Don’t Be Afraid – Using an ICD Means Having Difficult Conversations

Two facts: All humans die, and the main job of an implantable cardioverter defibrillator (ICD) is to prevent death from ventricular arrhythmia. The process of care surrounding an ICD, therefore, offers clinicians ample opportunity to help patients make decisions about the inevitable—death. The ICD differs from many therapies for heart failure because it confers potential risk of serious harms (shocks and anxiety), especially shocks that happen at the end of life (EOL). This fact enhances the need for thoughtful goal setting across the timeline of care. Yet, the evidence suggests patients rarely understand their options with an ICD.[1]

Physicians are reticent to initiate discussions on EOL, even in cases where simple and non-invasive procedures may mitigate risks to patient well-being (e.g. device deactivation).[2] The presence of heart failure amplifies the importance of these conversations as ventricular arrhythmia, shocks, and continued functional and clinical decline are a possible manifestation of the HF disease process.

To study the real-world conditions under which EOL discussions happen within ICD care, Thompson and colleagues used a survey sampled from patients in the Swedish national ICD and pacemaker registry [3]. They assessed experience in treatment, psychosocial well-being, ICD-specific knowledge, preferences for decision making and patient education. They found that patients with and without HF did not differ on EOL knowledge, but HF patients reported a number of QOL and experiential decrements associated with care. Patients who reported clinically significant anxiety or experienced shocks, were highly concerned about their ICD and were more likely to report having had a conversation about EOL care with their clinicians. Perhaps most striking was their finding that almost half (40%) of patients with and without heart failure did not want to ever discuss their illness trajectory or device deactivation. This was partially tempered by the fact that those with HF were more likely to want their doctor to choose when to deactivate the ICD near the end of their lives – an interesting finding which may be related to the closeness and honesty which often characterizes the long-term relationships between patients and their HF specialists.

Taken together, these findings expose a large gap between what patients ought to know about EOL with an ICD and what they actually do know. Most concerning is that the markers of being near the end of one’s life were not associated with understanding EOL options or having had a conversation about it. Rather, it was the group of patients who had experienced shocks, anxiety, and/or depression surrounding the ICD who were likely to have talked to clinicians about these concerns. The fact that ICDs continue to be a black box for patients, and that even those patients closest to end of life are not receiving counsel, represent an unmet need in heart failure care.

These findings also shed light on the difference between what shared decision-making (SDM) is perceived to be and what it actually is – at least in its ideal. Harmon et al writes that “for some, not wanting to discuss EOL issues with the provider may reflect a well thought out decision, while for others it may reflect poor knowledge of the ICD”.

The concern that clinicians may trample patient preferences by attempting to elicit patient preferences assumes that what’s important to clinicians is not part of the SDM equation. On the contrary, SDM should be thought of as an ongoing conversation in which all parties share their knowledge and expertise; clinicians about diseases, devices, and the experience they expect a patient to have, and patients about themselves, their lives, and what they want to get out of treatment. Within this meeting-of-experts definition of SDM, clinicians are not only are allowed to initiate EOL conversations with patients managing multitude illnesses, it is our responsibility to do so.[4] Here we would cite both the ethical principles of beneficence and nonmalfeasance.

Good intentions might tempt clinicians to engage in the soft paternalism of protection; avoiding the EOL conversation to spare patients momentary discomfort. But this choice undermines the far greater goals of preserving patients’ ability to make choices for themselves and building the bedrock relationship in which difficult – but eventually necessary - conversations can occur.

Who else are patients - especially those with heart failure, those who have had shocks or other bad experiences, or who are otherwise psychologically averse to thinking about death - going to talk to about this? Patients rely on the doctors and nurses for this information. Even within the much larger ecology of information available to patients online, informally from friends and family, or through other venues, our interactions during the clinic visit still hold the highest value in terms of patient education.[5]

While this is likely the rationale behind the new requirement for documented SDM occurring prior to treatment with a primary prevention ICD, there are no requirements for the topics that should be covered.[6] Indeed the greatest contribution of the findings from Thompson et al may be that it adds new context to the SDM documentation debate. Their work teaches clinicians that many patients with a life-limiting disease are reticent to discuss options at end of life. If patients always died peacefully in their sleep, and our heart failure therapies conferred only benefit and no potential for harm, we could easily comply with these wishes. But this is not the case, and thus, clinicians are bound to follow the Oath of Maimonides, which calls for us “to watch over the life and death of thy creatures.”

Of course, these conversations are hard. They require tact, skill and empathy. That may seem daunting, but in fact, most of what it takes to be successful in these talks is simple—caring.[7]