Respiratory mechanics and patient-ventilator interaction dataset from the ICU

The use of high-resolution data in the intensive care unit (ICU) environment allows for more detailed research into invasive mechanical ventilation (IMV) management and patient-ventilator interactions and their relationship with clinical outcome. The Real-time Agnostic Monitoring for Intensive Care (RAMIC-I) database is a single-center, de-identified dataset containing clinical and physiological data collected from 155 critically ill adults admitted to the ICU of Hospital Universitari Parc Taulí (Sabadell, Spain), between 2015 and 2019. RAMIC-I includes 13,433 hours of mechanical ventilation recorded during the first eight days of IMV, enriched with demographic information, bedside clinical variables (including clinical management variables, administered treatments, and clinical scores), vital signs, and mortality outcomes. Derived variables from respiratory waveforms, including patient-ventilator asynchronies, were aggregated into 10-minute resolution intervals to enable a comprehensive exploration of respiratory physiology during IMV. The dataset is structured according to FAIR data principles and is accessible through a federated repository based on the open-source infrastructure. This detailed resource supports research into predictive modeling, IMV optimization, and the study of patient-ventilator interactions in critical care settings.