End-of-life practices in Brazilian pediatric ICUs of three different regions

Objectives

To evaluate the frequency and types of end-of-life practices in five Brazilian pediatric intensive care units (PICU): Porto Alegre (two), São Paulo (one) and Salvador (two).

Methods

We conducted a cross-sectional and multicenter study based on a retrospective chart review including every death occurring between January 2003 and December 2004 in five Brazilian PICU of university-affiliated and tertiary hospitals located in Porto Alegre (two), São Paulo (one) and Salvador (two). Two fellows of each service filled a standardized protocol, searching for information regarding: demographic aspects, cause of death, frequency of cardiopulmonary resuscitation register plans, and medical practices in the 48 hours before death. The data were compared using the Student t test, ANOVA, chi-square and RR.

Results

A total of 332 death patients were identified in this period and 37 cases of brain death were excluded (11.1%). The mortality is similar in the five PICU. Only 120 (36.1%) patients had not been reanimated, with a statistical difference (P < 0.05) between the five PICU. Sixty-seven (55.8%) charts of non-reanimated patients had a register of an end-of-life plan made by the assisting team: 41 cases of life support limitation and 26 do-not-resuscitate orders, without differences between the five PICU (P = 0.2). There were no cases of withdrawing ventilatory support or a significant increase in sedatives and analgesics doses in the 48 hours preceding death.

Conclusion

Despite the increasing number of children who are not reanimated in the end of life in Brazilian PICU, we observed that withdrawing life-sustaining treatment preceding death is still insignificant. Moreover, we observed different medical practices in the five hospitals that can be a consequence of cultural, religious or even personal behaviors of each medical team in the three different Brazilian regions.