BACKGROUND:To enhance biological understanding of ARDS, pneumonia, and sepsis and to accelerate therapeutic development in these areas, the National Institutes of Health developed the ARDS, Pneumonia, and Sepsis (APS) Consortium. RESEARCH QUESTION:Is the APS Consortium study rapidly generating data and biospecimens from a large cohort of critically ill adults with ARDS, pneumonia, and sepsis that will facilitate phenotyping of these syndromes? STUDY DESIGN AND METHODS:The APS Consortium Phenotyping Study is a multicenter, longitudinal, prospective, observational cohort study aimed at enrolling 4,000 critically ill adults with ARDS, pneumonia, sepsis, or a combination thereof over 4 years. Data and biospecimens are collected to characterize many aspects of each participant's chronic health, acute illness, and long-term recovery to facilitate phenotyping, that is, subclassifying ARDS, pneumonia, and sepsis into precise biologically based subsets with shared pathophysiologic characteristics. Feasibility of the study was assessed by evaluating the first 1,000 participants in terms of recruitment pace, participant characteristics, biospecimen collection, and proportion with confirmed ARDS, pneumonia, and sepsis based on expert adjudication. RESULTS:The first 1,000 participants were recruited ahead of schedule in < 13 months. Median age was 64 years, 75% received vasopressors, 50% received invasive mechanical ventilation, and 25% died in the hospital within 4 weeks of enrollment. Biospecimen collection rates were high, with 99% of participants with blood samples, 98% with upper respiratory swabs, 37% with lower respiratory samples, 80% with urine samples, and 65% with gastrointestinal samples. Expert adjudication resulted in 40% classified with ARDS, 52% classified with pneumonia, and 89% classified with sepsis. INTERPRETATION:The APS Consortium Phenotyping Study is producing a cohort of critically ill adults with ARDS, pneumonia, and sepsis with high severity of disease and a rich set of data and biospecimens. The study will continue to full enrollment of 4,000 participants. CLINICAL TRIAL REGISTRATION:ClinicalTrials.gov; No.: NCT06521502; URL: www. CLINICALTRIALS:gov.

Background Pulse oximeter performance may vary by skin pigmentation, but most data are retrospective with key limitations. Research Question Does pulse oximeter bias in critically-ill adults differ by skin pigmentation? Study Design and Methods In this prospective single-center study of 631 ICU patients (2022–2024) we directly observed simultaneous pulse oximeter oxygen saturation (SpO2) and arterial blood functional oxygen saturation (SaO2). Skin pigment was assessed using the subjective Monk Skin Tone Scale and objective spectrophotometry measurement (Individual Typology Angle [ITA]). We quantified pulse oximeter bias [mean difference between SpO2 and SaO2, and average root mean square error (ARMS) and estimated adjusted effects of skin pigment on bias or ARMS with targeted maximum likelihood estimation. Results Among 1,760 paired measurements, median SaO2 was 98% (IQR 96%, 99%) with 40 episodes of stable hypoxemia (SpO2<90%). SpO2 systematically underestimated SaO2 [median bias= -1.70 IQR (-2.84, -0.50)]. Bias was less in patients with darker skin (ITA <-30°) [-1.05 (-2.44, -0.10)] vs lighter skin (ITA >30°) [-2.01 (-3.34, -1.00)] and remained significantly different in adjusted analyses. ARMS was 3.87 (95% CI 3.25, 4.53) overall and 4.49 (95% CI 2.63, 7.07) in patients with darker skin. Interpretation In this large ICU cohort, hypoxemia was rare, FDA-cleared pulse oximeter SpO2 systematically underestimated SaO2, and performance varied by skin pigment. Bias was less in patients with darker skin. Pulse oximeter inaccuracies in ICU patients may be more substantial than clinicians recognize. Additional real-world studies with multiple oximeter brands and more stable hypoxemia are needed to further understand bias variation with skin pigment.

