Elective discontinuation of life-sustaining mechanical ventilation on a chronic ventilator unit

Withdrawal of medical interventions has become common in the hospital for patients with terminal disease. Despite the widespread feeling that medical interventions may be futile in certain patients, many patients, families, and medical staff find withdrawal of care difficult and withdrawal of mechanical ventilation to be the most disturbing secondary to the close proximity of withdrawal and death. Presented is a 6-year retrospective review of elective withdrawal of life-sustaining mechanical ventilation on a chronic ventilator unit (CVU) in an academic nursing home. Of the 98 patients admitted to the 19-bed CVU during this period, only 13 underwent terminal weaning (TW). Statistically, these 13 patients did not differ significantly in age, gender, race, route of nutrition, decisional capacity, or length of stay on the unit compared with the 85 patients who were not terminally weaned (t-test P > .05). Stepwise logistic regression found that patients who were more alert at admission were more likely to have participated in TW (χ² = 5.22, coefficient for alertness P < .036). The decision to terminate mechanical ventilation was made by patients in eight cases and by family in five cases. The first step in the process leading to TW was a discussion with the patient and family about plan of care, including the patient's desires for attempted resuscitation, rehospitalization, advance directives, and family contacts. Plan of care was reviewed informally in a weekly multidisciplinary round and formally, to address each patient's care plan, in a multidisciplinary family meeting on a regular basis. The second step was to address TW when brought up by the patient, family, or medical staff. A request for TW by a patient or surrogate was referred to the medical staff, who screened the patient for depression or other remediable symptoms. The third step was to refer the patient and family to another formal meeting to discuss the request for TW and, if needed, in the case of multiple family members, to allow questions to be answered and consensus to be formed. Additional meetings were scheduled as needed. The next step occurred once a consensus was reached to proceed with TW; a date and time was set to reconvene the patient, family, and anyone else who wanted to be present at the TW. The TW process began when a peripheral intravenous catheter was placed and the patient was premedicated with low doses of morphine sulfate and a benzodiazepine. After premedication, the patient was removed from the ventilator. The physician, nurse, family, and physician assistant remained at the bedside and additional morphine or benzodiazepine was given, as needed, for symptom management. Death from TW occurred in all patients, at times ranging from 2 minutes to 10.5 hours (average 6.2 hours). A mean total dose of 115 mg morphine and 14 mg diazepam was given for symptom control. There was no correlation between dose of these medications and duration of survival off the ventilator.