventricular tachyarrhythmia. If a patient has a preference against survival with heart failure, this needs to be weighed against the longevity they would gain with an ICD.
The study has important limitations and does not illustrate how physician bias affects the overall decision on whether the patient ultimately receives the ICD—especially if there is decisional uncertainty. It is notable that a nontrivial number physicians in the study elected not to recommend an ICD to some eligible patients, not solely the mortality-minded ones.
In addition, the study does not address the reasons why some physicians may not appreciate the “absolute” mortality benefit of the ICD. Indeed, for many clinical situations (eg, congenital cardiomyopathies such as the long QT syndrome9) and patient populations (eg, women10 or elderly individuals11), the mortality benefit of ICDs is not well established. Furthermore, a recent analysis from the National Cardiovascular Data Registry's ICD Registry revealed that nearly 25% of ICD recipients did not meet evidence-based criteria for the device.12
Despite these notable exceptions, the dilemma facing patients eligible for ICDs remains unchanged: whether to receive an ICD or face the risk of SCD. In this context, it is easy to understand why the “mortality-minded” physicians represented in the present study may feel that the efficacy of the ICD is matched to necessity and thus does not require SDM.
Other investigations are needed to answer some of the remaining uncertainties. Are there patient-, physician-, or practice-level characteristics that can predict whether patient preferences are valued in the decision-making process? Would a different or larger population reveal a less uneven distribution of attitudes? Indeed, it is not surprising that this small subset of ACC members would be particularly motivated by the current evidence-base, emphasizing mortality over patient preference.
Overall, these results may be viewed as discouraging from both the medical community and the patient perspective, especially given that patient-centered care and SDM are part of high-quality clinical practice. The research suggests that patient preferences may be undervalued in this context. At the same time, it reinforces the fact that although decisions regarding ICD implantation are in a class of their own, the ultimate ideals of SDM should not be abandoned. Rather, it requires both patients and physicians to enter into a difficult dialogue that, by necessity, extends beyond the standard goal of SDM. Clearly, there exists an opportunity to create an equilibrium—while we expect physicians to carry a presumptive bias in favor of life-sustaining treatments, that bias has to be balanced against patient preferences and informed patient decision-making.
Correspondence: Dr Goldberger, Division of Cardiovascular Medicine, University of Michigan Health System, 6312 Medical Sciences Bldg 1, 1150 W Medical Center Dr, Ann Arbor, MI 48109-5604 (firstname.lastname@example.org).
Published Online: June 11, 2012. doi:10.1001/archinternmed.2012.2660
Financial Disclosure: None reported.
Additional Contributions: James F. Burke, MD, MS, at the University of Michigan Health System provided thoughtful comments on an earlier draft of the manuscript. No compensation was associated with his contributions to this article.
Goldberger ZD, Fagerlin A. ICDs—Increasingly Complex DecisionsComment on “Patient Preference in the Decision to Place Implantable Cardioverter-Defibrillators”. Arch Intern Med. 2012;172(14):1106-1107. doi:10.1001/archinternmed.2012.2660