Imminent Death: What Happens in the Last Days and Hours

Here are passages from a few more scholarly journal articles on the subject of what to expect when someone is dying.

The full text of this first article, “Ensuring Competency in End-of-life care: Controlling Symptoms” by Frank D. Ferris, Charles F. von Gunten, and Linda L. Emanuel  can be found here.  It includes 100 works cited, and the great thing here is that there are links to take you to the abstract or full text of the articles listed — if available freely. It looks to me that for at least half of these, such links are provided. Topics covered include:

Approaches to the medical management of pain, depression, anxiety, breathlessness (dyspnea), nausea/vomiting, constipation, fatigue/weakness and the symptoms common during the last hours of life… 

“Is This a Bad Day, or One of the Last Days? How to Recognize and Respond to Approaching Demise” by Geoffrey P. Dunn  and Robert A. Milch [Journal of the American College of Surgeons, Volume 195, Issue 6, December 2002, Pages 879-887] is freely available here. It has a lot of information, including good summary tables of changes that happen weeks to days and hours before death. A few highlights:

One of the most reliable criteria for differentiating a bad day from one of the last days is the patient’s own report. If the patient states “I am dying,” or has a sense of impending doom, experience often proves him right. . . .

To date, there are no validated, consistently accurate, generally accepted models for predicting life expectancy in either cancer or noncancer diagnoses. . . .

A study of 468 hospice referrals reflecting the national pattern of diagnoses admitted to programs showed that only 20% of prognostications of length of survival were accurate within 33% of actual survival. Sixty-three and one-half percent of physicians overestimated survival, averaging five times greater than the actual survival. More experienced physicians were more accurate, and those with a longer acquaintance with their patient tended to overestimate survival. . . .

Regardless of diagnosis, observable changes during the last days of life include (Table 4): profound weakness (asthenia) and bed-bound state, sleep much of the time, indifference to food and fluids, difficulty swallowing, disorientation to time with increasingly brief attention span, low or lower blood pressure not related to hypovolemia, urinary incontinence or retention caused by weakness, loss of ability to close eyes, oliguria [low output of urine], vivid dreams or nightmares, patient reports of seeing previously deceased individuals important to the patient, frequent references to “going home” (not their street address!), or speech content related to travel to a final destination. . . .
Very late changes (Table 5) related to changes in both brain stem and cortical function heralding demise within a day to hours include: changes in respiratory rate and pattern (Cheyne-Stokes, apneas); mottling and coolness of skin from vasomotor instability with venous pooling, particularly in the pretibial [shinbone] region; dropping blood pressure with rising, weak pulse; and mental status changes (delirium, restlessness, agitation, coma). The majority of patients are comatose at time of death, though up to 30% of patients are reported to be alert until moments before death.

“Care of the Dying Patient: The Last Hours or Days of Life” [BMJ. 2003 January 4; 326(7379): 30–34] can be read here. Drs. John Ellershaw and Chris Ward begin with this fairly amazing observation:

We searched Medline from January 2000 to March 2002 in the English language by using the terms “palliative care” and “terminal care.” The search yielded 253 references, but only a limited number of articles were directly related to the care of dying patients.

Their particular interest is the “complexities of diagnosing dying” in patients with heart failure.

Predicting when death is imminent is particularly difficult in patients with heart failure for several reasons. Worsening heart failure is not always the result of an inexorable progression of the underlying pathology. In many cases a reversible cause exists (for example, a chest infection, anaemia, an arrhythmia, or suboptimal or inappropriate heart failure drugs), the correction of which may induce a worthwhile symptomatic remission. Furthermore, the use of standard diuretics, inotropes, and vasodilators in varied combinations may produce an improvement, albeit only temporary. The variable effects on outcome of these clinical and iatrogenic scenarios may partly explain the failure of many attempts to identify sensitive biochemical or haemodynamic markers of the end of life in individual patients.

Experienced clinicians will recognise a subgroup of patients, admitted to hospital because of worsening heart failure, whose prognosis seems to be particularly poor. In our experience, currently the subject of a prospective review, these patients are often characterised by:

• Previous admissions with worsening heart failure
• No identifiable reversible precipitant
• Receiving optimum tolerated conventional drugs
• Deteriorating renal function
• Failure to respond within two or three days to appropriate changes in diuretic or vasodilator drugs.

While others steadily improve, such patients often continue to worsen, although they may survive for a week or more. Before this point is reached, the likelihood of recovery and the justification for continuing invasive treatments or monitoring should be reviewed and discussed with patients and carers. .  .  .

A constant source of frustration and anger voiced by bereaved relatives is that no one sat down and discussed the fact that their loved one was dying. If relatives are told clearly that the patient is dying they have the opportunity to ask questions, stay with the patient, say their goodbyes, contact relevant people, and prepare themselves for the death.

I couldn’t find a full text free source for the next article [abstract], but the passage I quote will give you a sense of it, and perhaps you can get into a University’s collection of e-journals. It seems to me that that should be easy at state schools, supported by your taxes, but that is a whole other topic. In “Terminal Care: The Last Weeks of Life, ” [Journal of Palliative Medicine; Oct. 2005, Vol. 8 Issue 5, pages 1042-1054], William M. Plonk and Robert M. Arnold write:

Several clinical features have been identified as indicators of death within days, but research investigating the reliability of these signs is scarce. Evidence does show that physicians consistently overestimate patient survival, and those most familiar with the patient are often the least accurate. One observational study in terminally ill patients with cancer noted that patients on average developed respirations with mandibular movement 8 hours, acrocyanosis 5 hours, and radial pulselessness 3 hours before death but there was wide individual variation, with most patients developing these symptoms less than 2.5 hours before they died. Decreased consciousness was identified in 84% at 24 hours
and 92% at 6 hours prior to death. Development of a death rattle is predictive of death within 48 hours but typically occurs in less than half of patients. With the exceptions of drowsiness, fatigue, and confusion, symptoms in patients with cancer followed at home tended to improve in the last days of life. According to expert opinion, other symptoms of near death include becoming bedbound, irregular breathing, tolerating sips of fluid only, and cool or mottled extremities.

I emphasized with bold type some phrases I think it is important to bear in mind.

Mandibular movement refers to the jaw moving, like someone is eating air. The death rattle is a gurgling sound. Acrocyanosis is the same as cyanosis of extremities and means that arms and legs may become cool to the touch and bluish. The radial artery is where a pulse is felt on the wrist, so radial pulselessness means the pulse is too weak to feel there.