Caverly TJ, Al-Khatib SM, Kutner JS, Masoudi FA, Matlock DD. Patient Preference in the Decision to Place Implantable Cardioverter-Defibrillators. Arch Intern Med. 2012;172(14):1104-1107. doi:10.1001/archinternmed.2012.2177
Author Affiliations: University of Colorado Anschutz Medical Campus, Aurora (Drs Caverly, Kutner, Masoudi, and Matlock); Denver Veterans Affairs Medical Center, Denver, Colorado (Dr Caverly); Duke Cardiovascular Center for Education and Research on Therapeutics, Duke Clinical Research Institute, and Division of Cardiology, Duke University Medical Center, Durham, North Carolina (Dr Al-Khatib); Colorado Cardiovascular Outcomes Research Institute, Denver (Drs Masoudi and Matlock); and Kaiser Permanente of Colorado Institute for Health Research, Denver (Drs Masoudi and Matlock).
Randomized controlled trials demonstrate that implantable cardioverter-defibrillators (ICDs) reduce mortality in certain patients with symptomatic heart failure and no history of sudden cardiac death (SCD).1 This trial evidence has led to guidelines recommending placement of an ICD for primary SCD prevention in patients with heart failure meeting specific criteria.2 More patients are receiving ICD therapy as indications for this therapy have expanded.3 However, ICDs are associated with important risks, including short-term procedural complications, the potential for worse quality of life if shocked, and increased hospitalization rates.4,5 Not surprisingly, patient preferences around ICD therapy vary.5 Patients who are older, have multiple comorbidities, or live with a higher burden of daily symptoms may see a trade-off between sudden cardiac death and living longer.6 In some cases, a patient's decision to forgo ICD therapy may create a conflict between the physician's desire to do good (beneficence) and the patient's preference (autonomy). We conducted a study to determine how physicians weigh patient preferences and the evidence of mortality benefit in their decision to recommend an ICD for primary prevention to potentially eligible patients.
We sent a web-based, cross-sectional, anonymous survey to a random national sample of 9969 physician members of the American College of Cardiology (ACC). Complete details of the survey and procedures have been discussed previously.7 We asked: “To what extent does each of the following factors affect your recommendations regarding an ICD for primary prevention of sudden cardiac death?” To minimize social desirability bias in responses, we included both mortality benefits and patient preferences with 12 other factors that might influence the decision. Responses were scored on a 5-point Likert scale from “not at all” (scored 0) to “a great deal” (scored 4). We also assessed physician attitudes toward ICDs by asking whether they agreed or disagreed with a series of statements using a 5-point Likert scale (“strongly disagree” to “strongly agree”). Physicians were grouped into those who rated patient preferences less important than estimated mortality benefits (“mortality benefit” group) and those who rated patient preferences at least as important as estimated mortality benefits (“patient preference” group). Pearson χ2 test was used to evaluate the differences in attitudes toward ICD therapy between the 2 groups.
A total of 1210 physicians responded to the survey (12% response rate); 1124 responses were analyzed after exclusions (pediatricians, physicians with no recent ICD experience, and non-US physicians). Demographics of responders were similar to those of nonresponders, but responders were more likely to practice in an academic setting (32.0% vs 21.4%; P < .001) and have board certification in electrophysiology (12.3% vs 4.3%; P < .001).
Mortality benefit data mattered “a great deal” in recommendations for an ICD for 962 of the respondents (85.9%); none reported that mortality benefit data mattered “very little” or “not at all.” Fewer than half of respondents (423 [37.7%]) thought that patient preferences mattered “a great deal,” and 138 (12.3%) responded that patient preferences mattered “very little” or “not at all.” Mortality benefit data were given a higher importance rating than patient preferences by 628 of respondents (56%). The group rating mortality benefit data higher than patient preferences generally had more positive attitudes toward ICDs (Table).
Despite the increasing focus on shared decision making and patient-centered care,8 over half of the respondents in this survey rated patient preferences lower than expected mortality benefits, and a considerable minority indicated that patient preference mattered “very little” or “not at all” in decision making around primary prevention ICD therapy. Physicians who rated patient preference as less important also had more favorable attitudes toward ICDs. These results illustrate 2 important points. First, they support previous qualitative findings that clear guidelines backed by strong evidence may compel physicians to adopt a beneficent or even paternalistic approach to decision making and “do the right thing,” which can take precedence over patient preferences.4 Second, more favorable attitudes regarding ICD therapy are associated with assigning less importance to patient preferences. Although this cross-sectional survey cannot determine cause and effect, it may be that the more convinced a physician is of the benefits of an intervention, the less important patient preferences become in decision making.
The response rate of 12% raises concerns about nonresponse bias. However, ACC members are a relatively homogeneous group, and to our knowledge, this sample of 1210 responders is the largest physician survey ever conducted regarding attitudes and recommendations toward ICDs.
Social norms promoted by guidelines may be a double-edged sword. On the one hand, they promote the consistent application of evidence-based therapies. On the other hand, they may also promote unilateral decision making that diminishes the importance of patient preferences, as suggested in this study. To minimize this effect, guidelines should consistently emphasize the importance of patient preferences in all clinical decisions.
Correspondence: Dr Matlock, Division of General Internal Medicine, Department of Medicine, School of Medicine, University of Colorado, Academic Office 1, 12631 E 17th Ave, Campus Box B-180, Aurora, CO 80045 (Daniel.Matlock@ucdenver.edu).
Published Online: June 11, 2012. doi:10.1001/archinternmed.2012.2177
Author Contributions:Study concept and design: Al-Khatib, Kutner, Masoudi, and Matlock. Acquisition of data: Matlock. Analysis and interpretation of data: Caverly, Al-Khatib, Kutner, and Matlock. Drafting of the manuscript: Caverly and Matlock. Critical revision of the manuscript for important intellectual content: Caverly, Al-Khatib, Kutner, Masoudi, and Matlock. Statistical analysis: Caverly. Obtained funding: Matlock. Study supervision: Kutner. Electrophysiology expertise: Al-Khatib.
Financial Disclosure: Dr Kutner serves as a Medical Editor for the Informed Medical Decisions Foundation. Dr Masoudi has contracts with the American College of Cardiology Foundation and the Oklahoma Foundation for Medical Quality.
Funding/Support: This project was supported in part by cooperative agreement number U18 HS016964 from the Agency for Healthcare Research and Quality (AHRQ), the NIH/NCRR Colorado CTSI grant No. UL1 RR025780, and the Division of General Internal Medicine, Department of Medicine at the University of Colorado. In addition, Dr Caverly is supported by an institutional T32HP1006 grant, and Dr Matlock was supported as a Hartford Geriatrics Health Outcomes Research Scholar and by a K23 (AG040696-01). Dr Masoudi has received grants from AHRQ, the National Heart, Lung, and Blood Institute, the ACC Foundation, and the Heart Rhythm Society. Dr Matlock has received grant support from the Informed Medical Decisions Foundation.
Funding for Less Is More: Staff support for topics research funded by grants from the California Health Care Foundation and the Parsemus Foundation.
Previous Presentation: The results of this analysis were presented in poster form at the American Heart Association Quality of Care and Outcomes Research Conference; May 20, 2010; Washington, DC.
Disclaimer: Contents are the authors' sole responsibility and do not necessarily represent the official view of the National Institutes of Health.
Additional Contributions: Caroline B. Emsermann, MS, and L. Miriam Dickinson, PhD, assisted with data analysis (for which they were compensated). Gillian D. Sanders, PhD, provided help with survey design and conduct.
This article was corrected for errors on September 7, 2012.