- Systematic review and meta-analysis of 32 studies including 11,943 critically ill patients, aimed at identifying prognostic factors for catheter-related thrombosis.
- Pooled adjusted odds and hazard ratios came from random-effects models, so the findings describe associations rather than tested interventions.
- Increasing age was associated with an aOR of 1.16 per 10-year increase (95% CI 1.06 to 1.27), and internal jugular insertion with an aOR of 3.76 (95% CI 2.82 to 5.01).
- The median incidence of catheter-related thrombosis was 13.3% (IQR 3.7% to 31.3%), and the authors describe substantial variation across studies.
Critical care
Research from anesthesiology, pain and critical care journals, sourced from PubMed.
Sorted by publication date, not import date. Recent filters exclude papers with only a publication year; find those under All years.
AI-generated abstract summaries can contain errors. Verify findings against the original paper before applying them to patient care.
Last successful paper check: 16 Nov 2026 17:18 UTC.
0 newly discovered papers saved; these are not necessarily published today.
2 papers awaiting AI summaries. Summary processing had an error; paper-fetch status is separate.
Background checks run while the server is active. This does not guarantee a daily run if hosting sleeps.
2 papers awaiting AI summaries. Summary processing had an error; paper-fetch status is separate.
Background checks run while the server is active. This does not guarantee a daily run if hosting sleeps.
- This exploratory secondary analysis reconstructed venous return parameters from 150 published septic shock hemodynamic observations, with no intervention.
- Observations were sorted by empirical quartiles into four recurrent circulatory phenotypes: reduced filling pressure/gradient, preserved venous return, elevated venous pressure, and high-resistance venous return.
- In exploratory adjusted analyses, venous return driving pressure was strongly associated with cardiac output (β 1.03 L/min/mm Hg; 95% CI 0.80-1.25; p < 0.001).
- The authors conclude that published septic shock data occupy distinct Guytonian phenotypes and that venous return geometry may help interpret circulatory heterogeneity.
- Retrospective multicenter cohort of 28,211 adults from 140 Japanese hospitals, with emergency ambulance admissions for community-acquired sepsis without initial organ support.
- ICU (n=1,156) and HDU (n=8,706) admissions were each compared with general ward admission, using overlap propensity-score weighting.
- Overall, neither HDU (RR 0.99, 95% CI 0.92-1.06) nor ICU (RR 1.06, 95% CI 0.92-1.22) was associated with lower in-hospital mortality.
- HDU subgroups with SOFA ≥6 or GCS ≤12 had lower mortality, and the GCS ≤12 subgroup was cost-effective under both the Japanese and U.S. thresholds.
- This multicenter retrospective cohort study of 1087 adults with acute kidney injury across eight South Korean tertiary hospitals included patients who survived at least 90 days after CKRT started.
- Patients were grouped by albumin status at CKRT initiation and on day 7 into persistent, resolved, new-onset, or no hypoalbuminemia, with no intervention assigned.
- The composite of end-stage kidney disease or 365-day all-cause mortality occurred in 33.4% (363/1087), and every hypoalbuminemia group had higher adjusted odds than the no-hypoalbuminemia group, with persistent hypoalbuminemia at aOR 2.64 (95% CI 1.73-4.04).
- The authors conclude that early albumin changes during CKRT are linked to long-term kidney failure and mortality, and that a decline in albumin is associated with worse prognosis even without baseline hypoalbuminemia.
- Secondary analysis of a prospective observational study of 33 mechanically ventilated adult ICU patients (median age 49, 27% women) with serial vastus lateralis biopsies at ICU day 5 and day 15, plus healthy controls.
- No intervention was given; the exposure was critical illness, with muscle mitochondrial structure and electron transport chain function compared against healthy controls.
- Intact mitochondrial profiles were reduced at day 5 (fold change 0.24, p = 0.039) and day 15 (0.20, p = 0.011), and complex III activity stayed reduced after citrate synthase normalization at both time points (day 15: 0.87, p = 0.007).
