Critical Care Medicine - List
http://www.pitt.edu/~crippen/

CASE#2: TERMINAL WEANING AN AWAKE PATIENT.

Mr. T was an unfortunate 28 y/o male with a past history of multiple medical problems resulting in a debilitated state. At age 14 he suffered a cervical fracture after diving into the shallow end of a swimming pool, resulting in spastic quadriplegia with only some minimal extensor use of his arms. However, he made the best of this handicap and ultimately married a woman with multiple musculoskeletal deformities who was wheelchair bound. Between them, they maintained a home and greatly relied on each other. His wife held a job in the social service department of a local hospital, and they were financially self sufficient.

The patient was originally admitted to the internal medicine service with hematuria. After an extensive work up he was found to have transitional cell carcinoma of the bladder, disseminated to Stage 3, with enlarged nodes in the para aortic region seen on CAT scan. He elected to undergo chemotherapy and endured a great deal of discomfort from nausea, vomiting, and septic episodes from neutropenia. He underwent several courses of antibiotics for sepsis. About two weeks after admission, a followup CAT scan of the abdomen showed painful obstructive uropathy of the left ureter and bilateral pleural effusions. He began to require large doses of narcotics for abdominal pain. At this point the patient began to become frustrated with his course and began some hostile acting out behavior and refused some therapies. He developed anasarca from poor nutritional status even in the face of parenteral feedings. He also had recurrent bowel ileus requiring an uncomfortable nasogastric tube. He became virtually dependent on narcotic analgesics.

About six weeks after admission, he began to complain of some progressive respiratory distress. A chest tube was inserted to drain a marked pleural effusion which ultimately showed malignant cells. At this point the patient was quite adamant in his desire not to be intubated should his condition deteriorate further, but seemed content to continue conservative and supportive therapy as long as there was some chance of palliation. However, as his respiratory insufficiency progressively increased, he became very uncomfortable and was afraid of dying in his sleep. After discussing this matter at length with his wife, he changed his mind and consented to endotracheal intubation. He was admitted to the Medical ICU.

On the same day of his MICU admission, he apparently changed his mind about the endotracheal tube and managed to extubate himself by squirming free of his hand restraint. His respiratory insufficiency progressed rapidly and he, again, became very uncomfortable. Shortly after his self extubation, he had a discussion with the attending critical care physician about re-intubation. During this discussion, he was lucid and competent. The intensivist informed the patient as accurately as possible what his options were under the circumstances. Those options were:

  1. The possibility of progressive respiratory failure and death, if not immediately reintubated.

  2. The possibility that re-intubation would be of no benefit to him other than to make him ventilator dependent, guaranteeing an ICU occupancy for the rest of his life, until he eventually succumbed as a result of his cancer or some complication thereof.

  3. The possibility of short term benefit as a result of a short period of intubation, aggressive pulmonary toilet and other aggressive therapy, after which it might be possible to extend his remaining days or weeks in a relatively comfortable manner.


After this discussion with the patient and his wife, they decided to opt for a reintubation on a short term basis predicated on the the possibility of short term benefit. It was also agreed that he would be adequately sedated during the intubation, and thereafter for his continuing comfort. It was further agreed that if, after some unspecified but finite length of time, there was no benefit forthcoming, he could opt for extubation and ensuing death under as comfortable a circumstance the intensivist could offer. The patient was adequately sedated and re-intubated approximately two months after his original hospital admission. Following re-intubation, the patient was aggressively treated in the MICU with pulmonary toilet, sclerosis of his pleural space in an attempt to reduce the malignant effusion, continuing antibiotics and general supportive care. Two days following intubation, several attempts to wean him from the ventilator failed.

It became apparent that the patient had become ventilator dependent and no lasting benefit had occurred as a result of his intubation. At this point, more options were discussed with the patient and his wife:

  1. The possibility of a trial extubation, after which he could effectively speak his mind about a third re-intubation if the trial extubation failed.

  2. An elective tracheostomy and transfer to the pulmonary rehabilitation service for his remaining days where he could be kept as comfortable as possible in an non-lCU environment.

  3. Extubation and comfort measures controlled by his physician during the dying process.

The patient chose the latter option. A consult was placed to the Ethics Committee for their evaluation.

After examining the chart, speaking to the patient and his wife, and discussing the situation, the committee found the following: the patient had an irreversible disease process, he had undergone the most aggressive care possible, this care had failed to restore him, he had made a deal with the intensivist for one last aggressive measure with a predetermined "stop order" limit if no benefit was to be had. There had not been any improvement following this aggressive therapy. The following notation was made in the chart: Accordingly the ICU physicians would like to now honor the agreement made at the time intubation was performed (extubation and inevitable death). His wishes should be honored".

Shortly thereafter, the patient's wife, as was her custom, crawled up on the patient's bed from her wheelchair and discussed this decision at great length with him. Other members of both families also became involved with these discussions. However, the patient's wishes ultimately prevailed and it was requested that the intensivist arrange transfer to a private room in the hospital where the patient could be extubated with his family at the bedside. The intensivist complied with these wishes and the patient was transferred to one of the private isolation rooms with ventilator capability. He was made comfortable in this area, still attached to the ventilator in the previous mode. One last time, the patient's wife crawled up next to his head and they discussed the issue at great length again, with the input of both families. Ultimately, the wife crawled back into her wheelchair and tearfully joined the rest of the families at the foot of the bed. When the intensivist entered the room after being paged, she stated that the patient had made up his mind and she respected his wishes. She desired that we would honor our commitment, and that we also honor our commitment to render the dying process a comfortable one for the patient. The wife reiterated her desire that the patient feel no discomfort during this process.