BACKGROUND:Some pulse oximeters perform worse in people with darker skin, and this may be due to inadequate diversity of skin pigment in device development study cohorts. Guidance is needed to accurately and equitably characterize skin pigment to ensure diversity in research cohorts. We tested multiple methods for characterizing skin pigment to assess comparability and impact on cohort diversity. OBJECTIVES:The objectives of this study were to assess reliability and comparability of common skin pigment measurement methods, compare findings from different anatomical sites and demonstrate that pigment cannot be assumed from US National Institutes for Health (NIH) race categories. METHODS:We used three subjective methods [perceived Fitzpatrick (pFP) scale, Monk Skin Tone (MST) scale and Von Luschan (VL) scale] and two objective methods [Konica Minolta CM-700d spectrophotometer and Delfin Skin Color Catch (DSCC) colorimeter] for individual typology angle (ITA) across multiple measurement sites in adults. We calculated ΔE to estimate operator perceptibility thresholds for subjective methods and to determine reproducibility for objective methods. We used each method to categorize participants as 'light, medium or dark' and compared the impact of method selection on cohort diversity. RESULTS:We studied 789 participants, with 33 856 assessments. The MST had the widest luminosity range, and the VL scale had the least discernible adjacent categories. With 'dark' defined as ITA < -30°, 14% of participants were categorized 'dark' as compared with 26% by pFP or 16% by MST. Approximately half of the 'dark' cohort had an ITA < -50°. With an ITA threshold < -50°, only 7% of the cohort was categorized as 'dark'. When 'Black or African American' self-identification was used to define 'dark', 23% of the cohort was categorized as such. Each self-assigned NIH race category included a wide range of ITA and subjective scale categories. Both ITA and L* from the KM-700d and DSCC demonstrated strong correlation (ρ > 0.7). CONCLUSIONS:Common methods for skin pigment characterization, especially the use of race or subjective scales, have significant limitations. When applied to the same cohort, different methods yield significantly different results and some may overestimate diversity. Previously published ITA thresholds for defining 'dark' skin are too light and lead to under-representation of people with darker skin.

The OpenOximetry Dataset stores clinical and lab pulse oximetry data. It supports measurements of arterial oxygen saturation (SaO2) by arterial blood gas co-oximetry and pulse oximetry (SpO2), alongside processed and unprocessed photoplethysmography (PPG) data and other metadata. This includes skin color measurements, finger diameter, vital signs (e.g., arterial blood pressure, end-tidal carbon dioxide), and arterial blood gas parameters (e.g., acid-base balance, hemoglobin concentration). All data, from desaturation studies to clinical trials, are collected prospectively to ensure accuracy. A common data model and standardized protocols for consistent archival and interpretation ensure consistent data archival and interpretation. The dataset aims to facilitate research on pulse oximeter performance across diverse human characteristics, addressing performance issues and promoting accurate pulse oximeters. The initial release includes controlled lab desaturation studies (CLDS), with ongoing updates planned as further data from clinical trials and CLDS become available.

Introduction: Several prior studies have suggested that baseline SpO2 in healthy adults may differ between females and males. However, most research has been limited by small sample sizes or the absence of concurrent arterial blood gas (ABG) measurements. Our study aims to assess potential sex-based differences in SpO2 and SaO2 in a controlled laboratory environment with simultaneous SpO2 and SaO2 collection. Methods: We analyzed prospectively collected data from healthy, non-smoking, non-pregnant adults enrolled in controlled desaturation studies at the UCSF Hypoxia Lab. Variables included age, body mass index (BMI), body surface area (BSA), heart rate (HR), finger diameter, SpO2 and pulsatility amplitude measured using the Nellcor PM1000N oximeter with the DS-100A1 probe. Additional measurements included total hemoglobin (tHb), carboxyhemoglobin (COHb), methemoglobin (MetHb), partial pressure of oxygen (PaO2), partial pressure of carbon dioxide (PaCO2), and arterial oxygen saturation (SaO2), obtained via ABG analysis using the Radiometer ABL90. All values were recorded while participants breathed room air. Skin color was assessed with individual typology angle (ITA) measured by the Konica Minolta Cm-700d spectrophotometer. Statistical analyses included the Mann-Whitney U test and Fisher's exact test for subject-level covariates, univariate generalized linear mixed effects and generalized estimating equations for measurement-level covariates. Results: We enrolled 90 females (median age: 25.7) and 77 males (median age: 26.6). Of these participants, 91 participants (46 females and 45 males) made repeat visits, resulting in a total of 502 paired SpO2-SaO2 samples (277 from females and 225 from males). As shown in Table 1, median SpO2 and SaO2 were not statistically different between females (99.0% and 98.5%) and males (98.0% and 98.1%) (p = 0.08 and p = 0.21). Females had significantly lower height, BMI, BSA, percent modulation, finger diameter, PaCO2, tHb, and had an ITA distribution with a larger proportion of lightly pigmented individuals than did the male cohort (p < 0.05). Conclusions: We did not detect a difference in SpO2 or SaO2 between healthy females and males. However, we observed several differences in our female cohort, including smaller stature, lower pulsatility amplitude, BMI, finger size, skin color, PaCO2, and tHb. While some of these factors could potentially contribute to previously observed differences in SpO2, further investigations are needed with larger sample sizes and more balanced variables to clarify how these factors may affect SpO2 and SaO2 and to determine if there is a sex-linked difference.