- The authors conclude that mitochondrial ultrastructure is altered by day 5 and remains abnormal at day 15, with persistent complex III/IV dysfunction, describing an early sustained remodeling phenotype.
- Systematic review of 11 eligible English-language studies, with 128,181 adults in critical care or post-ICU recovery settings, searched through March 2026.
- The studies measured language concordance, meaning a match between the patient's preferred language and the care team's language.
- Six studies examined clinical decision-making, but heterogeneity in analyses and outcomes meant the data could not be pooled quantitatively.
- The authors conclude current evidence suggests concordant care may influence clinically meaningful critical care processes, and that interventional studies are warranted.
- Pilot type 1 hybrid stepped-wedge cluster randomized trial with mixed-methods evaluation, across four Hong Kong emergency departments, in adults with suspected infection and NEWS-2 of 5 or higher.
- NEWS-1 care triggers the Surviving Sepsis Hour-1 Bundle when NEWS-2 is 5 or higher, compared with standard care.
- Lactate, blood culture and antibiotic compliance rose significantly, though overall bundle compliance reached only 56.1%, and the authors judged a full-scale trial feasible with a larger sample.
- Exploratory 30-day mortality did not differ significantly (3.9% vs 9.2%; absolute risk difference 5.3%, 95% CI -1.6% to 12.1%), and adverse events were rare.
- A systematic review and meta-analysis of 50 studies covering about 7.8 million pregnant and postpartum women in 20 countries, all from tertiary hospitals.
- Pooled ICU admission prevalence was compared across country income levels, with low-income countries showing the highest rate.
- Overall, 1.0% (95% CI, 0.8-1.3%) of pregnant and postpartum women were admitted to ICU; 72.3% of admissions had known obstetric causes, led by hypertension (26.9%) and hemorrhage (23.3%).
- Maternal mortality among ICU-admitted obstetric women was 8.5%, and the authors call for expanding ICU services in resource-limited settings.
- Nationwide observational registry study of consecutive adults with initial nonshockable out-of-hospital cardiac arrest in South Korea, 2016 to 2021.
- Patients were grouped by rhythm change during CPR into four transition patterns, with inverse probability of treatment weighting adjusting for measured confounding.
- Good neurologic outcome (CPC 1-2) rose from 0.101% with persistent asystole to 0.874% with persistent PEA; asystole-to-PEA OR 2.28 (95% CI, 1.78-2.93).
- Asystole converting to PEA was linked to better outcomes than persistent asystole, and PEA-to-asystole did better than asystole-to-PEA (OR 1.66, 95% CI 1.39-1.99).
- A Bayesian systematic review and meta-analysis of 54 studies, including 367,896 critically ill adults, examined new-onset atrial fibrillation in general ICUs and high-dependency units.
- It compared post-discharge systemic thromboembolism in patients with and without new-onset atrial fibrillation, with follow-up ranging from 1.5 months to 5 years.
- Pooled NOAF incidence was 10.6% (95% CrI 8.6-12.9%), and post-discharge thromboembolism risk was 7.1% (95% CrI 4.5-11.0%). Relative risk was 1.94 (95% CrI 1.23-3.29) across six studies, rising to 2.01 with adjustment.
- Anticoagulation use was low, at 32.9% in-hospital and 35.9% post-discharge, and 23.9% of patients with NOAF were later diagnosed with atrial fibrillation.
- Pre- and postimplementation observational study using surveys of families and clinicians across medical and surgical ICUs at one academic teaching hospital.
- The intervention was a structured, scripted multidisciplinary rounding template that included families and bedside nurses in daily ICU rounds.
- In the MICU, families showed statistically significant gains in perceived interest in patient needs, control of agitation and pain, and consistency and honesty of information, while SICU changes were not significant.
- The authors concluded that family-inclusive structured rounds were associated with greater family confidence in information quality and honesty and perceived staff compassion, and clinicians generally viewed family inclusion positively.
- This retrospective analysis of the ELSO registry included 115,849 adults on ECMO across 495 hospitals from 2012 to 2023.
- Hospital ECMO experience was measured by annual adult case counts and total ECMO hours, and the analysis compared centers with four or more adult ICUs to those with one.
