PURPOSE OF REVIEW To critically appraise recent advances that are shifting critical care nutrition from uniform nutrient targets towards phase-guided, phenotype-guided and response-guided care, with particular emphasis on major studies published in 2025-2026. RECENT FINDINGS Recent large trials do not support routine early full-dose energy or high-dose protein delivery in unselected critically ill patients, and some higher dose strategies have generated safety signals. Indirect calorimetry improves measurement of energy expenditure, but measured expenditure is not synonymous with the dose that should be delivered. Emerging data suggest that dynamic responses, including early phosphate trajectories, may help identify metabolic readiness for feeding, although no validated bedside marker exists. Pharmacological intravenous vitamin C has shown no benefit and recurring harm signals in shocked patients, whereas micronutrient care should distinguish physiological requirements and deficiency correction from megadosing. After ICU discharge, intake frequently falls during the transition from enteral to oral nutrition, and intake-guided protocols have achieved high nutritional adequacy in proof-of-concept studies. SUMMARY Critical care nutrition is moving from universal targets towards phase-guided, phenotype-guided and response-guided treatment. Acute-phase caution should not become prolonged underfeeding; repeated metabolic assessment, body composition, reliable intake monitoring and continuity through recovery are central priorities.
Purpose of review Patient–ventilator asynchrony is commonly treated as a technical mismatch requiring correction, although asynchronous breaths are not uniformly harmful and apparent synchrony may conceal injurious effort. This review summarizes evidence from the past 18 months that redefines asynchrony through inspiratory effort, the transition to assisted breathing, and continuous waveform monitoring. Recent findings Potential harm depends on effort magnitude and timing, breath stacking, ventilatory mode, lung and diaphragm vulnerability, dyspnea, and cumulative exposure. Double triggering, reverse triggering, and ineffective efforts may indicate injury, overassistance, or recovery depending on context. During transition, work shifts from ventilator to patient, exposing patients to both excessive and insufficient effort. Continuous analysis can identify burden, clustering, and physiological trajectories missed by intermittent inspection. Artificial-intelligence models increasingly estimate effort and dynamic driving pressure, but heterogeneous definitions, limited external validation, artifacts, and uncertain actionability constrain implementation. Summary Management should move beyond eliminating individual events toward minimizing time outside individualized ranges of acceptable effort and assistance. Monitoring should integrate waveforms with effort, lung stress, diaphragm loading, dyspnea, ventilator settings, and temporal context. Artificial intelligence should support bedside reassessment rather than autonomous adjustment. Multicenter validation and physiology-guided trials must establish whether this approach improves patient-centered outcomes.
PURPOSE OF REVIEW:For three decades, the Sequential Organ Failure Assessment (SOFA) score has been central to critical care, offering a simple, reproducible measure of organ dysfunction. Since the original score [Sequential Organ Failure Assessment-1 (SOFA-1)] [1], advances in organ support, drugs, and monitoring, and an ageing, comorbid population have reshaped practice and rendered several tools outdated. This prompted an international initiative to modernize SOFA while preserving its core strengths: simplicity, reproducibility, feasibility across diverse settings, and longitudinal assessment of organ dysfunction. RECENT FINDINGS:SOFA-2 was developed through a modified Delphi process with 60 international experts and dedicated methodology and validation committees. Expert consensus, systematic reviews, and analyses of multinational databases (high-income, middle-income, and lower-middle-income systems) established its face, content, and predictive validity. The six original organ systems were retained; the gastrointestinal and immune domains were examined but not included because current datasets did not support them. SOFA-2 keeps the architecture of SOFA-1 while updating definitions, reflecting contemporary therapies, and standardizing missing-data handling and the scoring of acute-on-chronic dysfunction. SUMMARY:SOFA-2 is an evolution rather than a reinvention. It must now be tested in the real-world - within definitions such as Sepsis-3, across its derived variables (admission, maximum, delta, discharge), and in varied settings, including low-income and middle-income countries.
PURPOSE OF REVIEW:To provide a practical, physiology-driven framework for advanced hemodynamic monitoring in patients with acute heart failure (AHF) and cardiogenic shock, highlighting and supporting clinicians in integrating monitoring tools into an individualized bedside approach. RECENT FINDINGS:Cardiogenic shock is increasingly recognized as a heterogeneous syndrome involving not only impaired cardiac output, but also systemic inflammation, vasoplegia, tissue hypoperfusion and multiorgan dysfunction. Recent studies have emphasized phenotype-based approaches for both diagnosis and treatment, integrating clinical evaluation, echocardiography, invasive hemodynamics and tissue perfusion markers. Machine learning derived phenotypes have further identified cardiorenal and cardiometabolic profiles associated with distinct prognostic implications. Critical care echocardiography has a central role for rapid diagnosis and serial reassessment, while transpulmonary thermodilution and pulmonary artery catheterization provide complementary information to echocardiography in refractory or complex cases. SUMMARY:Hemodynamic monitoring in AHF and cardiogenic shock should rely on individualized, multimodal and dynamic integration of clinical examination, biomarkers, echocardiography and invasive monitoring. The main goals are to define the underlying phenotype, assess congestion and tissue hypoperfusion, and evaluate treatment response. Echocardiography remains the cornerstone of bedside assessment, whereas invasive monitoring should be reserved for selected high-risk patients with persistent or complex shock.