Introduction: Shock is associated with mortality in patients admitted to the ICU with sepsis. Two latent class analysis derived subphenotypes are associated with differential prevalence of vasopressor use on ICU admission. The duration and intensity of vasopressor support after admission in each inflammatory subphenotype is poorly characterized. Whether vasopressor dose on admission is associated with differential mortality in each subphenotype is unknown. Methods: The cohort consisted of critically ill patients with sepsis prospectively enrolled in a large observational study (EARLI, n = 764). A parsimonious classifier model using interleukin-8, bicarbonate and soluble tumor necrosis factor receptor-1 assayed on day 1 of ICU admission assigned patients to subphenotype. The daily number of vasopressor agents used, and dose were captured. Maximum cumulative daily vasopressor dose was expressed in norepinephrine equivalents (mcg/kg/min). Outcomes were 60-day in-hospital mortality and vasopressor discontinuation by subphenotype. The relationship between admission vasopressor dose and mortality in each subphenotype was assessed using logistic regression and adjusted for age, cirrhosis, invasive mechanical ventilation and cardiac comorbidities. To evaluate the probability of vasopressor discontinuation by inflammatory subphenotype, we performed competing risk analysis using a cumulative incidence function treating death as a competing risk. Results: The hyperinflammatory subphenotype had higher prevalence of vasopressor use (54.6% vs 35.9%, p < 0.001) and maximal vasopressor dose on admission (median 0.36 vs 0.13 mcg/kg/min, p < 0.001), compared to the hypoinflammatory. The hyperinflammatory subphenotype had a greater daily proportion of patients on 1 and more vasopressors compared to the hypoinflammatory, and higher daily cumulative mortality (Figure panel A). Among patients on vasopressors on admission, the hyperinflammatory subphenotype received significantly higher daily doses compared to the hypoinflammatory (Figure panel B). 18.4% of the hyperinflammatory subphenotype patients on vasopressors died in the first 2 days after admission, compared to 2.5% in the hypoinflammatory. The hyperinflammatory subphenotype was associated with a lower adjusted hazard of vasopressor discontinuation compared to the hypoinflammatory (HR 0.57, 95% CI 0.45-0.72, Figure Panel C). Vasopressor use on admission and norepinephrine equivalents were associated with increased adjusted odds of mortality in the hypoinflammatory subphenotype while the relationship was attenuated in the hyperinflammatory (Figure panel D). Conclusion: In patients with sepsis, vasopressor intensity and duration were higher in the hyperinflammatory subphenotype compared to the hypoinflammatory. Admission vasopressor dose was significantly associated with mortality in the hypoinflammatory subphenotype.

Background: Functional arterial hemoglobin oxygen saturation is a critical clinical parameter and the reference for verifying accuracy of SpO2 values during performance testing of a pulse oximeter. Variable performance between blood gas analyzers that perform co-oximetry may complicate interpretation of research studies and regulatory guidance. This study quantifies systematic differences between multiple same model analyzers from two widely used brands of co-oximeters. Methods: Three sequential experiments were conducted using controlled hypoxemia in healthy adult volunteers according to regulatory clearance criteria for use of functional arterial oxygenation from co-oximetry analysis (sO2), to simultaneous SpO2 values over the range of 70-100%. To account for potential methodological confounders in experiments 1 and 2, a third experiment analyzed 31 arterial samples across multiple same model analyzers for co-oximetry (two Werfen GEM Premier 5000, and three Radiometer ABL90 Flex) in a randomized order after simultaneous preparation. Statistical analysis included paired differences, mixed-effects modeling, and variance analysis to isolate analyzer-specific effects while controlling for confounding variables. Results: The GEM analyzers consistently measured sO2 2.3% higher on average than ABL analyzers (95% CI 2.01-2.63%, P < .001). This difference was more pronounced at lower saturations (3.2% in the sO2 70s range vs 1.4% in the 90s range). Despite strong correlation between measurements (r = 0.999), 96.8% of paired samples differed by >1% and 61.3% by >2%. Within-manufacturer variability was similar between brands (mean absolute difference: ABL 0.38%, GEM 0.47%, P = .55). Conclusions: Clinically important systematic differences in sO2 measurement exist between GEM and ABL blood analyzers that perform co-oximetry, with potential implications for patient care and pulse oximeter verification studies. Larger studies with more co-oximeter brands, coupled with blood tonometry, are needed to better characterize these findings. In the meantime, regulatory standards and reports of pulse oximeter clinical trials should account for these findings and specify which manufacturer's co-oximeters were used.