- Higher annual case counts were associated with lower in-hospital mortality odds (OR 0.98, 95% CI 0.97-0.99 per 10 cases), with a similar but smaller association for ECMO hours.
- Centers delivering ECMO across four or more ICUs had higher adjusted mortality than one-ICU centers (OR 1.18, 95% CI 1.06-1.31), which the authors partly attribute to case complexity and offset by volume.
- Pre-post quality improvement study of 8,909 adults on mechanical ventilation across 14 hospitals in one Midwestern health system, 2022 to 2024.
- A daily afternoon tele-critical care nurse audit prompted bedside nurses when awakening or breathing trials were undocumented and no exclusion criteria applied.
- Awakening trial documentation improved faster after the intervention, rising 1.4% per 30 days versus 0.2% before (interaction p<0.001).
- Breathing trial documentation rose 12.8% (95% CI, 8.7 to 16.9%) in small ICUs but not large ICUs (interaction p=0.02), which the authors link to local ICU characteristics.
- Single-centre prospective cohort of 89 analysed adults (100 recruited) undergoing major vascular surgery, with point-of-care ultrasound and sidestream darkfield microscopy used to assess causes of hypotension.
- Hypotension (SBP under 90 mm Hg or vasopressor infusion) within 24 hours was the primary outcome, and its causes were vasodilation, myocardial dysfunction, severe hypovolemia and obstruction.
- Hypotension occurred in 23%; adjusted risk differences were 33% (95% CI, 14% to 53%) for vasodilation and 41% (95% CI, 3% to 79%) for myocardial dysfunction.
- Vasodilation was also linked to higher ICU admission (27%) and vasopressor need (22%), and severe hypovolemia and obstruction were not observed.
- Ten pigs were randomised to hypo- or hyperinflammatory ARDS models and imaged with dual-energy CT at individualised PEEP levels in supine and prone positions.
- Hyperinflammatory injury was induced with endotoxin and fluid loading, while hypoinflammatory injury used lung lavage followed by injurious ventilation.
- Hyperinflammatory injury had more severe V/Q mismatch, with normal V/Q at 35% (sd=10) versus 51% (sd=8) for hypoinflammatory injury at low PEEP supine (P<0.01).
- Prone positioning reduced high V/Q and dead space in hypo-inflammatory injury (interaction F1,50=4.32, P=0.04), whereas PEEP reduced low V/Q and shunt supine regardless of subphenotype.
- Secondary analysis of 3000 patients across 21 EPOCH trial centers, covering 4288 urgent ICU consultations grouped into care episodes.
- No intervention was assigned in this analysis; the exposure was whether patients received resuscitation during their first urgent ICU consultation.
- Half of last care episodes ended in urgent ICU admission (49.2%), with 0.1% in-consultation deaths; late escalation was 18% for first and second consultations.
- Among patients with a second consultation, prior resuscitation was linked to more resuscitation events (23.2% vs 3.4%, p<0.01), but late escalation difference was not significant (31.8% vs 15.6%, p=0.07).
- Prospective longitudinal mixed-methods study of 82 enrolled adults who had at least 72 hours of ICU stay for acute respiratory failure or sepsis, with 67 providing activity data.
- No intervention was given; daily step counts were tracked by accelerometer from hospital discharge through 3 and 6 months.
- Median daily steps rose from 1358 at baseline to 2305 at 3 months and 2826 at 6 months, while median sedentary time at 6 months was 300 min/day.
- Survivors improved gradually but stayed below recommended activity levels, and interviews with 11 participants identified motivational and contextual barriers that accelerometry did not capture.
- Retrospective external validation of SOFA-2 for ICU mortality in 64,015 adult ICU patients from MIMIC-IV at a single Boston center, 2008-2022.
- The comparison was across demographic subgroups (age, sex, race and ethnicity, language, insurance); no interventions were given.
- Overall AUROC was 0.77 (95% CI, 0.76-0.77), falling from 0.85 in ages 18-44 to 0.72 in ages 75 or older (ΔAUROC -0.14).
