Post-event analysis (PEA) is a collective support for healthcare teams put in place after each death of a hospitalized child. PEA is made up of two distinct parts: first a chronological reconstitution of the event; in a second time, analysis of practices by the multidisciplinary team. Its systematic and recurrent methodology implies regulation by inter-professional mediation; it guides exchanges thanks to the analysis of practices. It increases the development of the individual and collective act of work. It is a new channel of support linked to the creativity and collective expertise of the team, and this device accompanies the team progress over time. The team’s participation in the improvement of its practices and organizational conditions provides an additional resource to deal with the extreme psychological distress associated to the relationship with the child and the bereaved family. The article presents the PEA from a socio-psychoanalytic approach. It explains the main characteristics of the humanization of the healthcare sector. We present the current aim of healthcare improvement with a systemic approach. The practice improvement brings benefices for the patients (child and family) as well as for the worker’s team. In this process, the team workers are involved as actors with a real impact on their practices and on work organization. This process takes into account collective field experience, with strong attention paid to ethics. The PEA efficiency is explained by the permanence of the process with the auto-evaluation of practices and a memory of the decisions and their effects. This collective work analysis increases the team cohesion and responsibility, it helps to gain control upon the situation. All these factors increase the individual and the collective team’s resilience, when workers face painful work situations which are part of their work. Finally, this article questions the institutional responsibility in the choice of preventive devices to support situations causing suffering at work. This article is based on a fifteen-year experience; the model is illustrated by a concrete situation of child’s death.