Pulse oximeter performance may vary by skin pigmentation, but most data are retrospective with key limitations. To quantify pulse oximeter bias [mean difference between pulse oximeter oxygen saturation (SpO 2 ) and arterial blood functional oxygen saturation (SaO 2 )], and average root mean square error (A RMS ) and estimate adjusted effects of skin pigment on bias or A RMS in critically-ill adults Prospective single-center study of 631 ICU patients (2022–2024) directly observed SpO and SaO pairs. Skin pigment was assessed using the subjective Monk Skin Tone Scale and objective spectrophotometry measurement (Individual Typology Angle [ITA]). Adjusted effects were estimated with targeted maximum likelihood estimation. Among 1,760 paired measurements from 631 critically-ill adults, median SaO 2 was 98% (IQR 96%, 99%) with 40 episodes of stable hypoxemia (SpO 2 <90%). SpO 2 systematically underestimated SaO 2 [median bias= −1.70 IQR (−2.84, −0.50)]. Bias was less negative in patients with darker skin (ITA <-30°) [−1.05 (−2.44, −0.10)] vs lighter skin (ITA >30°) [−2.01 (−3.34, −1.00)] and remained significantly different in adjusted analyses. A RMS was 3.87 (95% CI 3.25, 4.53) overall and 4.49 (95% CI 2.63, 7.07) in patients with darker skin. Ear probes performed worse than finger probes [bias: −2.20 vs. −1.60; A RMS : 4.90 vs. 2.70]. In this large ICU cohort, hypoxemia was rare, pulse oximeters systematically underestimated SaO, performance varied by skin pigment and probe site. Bias was less negative in patients with darker skin. Pulse oximeter inaccuracies in ICU patients may be more substantial than clinicians recognize. Does pulse oximeter bias in critically-ill adults differ by skin pigmentation? In this prospective, single-center study of 631 critically-ill adults, oximeter mean bias was negative for all patients but relatively less negative for patients with darker pigment. Bias variation by pigment was nonlinear and larger in ear than finger probes. Pulse oximeter bias varies with skin pigmentation and may not always be positive in patients with darker pigmentation. Pulse oximeter inaccuracy may be larger than clinicians appreciate. Additional studies with multiple oximeter brands and more stable hypoxemia are needed to further understand bias variation with skin pigment.

Delirium is a neurologic syndrome characterized by inattention and cognitive impairment frequently encountered in the medically ill. Peripheral inflammation is a key trigger of delirium, but the patient-specific immune responses associated with delirium development and resolution are unknown. This retrospective cohort study of prospectively collected biospecimens examines RNA sequencing from peripheral blood mononuclear cells of adults hospitalized for COVID-19 to better understand patient-specific factors associated with delirium (n = 64). Longitudinal transcriptomic analyses highlight persistent immune dysregulation in delirium, marked by increasing expression trajectories of genes linked to innate immune pathways, including complement activation, cytokine production, and monocyte/macrophage recruitment. Genes involved adaptive immunity showed a declining trajectory over time in patients with delirium. Although corticosteroid treatment suppressed some aspects of immune hyperactivation, aberrant responses contributing to delirium were exacerbated. Delirium resolution was characterized by normalization of key transcripts such as CCL2 and innate immune markers. Novel associations with delirium were found in genes related to stress granule assembly and DUSP2 and KLF10 , which mediate T-cell responses. These findings provide insights into the peripheral immune responses accompanying delirium and their modulation by corticosteroids. Future trials targeting aberrant inflammatory responses may mitigate the severe outcomes associated with delirium due to COVID19.