- Discrimination was lower in non-English speakers (ΔAUROC -0.04), and patients with missing language data had 23.1% mortality versus 7.2% overall.
- Single-center, randomized, double-blind, placebo-controlled pilot trial of 100 critically ill adults with CAM-ICU-confirmed delirium in surgical and trauma ICUs.
- Patients received 2 mg guanfacine or matching placebo nightly for up to 14 days, or until ICU discharge or delirium resolution.
- Median delirium and coma-free days were 2 (IQR 0-6) with guanfacine versus 1 (IQR 0-4) with placebo (p = 0.02) in unadjusted analyses, but not significant after adjustment.
- Feasibility targets were mostly met, with 100% adherence, retention, and data completeness, though COVID-19 limited recruitment to 0.9 patients/week; authors judge the signal worth a multicenter trial.
- This methods paper uses data from the MENDS trial to examine composite endpoints for critical care delirium trials.
- It evaluates two existing composites: delirium- and coma-free days, and a ranked endpoint that treats death as worst and orders delirium- and coma-free days among survivors.
- The authors describe how to define treatment effects on each composite, assess their clinical interpretability, and visualize how effects accumulate over follow-up.
- They provide practical recommendations for composite endpoints that could extend to other randomized trials using similar endpoints.
- Target trial emulation using data from seven ARDS Network and PETAL Network randomized trials, with 292 patients assigned to each of normocapnia and mild hypercapnia.
- Normocapnia (target Paco2 35-45 mm Hg) was achieved by increasing respiratory rate, compared with mild hypercapnia (target Paco2 50-55 mm Hg), over 4 days of lung protective ventilation.
- 90-day in-hospital mortality was 28.8% in both groups (84 of 292 each; log-rank p = 0.842), and median respiratory rate was higher in the normocapnia group (33.0 vs 28.0 breaths/min).
- The authors concluded normocapnia was not associated with lower mortality than mild hypercapnia, and this persisted in older and obese subgroups.
- Multicenter prospective cohort of 480 adults with sepsis plus shock and/or acute respiratory failure, enrolled in medical and surgical ICUs in Nashville, TN.
- Plasma ccf-mtDNA was measured on study days 1, 3, and 5 and related to delirium and coma; no intervention was given.
- Higher ccf-mtDNA was associated with lower odds of delirium (OR 0.93, 95% CI 0.86-0.99; p = 0.048), but not with coma or 3- or 12-month cognition.
- Mortality association was marginal (HR 1.09, 95% CI 0.99-1.20; p = 0.08); the authors urge cautious interpretation until the finding is replicated.
- A cross-sectional, multicenter survey of 514 healthcare professionals working in ICUs, a cardiac telemetry ward, and operating rooms.
- Alarm fatigue was measured with the Dutch Charité Alarm Fatigue Questionnaire, plus self-rated fatigue and perceived false-alarm percentage; no intervention was applied.
- Telemetry staff reported the highest self-perceived fatigue (B = 0.65; 95% CI, 0.06-1.24; p = 0.032) and OR staff the lowest (B = -1.01; 95% CI, -1.59 to -0.43; p < 0.001).
- Physicians and staff with under 3 months' experience reported lower alarm fatigue, and the authors call for department-specific interventions and longitudinal research.
- Secondary analysis of a stepped-wedge cluster-randomised trial (ERIC) including 620 analysed patients from 10 German ICU clusters, assessed 6 months after ICU discharge.
- Patient-reported disability (WHODAS 2.0) was assessed as a predictor of handgrip strength and 2-Minute Walk Test distance, with no separate intervention arm analysed.
- Higher disability was moderately correlated with lower handgrip strength (Spearman -0.41) and shorter walk distance (Spearman -0.47), but prediction limits of agreement exceeded the MCIDs.
- The authors conclude that remote WHODAS 2.0 assessments cannot replace in-person physical function testing, and 59% of patients did not complete the 2MWT.
- Population-based cohort of 37,250 adults discharged alive from sepsis hospitalization in Ontario, Canada, 2010-2020.
