Análisis de las decisiones médicas en el triaje de pacientes en unidades de cuidados intensivos

In this dynamic ICU environment, medical decision making plays a critical role inresource management, especially in bed shortage scenarios. Moreover, this mismatchbetween demand for admissions and availability of ICU beds is expected to increase,leading to further rationing of these resources, i.e. triage decisions on patient dischargeand admission. The COVID-19 pandemic has also highlighted the importance of triage inICU bed management decisions. Triage is a crucial process that ensures that the mostcritical cases are prioritised and that resources are allocated efficiently and equitably. These ethical dilemmas are frequently faced by ICU healthcare professionals. In situationsof high bed occupancy, physicians must decide which patient will be assigned to the lastavailable bed. These problems have been widely discussed in the medical literature,identifying the influencing factors and analysing the consequences for patients’ health. Understanding how physicians make admission and discharge decisions in bedpressure situations is essential to improve ICU resource management. Understanding these factors can reduce the risk of bias, optimise bed utilisation and improve the quality of care for critically ill patients by developing more objective and equitable criteria for ICU admission. Current mathematical models used to study ICU management and patient Flow ignore many of the factors that clinicians consider when making triage decisions in the context of real-life situations. This thesis analyses medical decisions related to the triage of admissions and discharges in ICUs in situations of high bed occupancy and proposes a novel simulation framework to accurately represent the performance of these units. The development and application of data analysis and computational simulation techniques and methods have evolved a discrete event simulation model evolved into a digital twin that is able to reproduce, in a virtual environment, the real functioning of an ICU. The digital twin replicates the ICU admission and discharge processes and recreates patients' health status based on real clinical data. Validated by ICU professionals at the University Hospital of Navarra, this tool enables analysis of doctors' decision-making processes regarding patient admission and discharge. It also serves as a valuable learning and training resource. The simulator enables the collection of necessary data to examine decision making in relation to the admission and discharge of patients during periods of high bed pressure. It models three key bed control actions used by ICU managers: admitting or diverting patients from the ED, confirming or cancelling scheduled surgeries, and discharging or retaining patients admitted to the ICU. Physicians' admission and discharge decisions are analised from multiple perspectives, including a holistic assessment and a detailed examination of individual patient decisions, taking into account the level of bed occupancy. Preliminary results of the simulation carried out by different types of users, including intensive care physicians from different hospitals, reveal Certain variability in physicians' decision-making towards the last bed dilemma, both in general and in terms of the evolution of the admission and discharge process as the level of bed pressure increases. The main contribution of this work is the study of the decision-making process in the ICU using a digital twin, which has revealed the existence of variability in decision-making among physicians Whom are faced with identical scenarios with identical scenarios.