BACKGROUNDCritically ill patients with acute respiratory distress syndrome (ARDS) and sepsis exhibit distinct inflammatory phenotypes with divergent clinical outcomes, but the underlying molecular mechanisms remain poorly understood. These phenotypes, derived from clinical data and protein biomarkers, were associated with metabolic differences in a pilot study.METHODSWe performed integrative multiomics analysis of blood samples from 160 patients with ARDS in the ROSE trial, randomly selecting 80 patients from each latent class analysis-defined inflammatory phenotype (hyperinflammatory and hypoinflammatory) with phenotype probability greater than 0.9. Untargeted plasma metabolomics and whole-blood transcriptomics at day 0 and day 2 were analyzed using multimodal factor analysis (MEFISTO). The primary outcome was 90-day mortality, with validation in an independent critically ill sepsis cohort (EARLI).RESULTSMultiomics integration revealed 4 molecular signatures associated with mortality: (a) enhanced innate immune activation coupled with increased glycolysis (associated with hyperinflammatory phenotype), (b) hepatic dysfunction and immune dysfunction paired with impaired fatty acid β-oxidation (associated with hyperinflammatory phenotype), (c) interferon program suppression coupled with altered mitochondrial respiration (associated with hyperinflammatory phenotype), and (d) redox impairment and cell proliferation pathways (not associated with inflammatory phenotype). These signatures persisted through day 2 of trial enrollment. Within-phenotype analysis revealed distinct mortality-associated pathways in each group. All molecular signatures were validated in the independent EARLI cohort.CONCLUSIONInflammatory phenotypes of ARDS reflect distinct underlying biological processes with both phenotype-specific and phenotype-independent pathways influencing patient outcomes, all characterized by mitochondrial dysfunction. These findings suggest potential therapeutic targets for precise treatment strategies in critical illness.FUNDINGNIH National Heart, Lung, and Blood Institute and National Institute of General Medical Sciences.

Rationale: Living in a disadvantaged neighborhood is associated with adverse health outcomes, including sepsis mortality. The Area Deprivation Index (ADI) and the Healthy Places Index (HPI) are composite, area-based, metrics of neighborhood health consisting of demographic, socioeconomic and housing variables. Since 2023, Medicare has used the ADI score to adjust payments to selected Accountable Care Organizations. Recent work in New York suggests the ADI may overemphasize home value, limiting generalizability in regions with higher home prices. This study was designed to evaluate the association between neighborhood disadvantage and mortality among ICU patients with sepsis in the San Francisco Bay Area and compare the performance of the ADI and HPI. Methods: The cohort consisted of critically ill patients with sepsis admitted through the emergency department of two San Francisco hospitals who were enrolled in a prospective observational study (EARLI) from 2008-2023. Diagnosis of sepsis was determined by two-physician adjudication. Patient addresses were geocoded using ArcGIS Pro and matched to the ADI (via census block group) and HPI (via census tract). Neighborhood disadvantage is a higher score on the ADI (1-100) and a lower score on the HPI (-1.9 to 1.7). Patients were categorized in the most-disadvantaged group within our sample for either index: by ADI (highest 15th percentile) or by HPI (lowest 15th percentile). Association between each metric and 60-day in-hospital mortality was assessed through regression analysis with hospital discharge as a competing risk, adjusting for age, sex and comorbidities associated with sepsis mortality. Results: All 896 patients were from California, and 79% from San Francisco. The cohort had a mean age of 67 (SD±17) and was 44% female. The ADI distribution had a prominent right skew with a median ADI score of 3 (IQR 2-5, range 1-75) and HPI score of 0.61 (IQR 0.24-0.97, range -1.18-1.69). 56% of patients designated as most-disadvantaged by HPI were not classified as such by the ADI. There were 259 in-hospital deaths by 60 days (29%). Living in the most disadvantaged neighborhoods by HPI, but not ADI, was associated with 60-day hospital mortality in an adjusted regression model (Figure 1). Conclusion: Among ICU patients with sepsis predominantly from San Francisco, neighborhood deprivation by the HPI was associated with increased adjusted incidence of mortality, while no association was found with the ADI. Findings suggest the HPI may capture neighborhood health more accurately in this population and potentially other communities with similarly high median home prices.