- The exposure was the first outpatient neutrophil:lymphocyte ratio (NLR) measured 15-90 days after discharge, with a median NLR of 2.7.
- One-year rehospitalization or death rose across NLR quartiles, from 34.2% to 64.4%, with an adjusted HR of 2.25 (95% CI 2.15-2.35) for the highest quartile versus the lowest.
- The authors suggest NLR, a simple and widely available biomarker, may support early post-sepsis risk stratification, and they note NLR often returned to each patient's presepsis level, which was also prognostic.
- Retrospective population-based cohort of 237,075 patients who survived sepsis hospitalization in Catalonia, Spain, between 2005 and 2019.
- No intervention was studied; exposures were patient factors including age, comorbidity burden, and length of hospital stay.
- Two-year mortality was 42% (99,571 deaths), rising from 5.8% under age 15 to 62% over 84, and from 21% to 67% across Charlson index categories.
- Neoplasia (HR 2.011), cirrhosis (HR 1.562) and dementia (HR 1.304) were most strongly associated with death, and mortality declined during 2017–2019.
- Multistage psychometric study using provincial Ontario FS-ICU rollout data, plus an independent multicenter Canadian dataset for external validation.
- The 29-item FS-ICU was cut to 12 items (seven FS-Care, five FS-DM) using empirical criteria and expert ratings from a 12-member working group.
- The shortened form had high internal consistency (Cronbach's α 0.93 FS-Care, 0.86 FS-DM, 0.94 total), and its scores differed from the full version by ≤1.1 points on a 0-100 scale (ICC 0.97-0.99).
- The authors conclude the 12-item FS-ICUshort keeps the core content and psychometric quality of the original and supports routine use for ICU family experience measurement and benchmarking.
AI summary not available. Read the source abstract on PubMed.
- This scoping review of 20 studies (14 adult, 6 pediatric) from North America, Europe, and Asia included only observational or pre-post intervention designs.
- The studies examined early warning scores such as NEWS/NEWS2, MEWS, PEWS, Rothman Index and machine-learning models, measured at or near ICU discharge.
- Seventeen of 20 studies found higher scores at discharge were significantly linked to unplanned ICU readmission, while three found no significant association.
- The authors suggest these scores may support clinical judgement on discharge readiness, but they call for multicentre prospective studies to test whether EWS-guided protocols improve outcomes.
- Retrospective nationwide cohort using the Turkish Ministry of Health database, covering 3,362,539 adult ICU admissions from 2018 to 2023.
- Describes 22 ICU types and their admission profiles and outcomes, with no comparator group.
- Mixed ICUs had the largest share of admissions (31.4%), but their share of beds fell from 44.0% in 2018 to 25.1% in 2023.
- Authors describe a shift toward specialization and note that the impact of ICU diversity on resource use and outcomes needs further study; 7-day readmission was 3.0%.
- Retrospective multicenter cohort of 1,377 children with in-hospital cardiac arrest treated with ECPR, from ELSO registry hospitals in 2020 to 2023, split into 1,060 derivation and 317 validation patients.
- No intervention was compared; a penalized logistic regression (LASSO) model used 14 routinely available clinical variables to predict survival to hospital discharge, then fed hospital-level risk-standardized survival rates.
- The model showed fair discrimination (AUC 0.728 derivation, 0.678 validation) and good calibration (observed-to-expected ratio 0.995, Brier score 0.222 in validation).
- Between-hospital variation had a median odds ratio of 1.30 (95% CI, 1.11-1.93); the authors state external validation and case-mix review are needed before benchmarking.
AI summary not available. Read the source abstract on PubMed.
- This two-phase environmental scan combined website searches with 21 interviews of ICU stakeholders across Alberta, Canada, developed with Indigenous partners.
- The scan looked for critical-care-specific resources, programs, and policies for Indigenous patients and families, rather than testing an intervention against a comparator.
- The scan found a paucity of ICU-specific Indigenous resources, with only two national resources identified, and interviews confirmed the gap.
- Stakeholders described barriers including inadequate culture training, staff time and burnout, physical space limits, and too few Indigenous liaisons and social workers, which the authors link to equity and trust.
