Cardiopulmonary Physiology and Mechanical Support

CVS2 is an open-source closed-loop cardiovascular and pulmonary simulator written in MATLAB App Designer. It integrates a real-time hemodynamic engine (28 ODEs covering four cardiac chambers, valves, systemic and pulmonary Windkessel networks, pericardium, intrathoracic pressure, central sympathetic and parasympathetic autonomic integration, three afferent reflex arcs and twelve ICU drugs with receptor-level pharmacology) with bedside-monitor displays, PV-loop and Guyton analysis, a live HTML5 four-view echocardiography viewer and a literature-anchored scenario library (currently 39 calibrated presets). The tool is developed for ICU teaching and curriculum development. It is explicitly not a medical device.

Multicentre, open-label, cluster-randomised cross-over non-inferiority trial comparing two periods of 4 weeks per ICU (sequence AB or BA, with 2–4 weeks wash-out/training): one period withholding nutrition and glucose-containing maintenance solutions for 72 h, the other delivering standard early nutrition. Primary outcome: proportion of patients with ≥1 hypoglycaemic episode (BG <3.9 mmol/L) within 72 h. Key secondary outcomes: ICU/hospital and 90-day mortality, severe hypoglycaemia/hypernatraemia, refeeding syndrome, organ support. Trial registered as ISRCTN16339579.

Retrospective cohort of all adult ICU admissions at USB Basel between January 2019 and January 2026 (approximately 22,510 admissions). Primary outcome: incidence of hypoglycaemia (BG <3.9 mmol/L), two-tier severity classification (Level 1: 3.0–3.8; Level 2: <3.0). Risk-factor analysis (baseline comorbidity, severity scores, 72-h treatment exposures including insulin, total calories, corticosteroids), Cox PH for 30/90-day mortality with hypoglycaemia as time-dependent covariate, Fine–Gray competing-risk model for length of stay. GLMM with each glucose measurement as the unit of analysis to address ascertainment bias.

Retrospective secondary analysis of an existing porcine dataset (Liu et al. 2020) examining how respiratory loading manoeuvres (inspiratory and expiratory holds) and different volume states modify the components of the Venous Excess Ultrasound (VExUS) score and their relationship to invasive hemodynamic parameters. The analysis explores the concept of venous-right ventricular coupling. No new animal experiments are performed; only the previously collected data are re-analysed.

Development and in-silico validation of a model-based method for simultaneous estimation of cardiac output, intrapulmonary shunt, V/Q mismatch and recirculation during venovenous ECMO from non-invasive membrane-lung oxygen-transfer measurements. Forward model inverted by multi-start non-linear optimisation. Tested in 300 virtual ARDS patients across three noise tiers.

National retrospective registry study based on the SGI Minimal Dataset (MDSi) describing the incidence, injury pattern, ICU resource utilisation and outcomes of adult polytrauma and traumatic brain injury patients (AIS head ≥ 3) admitted to Swiss ICUs between 2005 and 2024. Outcomes include nursing workload (NEMS), ventilation days, ICU-based surgical interventions, ICU and hospital length of stay and ICU mortality. Pre-specified hypotheses examine the impact of the 2011 Highly Specialised Medicine (HSM) centralisation of trauma care, the 2012 introduction of SwissDRG flat-rate reimbursement and the COVID-19 pandemic on resource use, injury patterns and outcomes.

International multidisciplinary modified-Delphi consensus on a diagnostic and management algorithm for acute mesenteric ischaemia (AMI). ~50 panelists across vascular surgery, general surgery, gastroenterology, radiology and intensive care. Five algorithm flows (A–E, including a Definitions flow). Two-round REDCap survey using a modified delphi consensus process built in R.

Retrospective cohort study characterising the response to intravenous phosphate replacement in critically ill patients.

Baseline-stratified analysis (significant hypophosphataemia ≤0.6 mmol/L; mild 0.6–0.8; normo- and hyperphosphataemic groups), dose-response analysis, mixed-effects modelling of phosphate kinetics. Renal SOFA included as a previously untested covariate.

Prospective single-centre feasibility study assessing whether changes in thoracic electrical impedance measured by Electrical Impedance Tomography (EIT) can track changes in pulmonary artery pressure in intubated ICU patients with a clinical indication for invasive PA-catheter monitoring. Reference standard is the invasive PA pressure. Secondary outcomes include trending of cardiac output, ventilation parameters and right-ventricular–pulmonary-artery coupling. BMCC Project 2401, ethics approval BE 2024-00435.

Retrospective multicentre registry-based study (national)

Development and internal validation of a new Swiss ICU mortality prediction model (SWISS-MPM) and Standardized Resource Use Ratios (SRUR) for length of stay and NEMS based on the national MDSi registry. Cohort: 250,580 admissions across 77 Swiss ICUs (Jan 2021 – Jun 2024). Group authorship under "Swiss ICU Trial Group".