Comment#1: This case is a little trickier than what it might seem at first blush. Terminal weaning of the alert patient is a vastly different proposition than terminal weaning of moribund, unconscious patients. Patients who opt for death rather than the perpetual encumbrance of life support systems are alert enough to feel discomfort on discontinuation of these devices. Patients and their families not only demand their right to discontinue life support systems, but also their right to die in a "comfortable" fashion once removed. This puts us in the difficult position of having to diminish pain and discomfort during the dying process without directly facilitating death as a result. The principle of double effect buffers the situation somewhat (administering analgesia or sedation to the dying patient knowing that the dying process may be hastened as a result. However, if the patient succumbs as a proximal effect of a therapeutic or anti-therapeutic manipulation, we are skirting dangerously close to active euthanasia (in my opinion).

In fact, I think it is a true fact that judiciously administered narcotics or sedatives may ameliorate the discomfort of the dying process without actually becoming the proximate cause of death. Therefore, the administration of narcotics to palliate should not (in my opinion) hasten respiratory failure such that the patient succumbs from the effects of morphine rather than the underlying disease. I'll be the first to admit that this is a very fine line to tread. Should we worry about treading it? Withdrawal of futile extraordinary treatment results in a situation where the patient ultimately dies as a consequence of an untreatable disease. Euthanasia results in a situation where the patient dies as a proximate result of your treatment. If you are willing to accept these premises as at least arguable, how would you proceed with the nuts and bolts of ventilator withdrawal?


Comment#2: We found that when weaning the awake patient from mechanical ventilation, dyspnea was (is) the biggest problem. Dyspnea immediately precipitates a self perpetuating spiral of anxiety, increased carbon dioxide production from increased musculoskeletal hyperactivity, increased catecholamine release from hypercarbia and so on. There is no way weaning from mechanical ventilation can be totally discomfort free in the alert patient. We looked into the literature for some effective ways to blunt the sensation of dyspnea.

Since there was some evidence that opiates would give us the effect we needed, we locked onto fentanyl as our opiate of choice. Fentanyl would seem to be a very effective medication during the dyspnea stage, combining effective sedation, analgesia, anxiolysis and sympatholysis in a rapid acting and titratable manner. Treatment can be initiated by an initial bolus followed by a continuous infusion. Any resulting agitation or discomfort from the weaning process can be effectively ameliorated by increasing the fentanyl infusion as needed.

Since mechanical ventilation landed in the plus column, we elected not to physically extubate the patient as the ventilator would decrease dyspnea while in place. Endotracheal intubation would also stay in place to avoid gagging, agitation and panic as the airway collapsed as a result of progressive respiratory failure. Studies of alveolar anoxia suggest that the most rapid descent into unconsciousness with the least agitation occurs when hypoxia is allowed to progress in the face of normocarbia. The plan was to allow the patient to become hypoxic with assisted ventilations to maintain the pC02 at normal or below-normal levels. Since the patient had some degree of pulmonary insufficiency, decreasing the FiO2 would promote hypoxemia and tissue hypoxia but a minimum of agitation as a result of hypercarbia mediated catecholamine stimulation. Fentanyl was used as an adjunct in doses substantial enough to ameliorate the sensation of dyspnea, providing sedation, anxiolysis and sympatholysis, but not inducing respiratory failure. Once the patient became somnolent from tissue hypoxia, ventilation support could be discontinued, allowing hypoventilation and C02 narcosis, further blunting discomfort during the dying process.

The patient was quite comfortable on assist-control mode. It was elected to continue the assist-control ventilation mode and slowly decrease the FiO2 to room air while maintaining adequate ventilation under a continuous infusion of fentanyl. The FiO2 was progressively reduced from 50% to 21% over a period of about 10 minutes. During this period the patient progressively increased his ventilations, easily supported by the ventilator, and buffered with intermittent boluses of fentanyl administered by the physician at the bedside through the central line. "Stacking" of ventilations on the assist-control mode was not a problem, presumably because of ongoing intravenous sedation with fentanyl, which was administered empirically if the patient appeared to become uncomfortable. Other concurrent sedative drugs were considered but not used as the fentanyl alone seemed to be effective.

At several points during this process, the wife became quite concerned and requested more sedation be given because the patient seemed to be "in pain". After a period of about 15 minutes, the patient progressively became unresponsive and his hyperventilation subsided. At this time, the ventilator support was decreased to CPAP. The patient progressively developed agonal respirations and his heart rate slowed. He was pronounced dead approximately 30 minutes after the initiation of this "titrated terminal wean". A total of 500 mcg of fentanyl was administered over this 30 minute period, approximately 33 mcg of fentanyl a minute over the course of the weaning process.

In this manner we think we were able to tread the fine line between causing death as a result of intentional or unintentional narcotic dose and allowing the patient to die as a result of his natural fatal disease process, unencumbered by discomfort during the dying process. If anyone wants to quibble about 500 mcg of fentanyl over 30 minutes primarily causing ventilation failure, l have little resistance. However, at no time (at least visually) did I ever note the patient stop breathing from what I thought was a narcotic effect.