- Prospective multicenter observational study across 75 ICUs, including 55,963 preoperatively evaluated adult inpatient surgical patients, grouped by admission type.
- Of 2764 patients planned for ICU admission, 804 could not be admitted due to limited beds, and 411 did not need postoperative ICU care; 1980 were admitted.
- Overall ICU and 90-day mortality were 5.9% and 7.5%, highest in the emergent admission group at 18.1% and 18.9%.
- Emergency surgery (OR 12.4), intraoperative complications (OR 3.1), and ASA >2 (OR 2.7) independently increased unplanned admission risk; the authors conclude combined criteria help predict ICU need.
- Retrospective single-center cohort of 229 children with culture-positive postoperative sepsis after congenital heart surgery, 2018 to 2022.
- Serial neutrophil-to-lymphocyte ratio was measured preoperatively, at sepsis diagnosis, and on days 7 and 14, and compared between survivors and nonsurvivors.
- Sixty-six patients (28.8%) died in hospital; day 7 NLR had the best discrimination for mortality (AUC 0.74, 95% CI 0.67-0.81).
- After multivariable adjustment, NLR at sepsis diagnosis (OR 1.06, 95% CI 1.01-1.12) and on day 7 (OR 1.04, 95% CI 1.01-1.08) were each independently associated with in-hospital mortality.
AI summary not available. Read the source abstract on PubMed.
- A retrospective single-center cohort of 6591 adults receiving invasive mechanical ventilation had chest radiographs scored by a deep learning model for ARDS findings.
- The exposure was the deep learning ARDS-CXR score, which was compared with Pao2/Fio2 and APACHE-IV for predicting outcomes.
- Twenty-eight-day mortality was 35.1% with a positive score versus 13.5% with a negative score, and the AUROC for mortality was 0.68 (95% CI, 0.66-0.70) versus 0.64 for Pao2/Fio2.
- After adjustment, each 1-point higher ARDS-CXR score was independently associated with higher 28-day mortality or longer time to extubation, and the authors suggest possible uses in care, research enrichment, or ARDS definition.
- Prospective multicenter observational study of 303 adult ICU patients on vasoactive agents with radial arterial catheters, across 13 Polish ICUs.
- Invasive radial and noninvasive brachial mean arterial pressures were compared, with a gradient above 11 mm Hg defined as clinically significant; no intervention was given.
- Fifty-six patients (18.4%) had a gradient above 11 mm Hg, and vasoactive drug dose alone was independently associated with it (OR 1.68, 95% CI 1.12-2.52).
- ICU mortality was 48.0% with a significant gradient versus 30.1% without, and the authors link the gradient to greater illness severity.
- Retrospective cohort study of prospectively collected National Emergency Airway Registry for Children data from international PICUs and PCICUs, 2015-2023, including 27,773 intubations.
- TI-CA, defined as arrest or at least one minute of chest compressions within 20 minutes of tube placement, occurred in 1.2% (n=331) of intubations.
- Oxygenation failure (aOR 2.29), shock (aOR 4.67), and each added attempt (aOR 1.70) raised odds, while video laryngoscopy (aOR 0.76, 95% CI 0.6-0.97), ketamine, propofol, and neuromuscular blockade lowered them.
- The authors conclude that odds rose with each attempt and video laryngoscopy was protective, with reduced attempts accounting for 13% of its benefit on mediation analysis.
- Secondary analysis of data from a randomized trial, with 50 cardiac surgery patients propensity-matched 1:20 to 989 historic institutional patients receiving standard care.
- The comparison was continuous serratus anterior plane catheters versus standard care, with matching on age, sex, smoking, surgery type and pre-existing respiratory disease.
- Pneumothorax occurred more often with catheters (12% vs 3%; OR 5.3; 95% CI 2.1 to 13.6; P < 0.001), while ICU readmission, ventilation and length of stay did not differ.
- The authors found the catheter group had more pneumothorax but note causality is unclear, possibly reflecting trial monitoring differences and baseline population risk.