BACKGROUNDAccurate prognostic assays for COVID-19 represent an unmet clinical need. We sought to identify and validate early parsimonious transcriptomic signatures that accurately predict fatal outcomes.METHODSWe studied 894 patients enrolled in the prospective, multicenter Immunophenotyping Assessment in a COVID-19 Cohort (IMPACC) with peripheral blood mononuclear cells (PBMC) and nasal swabs collected within 48 hours of admission. Host gene expression was measured with RNA-Seq. We trained parsimonious prognostic classifiers incorporating host gene expression, age, and SARS-CoV-2 viral load to predict 28-day mortality in 70% of the cohort. Classifier performance was determined in the remaining 30% and externally validated in a contemporary COVID-19 cohort (n = 137) with vaccinated patients.RESULTSFatal COVID-19 was characterized by 4,189 differentially expressed genes in the peripheral blood. A COVID-specific 3-gene peripheral blood classifier (CD83, ATP1B2, DAAM2) combined with age and SARS-CoV-2 viral load achieved an area under the receiver operating characteristic curve (AUC) of 0.88 (95% CI, 0.82-0.94). A 3-gene nasal classifier (SLC5A5, CD200R1, FCER1A), in comparison, yielded an AUC of 0.74 (95% CI, 0.64-0.83). Notably, OLAH, the most strongly upregulated gene in both PBMC and nasal swab and recently implicated in severe viral infection pathogenesis, yielded AUCs of 0.86 (0.79-0.93) and 0.78 (95% CI, 0.69-0.86), respectively. Both peripheral blood classifiers demonstrated comparable performance in an independent contemporary cohort of vaccinated patients (AUCs 0.74-0.80).CONCLUSIONOur parsimonious blood- and nasal-based classifiers accurately predicted COVID-19 mortality and merit further study as accessible prognostic tools to guide triage, resource allocation, and early therapeutic interventions.FUNDINGNIH: 5R01AI135803-03, R35HL140026, 5U19AI118608-04, 5U19AI128910-04, 4U19AI090023-11, 4U19AI118610-06, R01AI145835-01A1S1, 5U19AI062629-17, 5U19AI057229-17, 5U19AI125357-05, 5U19AI128913-03, 3U19AI077439-13, 5U54AI142766-03, 5R01AI104870-07, 3U19AI089992-09, 3U19AI128913-03, 5T32DA018926-18, and K0826161611. National Institute of Allergy and Infectious Diseases, NIH: 3U19AI1289130, U19AI128913-04S1, and R01AI122220. National Center for Advancing Translational Sciences, NIH: UM1TR004528. The National Science Foundation: DMS2310836. The Chan Zuckerberg Biohub San Francisco.

Rationale: Two molecular phenotypes of the acute respiratory distress syndrome (ARDS) with divergent clinical trajectories and responses to therapy have been identified. Classification as "hyperinflammatory" or "hypoinflammatory" depends on plasma biomarker profiling. Limited data are available about the differences in the pulmonary biology of the molecular phenotypes. Objectives: To identify differences in the pulmonary biology of ARDS molecular phenotypes Methods: We compared tracheal aspirate gene expression between hyperinflammatory and hypoinflammatory phenotypes in bulk RNA sequencing (RNASeq) from coronavirus disease (COVID-19) and non-COVID-19 ARDS and single-cell RNASeq from non-COVID-19 ARDS. In a subset of subjects, we also compared plasma proteomic data. Measurements and Main Results: In bulk RNASeq analyses, 1,157 genes were differentially expressed (false discovery rate < 0.1) between phenotypes in non-COVID-19 ARDS, and 85 genes were differentially expressed between phenotypes in COVID-19 ARDS. Eighteen genes were reproducibly differentially expressed between phenotypes in both cohorts, including greater expression of IL32, HSPA8, and PPP3CC in hyperinflammatory ARDS. A total of 195 pathways were reproducibly enriched across the two cohorts by gene set enrichment analysis, including greater expression of granulopoiesis, T-cell and IFN signaling, and integrated stress response pathways in hyperinflammatory ARDS. Network analysis of single-cell RNASeq in a third group of patients identified greater T-cell signaling to other immune cells in hyperinflammatory ARDS. Conclusions: Hyperinflammatory and hypoinflammatory ARDS molecular phenotypes have distinct respiratory biology. Hyperinflammatory ARDS is characterized by an increased IFN-stimulated gene expression and T-cell activation in the lungs.

Myocardial injury is common in acute respiratory distress syndrome (ARDS) and sepsis and associated with increased mortality. Two latent class analysis derived subphenotypes are associated with differential risk of mortality in these populations, though the association of troponin-I with mortality within each subphenotype is unknown. The derivation (n = 597 in EARLI) and validation (n = 452 in VALID) cohorts consisted of patients with sepsis or ARDS admitted to the ICU and enrolled in two separate prospective observational studies. Patients with troponin-I measured between hospital presentation and within 24 h of ICU admission were included. A parsimonious classifier model using interleukin-8, soluble tumor necrosis factor receptor-1, and vasopressor use assigned patients to subphenotype. Association between peak troponin-I concentration and 60-day in-hospital mortality within each subphenotype was assessed through logistic regression adjusting for age, admission laboratory values, vasopressor use, invasive ventilation use, and cardiac comorbidities. Median peak troponin-I was significantly higher in the hyperinflammatory vs hypoinflammatory subphenotype in both cohorts (0.07 vs 0.04 ng/mL and 0.17 vs 0.07 ng/mL, both p < 0.05). The association between peak troponin-I and mortality differed between inflammatory subphenotypes (p-interaction 0.004, EARLI). In EARLI, each doubling of peak troponin-I was associated with increased adjusted odds of 60-day mortality (aOR 1.14, 95

