Le travail de recherche présenté dans cette thèse est l’un des premiers dans le domaine de la gestion optimisée des services d’aide médicale urgente en France. Il est conduit dans le cadre d’un projet ANR qui vise à proposer de nouveaux scénarios d’organisation pour le SAMU du département du Val-de-Marne pour offrir aux patients l’accès adapté aux soins, tout en disposant de ressources limitées. Pour cela, nous développons un modèle de simulation à évènements discrets qui modélise et évalue la performance actuelle de ce système complexe et identifie des pistes d’amélioration susceptibles de réduire la durée entre la réception de l’appel et l’arrivée d’une équipe sur le lieu de l’accident, appelé temps de réponse, qui est un aspect critique dans les systèmes d’urgence pré-hospitaliers. Ce modèle de simulation est utilisé pour quantifier l’impact de divers scénarii se rapportant aux nombres de ressources humaines et matérielles, à la localisation de ces ressources à travers le territoire du Val-de-Marne de manière statique et à l’affectation de ces ressources aux interventions. En outre, nous avons effectué des analyses de sensibilité sur différents paramètres du modèle comme le nombre d’appels reçus, les temps de trajet et les temps de service. Finalement, nous avons utilisé deux approches d’optimisation par simulation afin d’étudier l’impact de la relocalisation des équipes dans différentes bases du département plusieurs fois dans une journée afin de tenir compte des fluctuations des temps de trajet et du nombre de ressources. Les résultats de ces deux approches ont été analysés puis comparés au regard des temps de calcul et de la distribution du temps de réponse. / The research addressed in this thesis is one of the first studies to address the Emergency Medical Service in France, known as SAMU (which stands for the French acronym of Urgent Medical Aid Services). It is funded by the French National Research Agency and aims at improving the organizational processes of the Val-de-Marne department SAMU system in order to meet the population’s needs under limited resources. For this purpose, we develop a discrete event simulation model in order to assess the current performance of this complex system, as well as to investigate the effects of potential process changes that would lead to enhanced operational efficiency, in terms of response time performance (i.e. the period between the receipt of a call and the first arrival of a rescue team at the scene), which is a critical aspect for SAMU providers. This model was used as a decision-support tool for comparing the relative benefits of several scenarios mainly related to the needed resource levels and static location of rescue teams throughout the Val-de-Marne area and their assignment to incoming calls. Sensivity analyses were also performed by changing values of some input parameters such as arrival rates of calls, travel times and service times. Finally, we used two simulation optimization approaches to analyze the impact of rescue teams assignment to bases by considering temporal fluctuations of travel times and number of resources during a day. Experimental results of the two approaches were analyzed and compared regarding computational times and response time distribution.

[pt] As doenças do rim e trato urinário contribuem com cerca de 850 mil mortes a cada ano, sendo a décima segunda causa de morte do mundo. No Brasil, a diálise ainda é o procedimento mais utilizado, apesar de o transplante ser a modalidade mais recomendada, por oferecer melhor qualidade de vida ao paciente, uma possível redução do risco de mortalidade e menor custo que a diálise. Uma vez na fila, o paciente ainda se depara um conjunto de ineficiências do Sistema Nacional de Transplante. A presente pesquisa identifica que uma delas é o desalinhamento entre os atores do transplante (doador, receptor e unidade transplantadora) e que há um gap no que tange às atividades do ator unidades transplantadoras. Dessa forma, o objetivo da presente pesquisa é investigar, à luz da gestão de operações, o modelo de uma unidade transplantadora de referência, inserida na rede de transplante renal do estado do Rio de Janeiro. A partir do levantamento da literatura e de campo, são realizadas modelagens do processo de transplante renal, descrição e diagnóstico do modelo de operação da unidade transplantadora e proposições acerca do projeto de operação da unidade estudada / [en] Diseases of the kidney and urinary tract contribute about 850,000 deaths each year, being the 12th leading cause of death in the world. In Brazil, dialysis is still the most used procedure, although transplantation is the most recommended modality, because it offers a better quality of life for the patient, a possible reduction of mortality risk and lower cost than dialysis. Once in the queue, the patient still faces a set of inefficiencies of the National Transplant System. The present research identifies that one of them is the misalignment between the actors of the transplant (donor, receiver and transplantation unit) and that there is a gap with respect to the activities of the transplantation unit actors. Thus, the objective of this research is to investigate, in the light of operations management, the model of a reference transplantation unit, inserted in the kidney transplant network of the state of Rio de Janeiro. From the literature and field survey, modeling of the renal transplantation process, description and diagnosis of the operation model of the transplantation unit and propositions about the operation project of the studied unit are performed.