But,I Signed the Form

Recently, a visitor to Brown University shared with me his family's experiences as they journeyed with his wife through years of serious illness, and eventual dementia and death. His conveyed sorrow was laced with pride regarding the quality of life his wife enjoyed. It seemed he felt obliged to tell me that his wife received nutrition with a feeding tube near the end of life; and he continued by explaining his rationale and the conversations that had occurred. You see, he had read the nursing home studies that viewed “tube feeding” as a negative outcome. “It's a choice,” I told him. Further, I explained the concern is that people are not having the conversations (as he did) and thus decisions may not be well informed or consonant with patients' preferences. In May's Journal of Palliative Medicine Towsley and colleagues ¹ elaborated on these “missed conversations,” and somewhat surprising to me, on how having an advance directive contributed to the perception by some that other conversations might not be necessary. These qualitative findings are supported in this month's Journal of Palliative Medicine in the manuscript by Wallace and colleagues. ² Their research found advance directives were not associated with a lower likelihood of an intensive care unit (ICU) death or receipt of ICU services prior to a hospital death; but they did find having a palliative care consult at hospital admission was significantly associated with these outcomes. The conversations about preferences and goals of care that accompany palliative care consults it seems avoided undesired end-of-life ICU care that was experienced by many other cancer patients, even those with advance directives.

While the research by Wallace and colleagues ² shows numerous factors to be associated with hospitalized cancer patients dying with or without ICU services, the effect of having an admission palliative care consult was sizeable. Compared to dying in the hospital without ICU services, hospitalized cancer patients had a 95% lower likelihood of dying in the ICU and a 92% lower likelihood of receiving ICU services. The authors speculate that changing preferences over time—from when an advance directive was initiated to when hospitalization occurred—may be responsible for the advance directive findings. However, perhaps as Towsley and colleagues found, ¹ patients and their families may have felt further conversations weren't needed and thus didn't pursue further dialogue; or perhaps the advance directive remained unread in the patient's record until after the ICU admission when, as the authors note, reappraisal is common. Accordingly, the authors note that care within the ICU and days of ICU stay varied by advance directive status; and 50% of cancer patients who received ICU services but died outside of the ICU had a palliative care consult after hospital admission.

Also in this issue, Teno and colleagues lament the continuing substantial unmet needs in end-of-life care over time (2011–2013 compared to 2000); this is despite a 200% increase in hospice use and a large increase in the presence of hospital-based palliative care teams. ³ As observed in 2000, survivors of loved ones who received hospice in 2011–2013 were twice as likely to rate end-of-life care as “excellent” (compared to ratings for those without hospice); but the overall rating of “excellent” was lower in 2011–2013 compared to 2000 (47.0% versus 56.7%, respectively). Considering that 11% of survivors reported (care) decisions contrary to their loved one's wishes (20.9% if death was in an ICU) and that inadequate patient/family input into decisions was reported by one in seven (13.8%), it is clear that conversations about preferences are not consistently occurring or not occurring early enough to avoid undesired (but perhaps uncommunicated) care near the end of life.

As practitioners, researchers, and leaders, we are well aware of the problems contributing to the lack of family-patient and clinician-patient/family conversations regarding preferences and goals of care; and the 2015 Institute of Medicine report Dying in American: Improving Quality and Honoring Individual Preferences Near the End of Life ⁴ recommends education, public engagement, policy, and practices needed to improve communication and care overall for the seriously ill and dying. Public engagement in particular I believe is crucial to increase the occurrence of meaningful patient-family conversations regarding care preferences and to equip individuals with the knowledge and advocacy skills needed to ensure care preferences are heard and honored by providers. Patients and families need to be aware that a signed advance directive doesn't alleviate the need to communicate with providers about these documented preferences and to advocate for the honoring of such. Paula Span described the need for such communication in a recent New York Times article. ⁵ Two years after Ms. Span's father's advance directive wasn't asked for by emergency room staff, she found that her sister's advance directive didn't seem to matter either. However, it mattered to Ms. Span, since her sister's judiciously documented wishes enabled her to convey her sister's wishes to staff. Still, after hospital admission and until hospice admission occurred she felt the need to advocate for her sister so people wouldn't keep “doing things” to her. ⁶

My visitor may have feared being judged negatively when he told me about his wife receiving nutrition via a feeding tube, but I was impressed. While I would have chosen otherwise, his decision was reached after much inquiry and many discussions—it was thoughtful and well informed. There are no “wrong choices” when they reflect well-informed preferences. There are, however, missed conversations and perhaps an overreliance on signed forms.