PROSAFE data collection software, characterized by a modular structure, allows the activation of secondary data collections on different topics of clinical relevance. Among the active projects, infection surveillance is underway to continuously monitor the epidemiology of infections in intensive care units. Numerous collaborations have developed in this context. One of the most significant is the one with the Department of Public Health and Pediatric Sciences at the University of Turin, which involves a comparison of prevalence and incidence data of care-related infections collected in the Intensive Care Units adhering to the GiViTI group and incidence estimates measured through the third Italian Health Plan (PPS). In 2024, an ongoing collaboration was also initiated with the Sacco Hospital in Milan, where several specific projects are active, aimed at estimating the incidence of bloodstream infections acquired in the ICU and of pneumonia associated with mechanical ventilation in the participating wards. As part of an initiative coordinated by the Istituto Superiore di Sanità, some of the information collected is being shared with the European Centre for Disease Prevention and Control (ECDC), contributing to European ICA surveillance. In addition, analysis of samples collected with the AbioKin study, a multicenter, prospective observational study with the aim of studying the pharmacokinetics of some of the most commonly used antibiotics in critically ill patients in the ICU, is ongoing. The project is being conducted in centers using the MargaritaTre electronic medical record. The first phase, now completed, involved the collection of clinical data and plasma samples in patients undergoing antibiotic therapy for one of the antibiotics under study. In the second phase, which is still ongoing, pharmacokinetic models are being developed that can describe the evolution of plasma concentrations as a function of the patient's clinical condition in order to improve the personalization of antibiotic therapy in the critically ill patient.

Within a context of international collaboration aimed at the continuous improvement of the quality of ICU care, GiViTI has joined the LOGIC (Linking of Global Intensive Care) network https://icubenchmarking.com/ , a global platform that unites several national ICU registries. LOGIC represents an innovative initiative that currently involves 17 countries spread across several continents, united by the goal of systematically collecting and comparing clinical and organizational data. Through this network, participating countries can define strengths and areas for improvement in their respective healthcare systems. This exchange of knowledge and data allows for shared critical reflection, stimulating the adoption of best practices and evidence-based strategies on an international scale.

ICU departments are high-technology care settings in which highly qualified personnel are required. These characteristics result in conspicuous management costs and special attention to the appropriate use of resources. The StART study (Study on the Appropriateness of ICU Admissions) active in Tuscany and Piedmont fits into this context. The aim of the project is to evaluate the health care provided in the wards with respect to the intensity and severity of services, taking into consideration the levels of care that can be delivered as a function of both available equipment and personnel. The aim of the project is to assess any factors of inappropriateness, identify the reasons for them, and define goals for improvement.

The main objective of the project is to implement a permanent system for performance evaluation through the analysis of data entered into the MargheritaTre electronic medical record, currently used by about 70 departments in Italy. The secondary objective is the creation of an anonymized database accessible to the scientific community. The project is being coordinated by a multidisciplinary study group with medical, nursing, statistical and IT expertise.

The availability of appropriately structured large datasets offers significant advantages in health care research, enabling improved clinical practice and supporting decision making. In this context, clinical data sharing is critical to exploit the full potential of these databases to validate and extend research project results and disseminate scientific knowledge. Sharing clinical data requires a multidisciplinary effort, addressing legal, ethical, technical, and scientific issues to ensure patient privacy and data security on the one hand, and accessibility of information to researchers on the other.

Also using data collected through the Clinical Record, we are also working on a project using an analysis strategy called emulated target trial, aimed at studying the effects of one or more clinical treatments by simulating the evolution of clinical history from data collected on real patients. Specifically, it aims to simulate the clinical course of patients admitted to the ICU with a diagnosis of severe hypoxemia to evaluate the effect of different ventilation modes on the risk of ICU mortality. The first step is to clearly define all the elements of a hypothetical randomized clinical trial: inclusion and exclusion criteria, treatments to be compared, outcomes to be analyzed, duration of follow-up, and analysis plan. These elements are then applied to real data, collected from the MargheritaTre electronic medical record, to build a dataset consistent with the simulated study design. Next, predictive models are developed to estimate relevant variables and simulate the clinical course of patients in the so-called “natural course,” that is, without any intervention. This step is essential to verify the reliability of the models used. Finally, one or more variables are modified according to the defined intervention strategies to assess their impact on the mortality outcome.