Attitudes of older adults with serious competing health risks toward their implantable cardioverter-defibrillators: a pilot study

The major finding of this study was that older adults with multiple chronic conditions and ICDs remained enthusiastic about device implantation despite high short-term risk of death or functional decline. Nearly half overestimated the survival benefit of ICDs. The majority of patients did not consider their competing health risks in their decision of whether or not to accept an ICD, perhaps because they were unaware of their prognosis, or because of a societal bias to embrace potentially curative treatment, regardless of potential consequences.

Patients revealed considerable gaps in their understanding of ICDs. Importantly, 7 participants who had never been shocked thought the ICD had saved their life and 10 were unsure. Using a question from prior ICD survey research [21], we found that nearly half of those surveyed expected an ICD to save greater than 50 lives per 100 during 5 years. The multiple choice question presented absolute risks using frequencies, rather than relative risks. This method has been shown to improve patients’ understanding of risks and benefits, and is considered “state of the science” in shared decision making [25]. Poor grasp of complex medical information is not isolated to defibrillator technology, nor does it necessarily indicate that accurate information was not conveyed to patients at the time of implantation. A variety of factors may prevent older adults from achieving truly informed consent [26]. Marginal health literacy may be one contributing factor; the results of the health literacy screen suggest that patients in our sample performed well with simple medical tasks but may sometimes have had difficulty comprehending more complicated medical information [22]. The patients in our study were recalling discussions that happened years ago; we do not know what they were told. Previous research has suggested that cardiologists may downplay information about psychological and long-term risks when discussing ICDs with eligible patients [10, 27]. Our data do not allow us to reach conclusions about the reason for patients’ poor understanding of ICDs. In addition, we did not have information on patients’ perceived health status. It is possible that the participants considered themselves to be in relatively good health with reasonable life expectancy, though many had high VES-13 scores. This misunderstanding is a barrier to informed consent and patient-centered care.

Our findings are consistent with previous studies, which found that patients perceive an obligation to undergo cardiac interventions despite advanced age, overestimate the survival benefits conferred by ICDs, and are unwilling to deactivate their devices even when presented with hypothetical scenarios in which they are likely to die soon of cancer or another noncardiac cause [21, 28]. Our data expand upon previous qualitative studies of ICD decision-making, which were small, included patients as young as 18 years [20, 29], or examined preferences for ICD deactivation in the context of hypothetical health scenarios [20, 21]. One study found that the majority of patients (mean age 71.4 years), after hearing an informational script about the benefits and burdens of ICDs, would opt for ICD deactivation in hypothetical end-of-life scenarios such as advanced dementia or prolonged mechanical ventilation [20]. Yet it has long been recognized that “the willingness to suffer for a chance to postpone death may be felt more acutely by those nearer to death,” [30] so it is not clear how such patients would decide if in these clinical situations. Our study addressed a slightly different question, ICD placement rather than deactivation. A major strength of our study was that the patients were older and had a higher prevalence of functional impairment than patients in previous studies; 43 % of patients in our study had VES-13 scores ≥3, conferring a 50 % likelihood of further functional decline or death within 2 years.

Our findings are relevant given the large number of older adults undergoing ICD implantation [2]. The decision to implant an ICD in older adults is complex, particularly for primary prevention in patients who are at risk for sudden cardiac arrest because of systolic dysfunction but have never experienced a ventricular arrhythmia. A substantial number of older adults currently receiving ICDs may not actually benefit from them, either because they are not at high risk of sudden cardiac death or because they are more likely to die of other causes; only one-quarter of patients receive appropriate ICD shocks within 5 years [31]. Many older patients have multiple chronic conditions, increasing their risk of death from non-arrhythmic causes. Current ICD guidelines recommend against defibrillator implantation in patients with life expectancy less than 1 year [6]. Functional status measures such as those in the VES-13 are not routinely obtained during cardiology office visits [18]. Functional status falls outside the traditional disease model and is often under-recognized by non-geriatricians, though it predicts mortality and may impact the effectiveness of ICD use, as well as an older adult’s decision to accept an ICD. Older adults with multiple coexisting diseases and functional impairments may weigh the potential benefits and harms of ICD therapy differently if consent were truly informed about the benefits and harms of ICDs and the risk of competing health events [32, 33]. For those with multiple chronic conditions and poor functional status, the “rescue culture” of modern medicine can result in a spiral of aggressive therapies that may have little potential for benefit and may instead increase distress near the end of life [13, 34, 35]. The societal imperative to avoid death in frail, older patients may result in “new pathways to death and new qualms for patients and families” [13, 28].

