Prolonging the withdrawal of life support in the ICU affects family satisfaction with care

Clinicians in the intensive care unit (ICU) often care for patients who are on several life support measures at once. When such a patient is dying and the decision is reached to withdraw life support, these clinicians may make an imperfect compromise in seeking to balance the complex needs of the patient and the patient's family - they may remove the life support measures one at a time over a period of days, rather than withdrawing all at once.

According to a paper published in the Oct. 15, 2008, issue of the American Journal of Respiratory and Critical Care Medicine, this practice, referred to as sequential withdrawal, may be relatively common, and may have a varying impact on the family's satisfaction with ICU care. This study was funded by the National Institute of Nursing Research (NINR), the lead Institute for end-of-life research at the National Institutes of Health (NIH).

"We found that sequential withdrawal of life support is not as rare a phenomenon as previously thought," said J. Randall Curtis, M.D., MPH, the principal investigator of the study. "It occurred in nearly half of the patients we studied."

"The decision to remove a loved one from life supporting treatment is typically very traumatic for families," said NINR Director Patricia A. Grady, Ph.D., RN, FAAN. "Once a patient enters the ICU, clinicians may need to help family members develop realistic expectations based on the patient's prognosis. This study highlights the importance of open communication between clinicians and the family."

The data for this study came from a larger project involving ICUs in 15 hospitals across the Seattle and Tacoma area in Washington state. Dr. Curtis and colleagues examined the life support withdrawal process for 584 patients who died in the ICU or within 24 hours of discharge from the ICU, and surveyed the family members on their perceptions of the care provided.

The patients ranged in age from 19 to 99 years, with an average age of 72 years, 91 percent were white, and 61 percent were male. A medical chart review found that the patients received a median number of four life support therapies or interventions within their last five days of life: 99 percent had laboratory tests done, 83 percent were on mechanical ventilation, 76 percent received intravenous fluids, 52 percent were on vasopressor medications to sustain blood pressure, 33 percent received tube feedings, and 9 percent received renal dialysis.

When the decision was reached to remove life support, all measures were withdrawn on the same day for 54 percent of the patients. For the remaining 46 percent, the process of withdrawal took at least two days. Among these patients, dialysis was most often the first therapy withdrawn, and mechanical ventilation the last. Older patients, those with cancer, neurologic, or respiratory disease, and those experiencing pain tended to have a shorter duration of the withdrawal process, while trauma patients had the longest. The presence of a living will did not influence the duration of the withdrawal process.

Among the families, 95 percent participated in a family conference with the ICU clinicians during the last week of the patient's life. Having more family members involved in making decisions tended to prolong the withdrawal process. When surveyed one to two months after the death, family members of patients who had a short ICU stay reported a lower satisfaction with the ICU care if the withdrawal process was extended over more than one day. However, for family members of patients who had a long ICU stay (eight days or more), satisfaction with care increased with a more extended duration of the withdrawal. In addition, family satisfaction with care was higher if the patient was off the ventilator at the time of death.

Of particular interest was the finding that the families of patients who had experienced a longer ICU stay tended to prefer the more extended, sequential withdrawal process. "This finding is in the opposite direction of our original hypothesis," stated Dr. Curtis. "We believed that extending the withdrawal process would lower the satisfaction with care among all families. A longer duration of withdrawal of life support seems unlikely to benefit the patient, because it prolongs non-beneficial and sometimes painful therapies."

"After making the decision to withdraw life support measures from a dying patient in the ICU, some physicians may slow down the withdrawal process to give the family more time to cope," noted Dr. Grady. "The outcome of this study indicates that nurses and physicians need to continue to work with the family throughout the patient's ICU stay to provide them with accurate information on which to base decisions, and prepare them emotionally for the possible loss of their loved one."