Discussing Withdrawal or Withholding of Life Support: Family Perspectives

Abstract & Commentary

Synopsis: When the next of kin of patients who had died in the ICU of a teaching hospital were surveyed 1 year later, nearly half of them reported that there had been conflict about withholding or withdrawing life support, involving such things as staff behavior, communication, and family dynamics. Family members identified the attending physician as the ICU leader despite the presence of house staff.

Source: Abbott KH, et al. Crit Care Med. 2001;29:197-201.

In this cross sectional survey, abbott and colleagues were trying to assess the quality of care as it relates to dying patients in ICUs. The issue at hand was end-of-life care discussions with next of kin. Abbott et al invited selected family members to participate in the study 1 year after the family member had had a discussion with the patient’s caregivers about withdrawing or withholding life-sustaining support and the patient had died. The attending physician taking care of the dying patient had to agree for the family member to be interviewed.

Patients were identified as a prospective cohort during a 10-week period at 6 ICUs of Duke University Medical Center. Of the 102 patients about whom there were end-of-life care discussions with family members, 98 next of kin were identified. Two of these had died at the time of contact and 22 could not be contacted. In 9 cases, the attending physician refused to allow Abbott et al to study the family member of their patient. Of the 65 family members contacted, 48 agreed to be included in the study. After an in-person or telephonic semi-structured interview, the transcript was coded by a systemic coding system. Various aspects of conflict and psychosocial support were coded. Two raters coded all interviews and inter-rater agreement was defined.

Abbott et al found that almost half of the respondents (46%) reported conflict during their family member’s stay in the ICU. The conflicts were expressed over perceived unprofessional staff behavior (31%), the care of the patient (19%), over communication (33%), between staff and family (40%), and over treatment decisions (15%). In terms of social support, 92% of the family members reported involvement of multiple family members in decision-making process. About half of the family members (48%) reported faith or spirituality as a significant and reassuring aspect of the hospital stay. Sixty-three percent of the studied family members reported having had a previous conversation with their loved one about end-of-life care preferences. A large number of family members suggested that more space should be provided for such discussions. An open visitation policy of many ICUs was considered very comforting to family members. Nearly half of the family members thought that their patient’s attending physician was the leader of the ICU; all ICUs were teaching services and had house staff as primary providers, yet rotating house staff was a negative factor for only 17% of families.


So what does such a small, potentially biased study mean to us in clinical practice? In fact, this is a wonderful study for almost any ICU team. It provides a framework on which one can start a quality control program in the ICU. It is also helpful in terms of the end-of-life care discussion format. Families sought even more information than was provided, and they identified the attending physician as a leader, seeking to talk to him or her more often than not. One could thus formulate a plan to have at least some time spent by the attending physician with the family members of moribund patients. Multiple family members, previous discussions with patients, and spirituality were important support systems for family members making the decision; these things can also be looked upon when engaging in end-of-life care discussions with family members. As long as a leader could be clearly identified by the patients’ families, rotating house staff did not have a negative effect on the families’ perceptions of care. This suggests an important role for the staff intensivist, in identifying himself or herself as a leader in the multidisciplinary teams in the ICU.

Unfortunately, quality is a difficult thing to quantify. Isn’t it an oxymoron to try to quantitate quality? If medicine is becoming a business, then quality of care is an important marker of how well you are doing your business. Needless to say, the ICU takes care of a patient population that experiences a high mortality rate. For this reason, mortality rate is difficult to use as a quality of care standard. A lot of end-of-life care discussions happen in the ICU, and withdrawing and withholding life support are as much part of ICU practice as are intubation and mechanical ventilation. Hence, even when patients are dying, how well we take care of them and how families perceive our care are extremely important to any quality assurance program.

Dr. Nanavaty is Senior Research Fellow, Department of Critical Care Medicine, National Institutes of Health, Bethesda, Md.