OBJECTIVES:Hyperinflammatory and hypoinflammatory molecular subphenotypes in sepsis and acute respiratory distress syndrome have divergent mortality and treatment responses in secondary analyses of randomized controlled trials. However, the prevalence of immunocompromise is low in these populations, and how preexisting immunocompromise contributes to subphenotypes is unknown. We studied two observational sepsis cohorts to test associations between immunocompromise and the hyperinflammatory subphenotype and to assess whether the prognostic relevance of molecular subphenotypes is generalizable to immunocompromised populations. DESIGN:Observational cohort study. SETTING:Prospective data from two ICU cohorts in the United States. PATIENTS:We included 1826 patients from two combined sepsis cohorts. INTERVENTIONS:None. MEASUREMENTS AND MAIN RESULTS:We defined immunocompromise as a history of solid organ transplant, AIDS, hematologic malignancy, solid malignancy on chemotherapy, or immunosuppressive medication use. Subphenotype was previously assigned using latent class analysis. We used logistic regression to investigate associations between type of immunocompromise and hyperinflammatory subphenotype. Models were repeated with individual covariates known or hypothesized to be associated with the hyperinflammatory subphenotype. Kaplan-Meier survival plots were used to assess mortality differences by subphenotype. Hematologic malignancy was strongly associated with the hyperinflammatory subphenotype (odds ratio [OR], 4.3; p < 0.0001), an association that persisted after adjustment for identified pathogen, presence of bacteremia, or illness severity. History of solid organ transplantation was also associated with the hyperinflammatory subphenotype (OR, 1.6; p = 0.02) but was no longer significant after accounting for bacteremia. Hyperinflammatory classification was associated with a decreased likelihood of survival in hematologic malignancy, but not in organ transplant or solid malignancy populations. CONCLUSIONS:Preexisting immune status is associated with subphenotype assignment and may influence its prognostic utility.

Rationale: Two molecular subphenotypes, (hyperinflammatory and hypoinflammatory) have been described in sepsis with divergent mortality in secondary analyses of randomized controlled trials (RCTs). Importantly, prevalence of immunocompromise is low in RCTs, and how pre-existing immunocompromise contributes to molecular subphenotypes is unknown. We designed a study to investigate the association between pre-existing immunocompromise and sepsis molecular subphenotypes.Methods: We combined data from two prospective observational cohorts of patients enrolled within 24 hours of admission to the ICU at UCSF and Vanderbilt University from 2008-2018 and 2007-2019 respectively. Included patients met clinical criteria for sepsis on enrollment, as determined by trained physicians, and had a sepsis subphenotype previously assigned through latent class analysis (Sinha, LRM 2023). Type of immunocompromise was defined as history of 1) solid organ transplant 2) AIDS (HIV, CD4<200) 3) hematologic malignancy 4) solid malignancy on chemotherapy or 5) immunosuppressive medication use. We used logistic regression to investigate associations between hyperinflammatory subphenotype and type of immunocompromise with original cohort included as a covariate to adjust for cohort-level effects. We repeated individual regression models to include covariates known or hypothesized to be associated with the hyperinflammatory subphenotype, including bacteremia, pathogen type, severity of illness and presence of leukopenia.Results: Of the 1,868 patients who met inclusion criteria, 605 (32%) patients were immunocompromised. Comorbidities were similar between groups, with exception of more cirrhosis in the immune competent group (p = 0.01). Immunocompromised patients had higher APACHE II scores (p = 0.01) and mortality at 30 days (p = 0.004). Both a history of hematologic malignancy (OR 4.3, 95% CI 3.1-6.0, p <0.0001) and a history of solid organ transplant (OR 1.6, 95% CI 1.1-2.4 p = 0.01) were associated with the hyperinflammatory subphenotype (Figure 1A). There was no association between subphenotype and AIDS, solid malignancy or immunosuppressive medication use. In multivariate analysis to identify potential mediators, hematologic malignancy remained a strong independent predictor of hyperinflammatory subphenotype (Figure 1B). In contrast, when accounting for bacteremia and leukopenia the association between solid organ transplantation and hyperinflammatory subphenotype was no longer significant (Figure 1C).Conclusions: Among septic patients admitted to the ICU, a history of hematologic malignancy and solid organ transplantation are associated with the hyperinflammatory subphenotype. Perhaps surprisingly, hematologic malignancy patients were more likely to be classified in the hyperinflammatory subphenotype, independent of causative pathogen, presence of bacteremia or severity of illness. Our results indicate that pre-existing immune status is a key determinant of sepsis subphenotypes.

