book cover

Palliative Care Perspectives : Chapter 11: The Final 48 Hours : Predicting Active Dying

We have learned the most about active dying from following cancer patients, because their dying trajectories tend to be the most predictable. However, patients with other disease processes can certainly enter a pathway largely indistinguishable from that of cancer death. Morita identified four signs that heralded impending death in 100 cancer patients: the "death rattle," respiration with mandibular movement, cyanosis of the extremities, and lack of radial pulse. He measured the median time to death from the onset of these signs. They tended to occur in a rather orderly fashion, with the death rattle preceding respiration with mandibular movement (74% of the time), which in turn preceded cyanosis and pulselessness (63% of the time). The median time until death following the death rattle was 57 (+ or - 23, SD) hours, 7.6 (2.5) hours following respiration with mandibular movement, 5.1 (1.0) hours following cyanosis, and 2.6 (1.0) hour following lack of radial pulse.3 (I was struck by this study because until I read about mandibular movement, I had been unable to see it in my dying patients.)

As this study suggests, many patients who are actively dying have "noisy respirations." These sounds come from retained secretions in the pharynx and the upper respiratory tree. Sometimes called terminal pneumonia, it is unclear how often such secretions represent true bacterial pneumonias compared to retained normal secretions in patients who are no longer able to cough.

This terminal syndrome, as we might call it, is easiest to identify in solid tumor cancer patients and is usually preceded by a bedridden status and little, if any, oral intake of food or fluids. The patient rarely speaks, or speaks only in brief phrases. A weak cough may be present initially but then disappears. The respiratory rate is variably increased and often becomes irregular, sometimes with frank Cheyne-Stokes respirations. The stethoscope is of minimal use. The buildup of respiratory secretions in the bronchi and bronchioles makes localization of underlying alveolar involvement, manifested by rales, difficult, if not impossible, to detect. More useful is direct palpation of the chest wall for vibrations that represent the buildup of secretions, a form of fremitus. If such vibrations occur only in the center of the anterior chest over the trachea, this may reflect only tracheal secretions and generate a false positive finding for active dying. These may subsequently clear if cough returns. However, peripheral fremitus appears to be more suggestive of terminal retained secretions. Fever is often absent, particularly if steroids have been used. The pulse may be strong initially but becomes threadier and eventually will not be palpable as blood pressure falls. Despite the lack of fever, peripheral vasodilatation may occur if there is underlying sepsis. In such cases the pulse initially is rapid and often hyperdynamic, which can be erroneously read as a "strong pulse." Most likely, this pulse results from a widened pulse pressure, because enhanced cardiac output under adrenergic stimulation is accompanied by a fall in systemic vascular resistance, especially if sepsis is present. I have found that feeling the shins for warmth is useful in evaluating for this. Because the shins have poor vascularization and normally are cool (especially in the presence of hypotension with reflexive vasoconstriction), warm shins conversely suggest vasodilatation. (Note that the sensitivity and specificity of these suggested examinations have not yet been tested.)

Although further studies need to be done on the short-term prediction of dying, I have found these signs useful but not infallible. Some patients with obstructive lung disease, those prone to chronic aspiration such as stroke, and dementia patients may rattle with retained respiratory secretions and yet not be actively dying. Cyanosis and mottling of the upper extremities appear more specific for impending death than do such changes in the lower extremities, where they commonly reflect peripheral vascular disease. I have witnessed false positives for mandibular movement in patients who have obstructive lung disease. The exaggerated use of strap muscles in breathing may result in jaw movements that mimic true respiration with mandibular movement. I suspect this sign in dying people results from relaxed muscular tone in the jaw combined with deep breathing. This might explain additional false positives I have seen for this sign in patients with benzodiazepine overdoses and amyotrophic lateral sclerosis (two patients each in my experience), both situations that involve relaxed muscular tone.*

*To my knowledge, mandibular movement as a little-known sign of active dying has been studied only in cancer patients. While many people sleep with their mouths open, commonly called the "O" sign, sleeping patients do not usually exhibit this gentle opening and closing of the mouth. Having watched for this sign for several years, my impression is that it is a relatively specific sign, if present, for active dying, but that it lacks in sensitivity. The few false positives I have observed were in patients with profound neuromuscular weakness and, occasionally, chronic obstructive pulmonary disease (COPD) as described in the text. The cause of this finding is unknown, but based on these observations, I suspect it results from incomplete relaxation of the facial muscles. If fully relaxed, as in REM sleep, the mouth does not move. This suggests to me that while there are similarities between dreaming and active dying, important differences, as yet unstudied, also exist.

A common mistake I have made and witnessed in others who work in hospice is to inappropriately extrapolate from cancer to other illnesses in predicting active dying. Most cancer patients do seem to follow a common dying trajectory, and hospice workers become quite good in recognizing when cancer patients enter the active dying phase. Mistakes occur when clinicians excessively extrapolate from this experience to other illnesses. In Chapter 2 I discussed "sine-waving," a vacillating dying trajectory in which patients with certain illnesses such as congestive heart failure and dementia may deteriorate and then improve - over and over again. For sine-waving trajectories, it is more difficult to state definitively that any given clinical deterioration will, in fact, lead to death.

Clinicians may also fail to predict the deaths of patients for a different reason - the tendency for patients with certain illnesses to die relatively suddenly. Obviously, patients at risk for catastrophic events such as cardiac arrhythmia or exsanguination may die suddenly and unpredictably, although relatively few such patients are followed in hospices. The more common mistake in hospice relates to patients at risk for sudden respiratory failure. Such patients are characterized by a lack of any respiratory reserve. Patients with severe intrinsic lung disease such as obstructive lung disease and pulmonary fibrosis are at risk for this trajectory. Less obviously, those patients who lack respiratory reserve because of neuromuscular disorders such as amyotrophic lateral sclerosis or Guillain-Barré syndrome often die suddenly and unpredictably. My guess is that in such patients who lack respiratory reserve, any pulmonary "insult" such as a mucous plug, food aspiration, bleeding, or early pneumonia is enough to tip the balance toward CO2 retention and sudden death.

I stress this point because I have generally found it helpful to explain to families (and some patients) the possibility or probability that we may not be able to predict impending death for patients with such illnesses. If, in fact, such a patient is found dead, family shock seems dampened by foreknowledge of this possibility. Although some may be disturbed to learn of the inherent uncertainty in predicting such deaths, many are relieved to learn that active dying will not be drawn out and that most such deaths seem peaceful. Patients who suffer from severe baseline dyspnea, as do most such patients, tend to be more terrified of worsening dyspnea and suffocation than they are of dying. They are often quite relieved to be informed that the active dying phase is likely to be very short - measured in minutes to hours, rather than days.

<<< Previous Next >>> [ Go Up ]

book cover

Palliative Care Perspectives

James L. Hallenbeck, M.D.

Order hardback from
Order paperback from

Copyright © 2003 by Oxford University Press, Inc.

The online version of this book is used with permission of the publisher and author on web sites affiliated with the Inter-Institutional Collaborating Network on End-of-life Care (IICN), sponsored by Growth House, Inc.