Background/Objectives: Artificial intelligence (AI), machine-learning (ML), and deep-learning (DL) models are increasingly used for prognostic prediction in intensive care unit (ICU) sepsis, but their clinical readiness remains uncertain. This systematic review aimed to evaluate AI-, ML-, and DL-based models for mortality and clinically relevant outcome prediction in adult ICU patients with sepsis or septic shock. Methods: PubMed/MEDLINE, Scopus, and the Cochrane Library were searched up to April 2026. Eligible studies included adult ICU sepsis or septic shock cohorts evaluating AI/ML/DL-based prognostic models. Screening, full-text assessment, and data extraction were performed independently by two reviewers. Outcomes, model families, validation strategies, discrimination, calibration, clinical utility, explainability, comparative performance versus conventional severity scores, risk of bias, and reporting completeness were synthesized. Risk of bias was assessed using PROBAST domains supplemented by PROBAST + AI considerations, and reporting completeness was evaluated according to TRIPOD/TRIPOD + AI domains. Results: Seventy-five studies were included, comprising 50 PubMed-derived and 25 additional Scopus-derived studies. AUROC or C-statistic was extractable in 64 studies, external validation was reported in 27, prospective evaluation in three, calibration in 38, decision-curve analysis or clinical utility assessment in 37, and explainability in 64. Across 17 directly extractable within-study comparisons from nine studies, AI/ML models usually, but not uniformly, achieved higher discrimination than conventional severity scores, with a median paired ΔAUROC of +0.108 (IQR, +0.082 to +0.148; range, −0.013 to +0.203). Externally validated fixed-horizon models showed clinically relevant but heterogeneous discrimination across sepsis phenotypes, with stronger evidence in selected sepsis-induced coagulopathy cohorts and more variable transportability in respiratory and liver-injury subgroups. However, 45 studies were judged at high risk of bias, mainly because of limitations in the analysis domain. Conclusions: AI/ML models for adult ICU sepsis show a recurrent signal of prognostic discrimination and often perform comparably to or better than conventional severity scores in directly extractable within-study comparisons; however, this signal should be interpreted cautiously given clinical and methodological heterogeneity, limited prospective validation, incomplete calibration, and frequent high risk of bias. The strongest evidence comes from externally validated, phenotype-specific models, although routine clinical implementation remains limited by heterogeneous endpoints, incomplete calibration, insufficient prospective validation, and scarce workflow-level evaluation. Future studies should shift from retrospective AUROC optimization toward calibrated, externally validated, clinically actionable, and workflow-integrated decision-support tools tested in prospective ICU settings.

Background: Asthma is a heterogeneous chronic airway disease characterized by inflammation, airflow obstruction, hyperresponsiveness, and remodeling. Severe eosinophilic asthma is driven by eosinophilic inflammation, which contributes to tissue damage, recurrent exacerbations, and progressive impairment of airway structure and function. Eosinophils play a central role through the release of cytokines, cytotoxic granule proteins, and extracellular traps, and their persistence in the airways is sustained by type 2 inflammatory pathways, particularly interleukin-5-mediated signaling. A better understanding of eosinophil biology has promoted the development of targeted therapies, including anti-interleukin-5/interleukin-5 receptor agents and biologics that indirectly modulate eosinophilic inflammation, such as anti-interleukin-4 receptor alpha and anti-thymic stromal lymphopoietin antibodies. Aim: This narrative review summarizes the immunopathology of eosinophilic asthma and links eosinophil biology to current and emerging pharmacological strategies. We discuss biologics that directly target the IL-5/IL-5 receptor axis, as well as agents that indirectly modulate eosinophilic inflammation, including IL-4 receptor alpha and TSLP blockade. We also review the clinical positioning of available biologics, focusing on blood eosinophils, FeNO, exacerbation history, oral corticosteroid exposure, lung function, type 2 comorbidities, treatment response, remission and switching. Conclusions: Overall, eosinophilic inflammation remains a central therapeutic target and a key component of precision medicine in severe asthma, but biologic selection should be individualized and reassessed through multidomain clinical outcomes.

MicroRNAs (miRNAs) are small non-coding RNA molecules approximately 20–22 nucleotides in length that regulate gene expression at the post-transcriptional level. By binding to target messenger RNAs (mRNAs), miRNAs inhibit translation or induce degradation, thus influencing a wide array of biological processes including development, inflammation, apoptosis, and tissue remodeling. Owing to their remarkable stability and tissue specificity, miRNAs have emerged as promising biomarkers in both clinical and forensic settings. In recent years, increasing evidence has demonstrated their utility in cardiovascular diseases, where they may serve as diagnostic, prognostic, and therapeutic tools. This systematic review aims to comprehensively summarize the role of miRNAs in cardiovascular pathology, focusing on their diagnostic potential in myocardial infarction, sudden cardiac death (SCD), and cardiomyopathies, and their applicability in post-mortem investigations. Following PRISMA guidelines, we screened PubMed, Scopus, and Web of Science databases for studies up to December 2024. The results highlight several miRNAs—including miR-1, miR-133a, miR-208b, miR-499a, and miR-486-5p—as robust markers for ischemic injury and sudden death, even in degraded or formalin-fixed autopsy samples. The high stability of miRNAs under extreme post-mortem conditions reinforces their potential as molecular tools in forensic pathology. Nevertheless, methodological heterogeneity and limited standardization currently hinder their routine application. Future studies should aim to harmonize analytical protocols and validate diagnostic thresholds across larger, well-characterized cohorts to fully exploit miRNAs as reliable molecular biomarkers in both clinical cardiology and forensic medicine.

