ABSTRACT.Objective. Approximately 60% of deaths in pediatric intensive care units follow limitation or withdrawal of life-sustaining treatment (LST). We aimed to describe the circumstances surrounding decision making and end-of-life care in this setting.Methods. We conducted a prospective, descriptive study based on a survey with the intensivist after every consecutive death during an 8-month period in a single multidisciplinary pediatric intensive care unit. Summary statistics are presented as percentage, mean ؎ standard deviation, or median and range; data are compared using the Mantel-Haenszel test and shown as survival curves.Results. Of the 99 observed deaths, 27 involved failed cardiopulmonary resuscitation; of the remaining 72, 39 followed withdrawal/limitation (W/LT) of LST, 20 were do not resuscitate (DNR), and 13 were brain deaths (BDs). Families initiated discussions about forgoing LST in 24% (17 of 72) of cases. Consensus between caregivers and staff about forgoing LST as the best approach was reached after the first meeting with 51% (35 of 68) of families; 46% (31 of 68) required >2 meetings (4 not reported). In the DNR group, the median time to death after consensus was 24 hours and for W/LT was 3 hours. LST was later withdrawn in 11 of 20 DNR cases. The family was present in 76% (45 of 59) of cases when LST was forgone. The dying patient was held by the family in 78% (35 of 45) of these occasions.Conclusions. More than 1 formal meeting was required to reach consensus with families about forgoing LST in almost half of the patients. Families often held their child at the time of death. The majority of children died quickly after the end-of-life decision was made. In pediatric ICUs (PICUs), retrospective studies done during the past decade indicate that 40% to 60% of all deaths follow such a decision. [3][4][5][6][7] However, few of these reports describe how these decisions were reached. 7-10 A more recent study analyzed 53 deaths in 3 PICUs in the United States in which LST was forgone. The author's focus was on the medications given at the end of life and the physicians' and nurses' levels of satisfaction with the care provided. 11 The revelation about the presence of paralyzing agents in some patients at the time of withdrawal of LST and that 2% of the involved professionals believed that hastening death is an acceptable goal in itself generated great controversy. 12-14 Consequently, more information about pediatric intensivists' actual practices regarding forgoing LST in children 8,10,15,16 is warranted. Some details about these events, such as timing of decision making, terminal sedation, barriers to achieve consensus with families, and their participation in the process, remain topics that need additional exploration. 17,18 The purpose of this study was not only to determine the modes of death in a single large multidisciplinary PICU in Canada but also to describe the decision-making process and the end-of-life care. We also examined the level of difficulty to reach consensus with families or ...