Predicting mortality risk in patients with COVID-19 remains challenging, and accurate prognostic assays represent a persistent unmet clinical need. We aimed to identify and validate parsimonious transcriptomic signatures that accurately predict fatal outcomes within 48 hours of hospitalization. We studied 894 patients hospitalized for COVID-19 across 20 US hospitals and enrolled in the prospective Immunophenotyping Assessment in a COVID-19 Cohort (IMPACC) with peripheral blood mononuclear cells (PBMC) and nasal swabs collected within 48 hours of admission. Host gene expression was assessed by RNA sequencing, nasal SARS-CoV-2 viral load was measured by RT-qPCR, and mortality was assessed at 28 days. We first defined transcriptional signatures and biological features of fatal COVID-19, which we compared against mortality signatures from an independent cohort of patients with non-COVID-19 sepsis (n=122). Using least absolute shrinkage and selection operator (LASSO) regression in 70% of the COVID-19 cohort, we trained parsimonious prognostic classifiers incorporating host gene expression, age, and viral load. The performance of single and three-gene classifiers was then determined in the remaining 30% of the cohort and subsequently externally validated in an independent, contemporary COVID-19 cohort (n=137) with vaccinated patients. Fatal COVID-19 was characterized by 4189 differentially expressed genes in the peripheral blood, representing marked upregulation of neutrophil degranulation, erythrocyte gas exchange, and heme biosynthesis pathways, juxtaposed against downregulation of adaptive immune pathways. Only 7.6% of mortality-associated genes overlapped between COVID-19 and sepsis due to other causes. A COVID-specific three-gene peripheral blood classifier ( CD83, ATP1B2, DAAM2 ) combined with age and SARS-CoV-2 viral load achieved an area under the receiver operating characteristic curve (AUC) of 0.88 (95% CI 0.82–0.94). A three-gene nasal classifier ( SLC5A5, CD200R1 , FCER1A ), in comparison, yielded an AUC of 0.74 (95% CI 0.64-0.83). Notably the expression of OLAH alone, a gene recently implicated in severe viral infection pathogenesis, yielded an AUC of 0.86 (0.79–0.93). Both peripheral blood classifiers demonstrated comparable performance in vaccinated patients from an independent external validation cohort (AUCs 0.74– 0.80). A three-gene peripheral blood signature, as well as OLAH alone, accurately predict COVID-19 mortality early in hospitalization, including in vaccinated patients. These parsimonious blood- and nasal-based classifiers merit further study as accessible prognostic tools to guide triage, resource allocation, and early therapeutic interventions in COVID-19.

Delirium is a neurologic syndrome characterized by inattention and cognitive impairment frequently encountered in medically ill older adults. As a hallmark of age-related brain vulnerability, delirium offers a clinical model to investigate how peripheral immune responses contribute to acute brain dysfunction. Peripheral inflammation is a key trigger of delirium, but the patient-specific immune responses that drive delirium onset and recovery remain poorly understood. This retrospective cohort study of prospectively collected biospecimens examines RNA sequencing from peripheral blood mononuclear cells of adults hospitalized for COVID-19 to better understand patient-specific factors associated with delirium (n = 64). Longitudinal transcriptomic analyses highlight persistent immune dysregulation in delirium, marked by increasing expression trajectories of genes linked to innate immune pathways, including complement activation, cytokine production, and monocyte/macrophage recruitment. Genes involved adaptive immunity showed a declining trajectory over time in patients with delirium. Although corticosteroid treatment suppressed some aspects of immune hyperactivation, aberrant responses contributing to delirium were exacerbated. Delirium resolution was characterized by normalization of key transcripts such as CCL2 and innate immune markers. Novel associations with delirium included transcripts related to stress granule assembly and the T cell regulators DUSP2 and KLF10. Delirium in COVID-19 is associated with distinct and dynamic peripheral immune trajectories that are modulated by corticosteroids. Further understanding these mechanisms has important implications for preventing delirium in older adults. These findings provide novel mechanistic insights with translational relevance for immunomodulatory strategies targeting maladaptive immune responses to prevent or treat delirium in medically ill populations.