Background/Objectives: Extracorporeal membrane oxygenation (ECMO) is increasingly used in adult critical care, but tracheostomy during ECMO remains controversial because of bleeding risk, anticoagulation exposure, and variability in patient selection. This systematic review evaluated the safety, timing, and clinical outcomes of tracheostomy in adult patients receiving ECMO. Methods: A systematic search of PubMed/MEDLINE was supplemented by additional searches in the Cochrane Library/CENTRAL and Scopus to identify studies evaluating tracheostomy in adult ECMO patients. The review was conducted according to PRISMA 2020 guidelines. After database searching and screening, 13 observational studies were included in the qualitative synthesis. Results: Across 13 studies encompassing 1918 patients, tracheostomy during ECMO was feasible and was not associated with procedure-related mortality. Bleeding was the main procedural complication, with reported rates varying according to study design, ECMO configuration, timing of tracheostomy, anticoagulation management, and bleeding definitions. Tracheostomy performed during active ECMO support was generally associated with a higher burden of bleeding or minor procedure-related complications than tracheostomy performed after decannulation. Tracheostomy was performed using percutaneous, surgical, open, or hybrid techniques, although comparative evidence between approaches remained limited. Early tracheostomy was associated with shorter ECMO duration, shorter mechanical ventilation, or improved clinical outcomes in selected cohorts, but timing definitions and outcome measures were heterogeneous. Conclusions: Tracheostomy during ECMO may support airway management, sedation reduction, and ventilatory progression, particularly in prolonged ECMO courses, but it carries a relevant bleeding and transfusion burden. Timing, anticoagulation management, patient selection, ECMO configuration, and procedural technique are likely to influence the risk-benefit balance. Prospective studies are needed to standardize definitions, compare techniques, and clarify optimal timing strategies.

Background: Admission to an intensive care unit (ICU) exposes family members of adult patients to substantial informational, emotional, and decisional burden. In recent years, digital tools have increasingly been used to support communication, information delivery, virtual visiting, psychological support, diary writing, and surrogate decision making in ICU settings, although the available literature remains heterogeneous in terms of intervention type, purpose, timing, and outcomes assessed. Methods: A scoping review was conducted according to Joanna Briggs Institute methodology and reported following PRISMA-ScR. The literature search was performed between January and March 2026 in PubMed/MEDLINE, Scopus, and CINAHL. After duplicate removal, title/abstract screening, and full-text assessment, 32 studies were included in the qualitative synthesis. Results: The included studies were published between 2016 and 2026, used heterogeneous methodological designs, and originated from different international contexts. Six main categories of digital tools were identified: educational websites and online information resources; decision aids and tablet-based tools; virtual visiting and video communication systems; digital diaries and writing practices; psychological support or self-management applications; and digital assessment or family-engagement platforms. Overall, informational and communication-oriented tools appeared to provide the clearest signals of usefulness for family orientation, information access, communication, and relational continuity, whereas evidence regarding psychological and decisional outcomes remained more variable and largely preliminary. Conclusions: Digital tools for family members of adult ICU patients represent a relevant and evolving component of family-centered critical care. Their value appears to depend on the family need addressed, the timing of implementation, and their integration into clinical workflows. Overall, the available literature suggests that digital tools may be particularly useful for family orientation, information access, and communication, whereas their impact on psychological and decisional outcomes remains less certain and requires further investigation.

Fungal biomarkers may support antifungal stewardship in critically ill adults, but their clinical effect depends on the decision rule linking test results to treatment. We systematically reviewed ICU or ICU-relevant adult studies in which fungal biomarkers or rapid fungal diagnostics explicitly informed antifungal management. Searches were updated through 24 July 2026 (PROSPERO CRD420261432481), and clinically compatible Candida randomised outcomes were synthesised by rule direction. Twenty-one reports represented 19 independent studies. Five core Candida trials randomised 677 participants (661 analysed). Rule-in assignment increased systemic antifungal receipt (two trials; RR 2.06, 95% CI 1.58–2.67) without demonstrated patient benefit. In three rule-out trials, mortality was 34/126 versus 38/132 (RR 0.94, 95% CI 0.63–1.40), while only 16 post-randomisation invasive Candida events informed highly uncertain fungal-safety estimates (RR 1.85, 95% CI 0.51–6.65). The evidence is therefore insufficient to establish a clinically acceptable fungal-safety margin for biomarker-guided discontinuation. Biomarkers should therefore support, not replace, risk-based reassessment: timely negative results may justify supervised discontinuation, whereas isolated positivity should not be a stand-alone treatment trigger.