Of the 1,009 HCA members who viewed the survey, 676 provided responses, of whom 546 were patients with HC. The demographic features of these respondents are listed in . The mean age was 49.7 years, and 47% of the patients were women. Most patients self-identified as white and had completed education beyond high school (n = 413, 76%).
Patient characteristics (n = 546)
ICDs were currently present in 311 patients (57%), 229 (42%) had never had a device, and 6 (1%) had previously had an ICD but did not currently have a device. Device-related events were relatively common among those who had received ICDs, including shocks (n = 94, 30%), recalls (n = 77, 25%), lead problems (n = 46, 15%), and generator problems (n = 21, 7%). Many had experienced life-threatening events or serious illnesses. A few patients (n = 36, 7%) had experienced sudden cardiac death, and 27 (5%) had received cardiopulmonary resuscitation. Intensive care unit hospitalizations had been experienced by 152 patients (28%), and 76 (14%) had previously required support with mechanical ventilation.
Only 255 of the respondents (46%) had completed an advance directive. Of the 198 patients with ICDs who completed an advance directive, only 15 (8%) of these had specifically mentioned their CIEDs. Similarly, only 278 of patients (51%) had a designated healthcare proxy. Of the 218 with CIEDs who had a healthcare proxy, only 91 (42%) had discussed their views regarding their CIEDs specifically with the designated persons. Those patients >50 years old were approximately twice as likely to have an advance directive (61% vs 30%, p <0.0001) or healthcare proxy (67% vs 33%, p <0.0001), although no more likely to have included a discussion of their CIEDs in either case. Patients with education beyond high school were also more likely to have completed an advanced directive (51% vs 33%, p = 0.0014) or to have a healthcare proxy (55% vs 39%, p = 0.007). No significant differences were seen in either advance directive or healthcare proxy designation according to gender or the presence of an ICD.
Widespread uncertainty was found regarding the legal status of CIED deactivation and end-of-life care generally. For PMs, 45 (8%) responded that, in general, it was not legal to withdraw therapy when a patient with decision-making capacity makes such a request; 183 (34%) responded that it was legal; and 316 (58%) were not sure. For ICDs, 33 (6%) responded that it was not, in general, legal to withdraw therapy at a patient’s request; 201 (37%) responded that it was legal; and 305 (57%) were not sure. Many respondents were uncertain (n = 235, 43%) whether the degree of PM-dependence influenced the legality of PM deactivation (it does not). Similarly, >50% were uncertain whether a terminal illness must be present for either PM (n = 294, 54%) or ICD (n = 280, 52%) deactivation to be legal (it does not). Age, gender, the presence of a device, and education did not significantly influence these results.
Fewer than half (n = 227, 42%) correctly identified that euthanasia is illegal everywhere in the United States. A total of 159 patients (29%) thought that physician-assisted suicide was illegal everywhere in the United States, and only 134 (25%) correctly identified it as being legal in at least one state (it is legal in Montana, Washington state, and Oregon). Level of education did not markedly influence these results.
A substantial proportion of respondents characterized PM deactivation in a PM-dependent patient as physician-assisted suicide (n = 136, 25%) or euthanasia (n = 22, 4%); many (n = 224, 42%) were not certain. Respondents less frequently characterized ICD deactivation as euthanasia (n = 15, 3%) or physician-assisted suicide (n = 77, 14%), although nearly half (n = 244, 45%) were uncertain. Patients viewed the moral similarity of CIED withdrawal to cessation of other therapies differently for ICDs, PMs, and PMs in dependent patients (). More than half of the patients viewed ICD deactivation as morally different from stopping cardiopulmonary resuscitation, mechanical ventilation, renal dialysis, or feeding tubes. PM deactivation in a PM-dependent patient was viewed as morally different from cardiopulmonary resuscitation (n = 160, 31%), mechanical ventilation (n = 122, 24%), and renal dialysis (n = 147, 29%). For non–PM-dependent respondents, however, >75% viewed PM deactivation as morally different from the other 3 examples of care withdrawal. For all 4 therapies, the difference in response according to CIED type was statistically significant (; p <0.0001 for each comparison).
Figure 1 The percentage of patients who viewed withdrawal of PM in a PM-dependent patient (red-hatched bars), ICD therapy (blue-solid bars), or PM in a non–PM-dependent patient (green-dotted bars) or therapy to be morally different compared to withdrawal (more ...)
The respondents expressed important concerns regarding the potential for clinical conflicts in CIED deactivation related to religion, ethical and legal uncertainty, and informed consent (). Although the patients identified physician refusal to deactivate a CIED as potentially concerning, only 13 respondents (2.5%) reported having directly observed such a circumstance.
The percentages of patients who rated their level of concern as ≥8 on a 0 to 10 scale regarding the influence of each variable on deactivation of a CIED shown.