The second limitation is that only 44 of 104 eligible subjects (excluding 4 who had died) completed the survey. The 33 patients who were lost to clinic follow-up and 12 who did not respond to repeated phone calls were likely different than the patients who agreed to participate. Patients who declined ICDs, no longer followed up with the electrophysiology clinic, or died from ICD complications or comorbidities likely have negative viewpoints that we did not document. Only one participant had experienced an ICD-related complication. Therefore, the generalizability of our study to the broader population of older adults with ICDs is uncertain. Despite these caveats, our findings are important because they reveal a high degree of misunderstanding about the potential benefits and harms of ICDs.

The third limitation is that we were unable to distinguish between patients who received their ICDs for primary or secondary indications. Patients’ attitudes may differ depending on the indication for ICD therapy. Fourth, our patients were recruited from a single academic cardiac device clinic. Their experiences may not be representative of patients at other institutions. Although we did not formally assess the rate of periprocedural or late (≥90 days) complications among the study population, we estimate that the rate at our institution is consistent with that reported in the literature [7]. Patients at referral centers may be more likely to accept treatment with invasive medical technologies [21]. Fifth, the investigator who made the initial contact with potential participants is well-known to them because he conducts all of their device interrogations, and this may have positively influenced their perspectives toward ICD therapy. Patients’ acceptance of ICDs is strongly influenced by cardiologists’ opinions [29]. Sixth, our data did not permit us to determine the timing of ICD shocks. Patients may feel heightened anxiety and regret about their ICD in the immediate post-shock period. Seventh, the small sample size limits generalization.

A final limitation is that some of the patients had pacemaker-defibrillators and may have conflated the two separate functions of the device. For example, the patient who said, “The only thing keeping me going is the defibrillator” may have had severe sinus node dysfunction that improved after receiving a pacemaker-defibrillator. However, our consent script and questions explicitly stated that the study was about defibrillators. Many patients have indications for pacemakers and defibrillators, and they should understand the distinction so that they can make an informed choice. An ICD itself does not improve quality of life. A patient with poor functional status and limited life expectancy due to other comorbidities may want a pacemaker but wish to forego the possibility of defibrillator shocks.

Despite these limitations, this study expands previous literature and raises important concerns about the quality of ICD decision making for older adults with competing health risks, and substantial patient misunderstanding regarding the potential benefits and harms of ICD therapy. Although the nation’s leading cardiology and heart failure organizations have called for shared decision making regarding cardiac implantable electronic devices and consideration of prognosis, physical function and quality of life, cardiologists receive little if any formal training in how to conduct such conversations with patients about whether or not to pursue treatment with invasive technologies [36‐38]. Furthermore, physician quality metrics and professional society guidelines may serve as barriers to patient-centered care by heavily influencing cardiologists’ decision making regarding ICDs [29, 39]. As a result, clinicians may be reinforcing patient misperceptions, and the care patients receive may frequently not be the care they or their families would choose if properly informed [21, 32, 40]. Decision aids are a potential solution to improve ICD decision making [41]. A recent systematic review identified 4 decision aids for patients considering ICD insertion and found that the tools “included comprehensive content on technical aspects of insertion, but made limited references to implications for quality of life and generally lacked balance in terms of how the decision to insert was presented.” [42] The development of high-quality decision aids that address not only the technical aspects of ICD insertion, but also issues related to how ICDs may affect quality of life, is needed.

Incorporation of comprehensive geriatric assessment has been recommended within the fields of geriatric oncology and nephrology in order to identify risk factors for increased vulnerability and inform treatment decisions of older patients with cancer and end-stage renal disease [43, 44]. Similarly, awareness of geriatric impairment among older adults who are eligible for ICDs could lead to improved clinical decision making.