Background: Metagenomic next-generation sequencing (mNGS) is a promising technique for pathogens diagnosis. However, application of mNGS in immunocompromised adults with severe community-acquired pneumonia (SCAP) is relatively limited.Methods: We retrospectively reviewed 23 immunocompromised and 21 immunocompetent SCAP patients with mNGS detection from April 2019 to December 2019. The performances of pathogenic diagnosis and subsequently antibiotic adjustment in immunocompromised SCAP patients were compared to immunocompetent SCAP patients. The defined by days of therapy (DOT) method was used for estimate daily antibiotic use.Results: There was a significant difference in the diagnostic positivity rate between mNGS and conventional test in both groups (P<0.001). Compared to immunocompetent patients, more mixed pathogens in immunocompromised patients were found (P=0.023). Before the availability of mNGS, the DOTs in immunocompromise patients were higher than immunocompetent patients (3.0 [3.0, 4.0] vs. 3.0 [2.0, 3.0], P=0.013). Compared to immunocompetent patients, immunocompromised patients had fewer full pathogen covered empirical antibiotic therapy (14.7% vs. 57.1%, P=0.022), more adjustments of antibiotic treatment (87.0%) vs. 57.1%, P=0.027). More than a half (13 of 23) SCAP patients in immunosuppressed group had reduced or downgraded antibiotic adjustments based on the results.Conclusions: mNGS may be a useful technique for detecting mixed pathogens and personalized antibiotic treatment in immunocompromised SCAP patients.

Objectives: To evaluate the incidence and mortality of acute respiratory distress syndrome (ARDS) in medical/respiratory intensive care units (MICUs/RICUs) to assess ventilation management and the use of adjunct therapy in routine clinical practice for patients fulfilling the Berlin definition of ARDS in mainland China. Methods: This was a multicentre prospective longitudinal study. Patients who met the Berlin definition of ARDS were included. Baseline data and data on ventilator management and the use of adjunct therapy were collected. Results: Of the 18,793 patients admitted to participating ICUs during the study timeframe, 672 patients fulfilled the Berlin ARDS criteria and 527 patients were included in the analysis. The most common predisposing factor for ARDS in 402 (77.0) patients was pneumonia. The prevalence rates were 9.7% (51/527) for mild ARDS, 47.4% (250/527) for moderate ARDS, and 42.9% (226/527) for severe ARDS. In total, 400 (75.9%) patients were managed with invasive mechanical ventilation during their ICU stays. All ARDS patients received a tidal volume of 6.8 (5.8–7.9) mL/kg of their predicted body weight and a positive end-expository pressure (PEEP) of 8 (6–12) cmH₂O. Recruitment manoeuvres (RMs) and prone positioning were used in 61 (15.3%) and 85 (16.1%) ventilated patients, respectively. Life-sustaining care was withdrawn from 92 (17.5%) patients. When these patients were included in the mortality analysis, 244 (46.3%) ARDS patients (16 (31.4%) with mild ARDS, 101 (40.4%) with moderate ARDS, and 127 (56.2%) with severe ARDS) died in the hospital. Conclusions: Among the 18 ICUs in mainland China, the incidence of ARDS was low. The rates of mortality and withdrawal of life-sustaining care were high. The recommended lung protective strategy was followed with a high degree of compliance, but the implementation of adjunct treatment was lacking. These findings indicate the potential for improvement in the management of patients with ARDS in China. Trial registration: Clinicaltrials.gov NCT02975908. Registered on 29 November 2016—retrospectively registered.

Objectives: An ongoing outbreak of coronavirus disease 2019 is spreading globally. Acute hypoxemic respiratory failure is the most common complication of coronavirus disease 2019. However, the clinical effectiveness of early high-flow nasal oxygen treatment in patients with coronavirus disease 2019 with acute hypoxemic respiratory failure has not been explored. This study aimed to analyze the effectiveness of high-flow nasal oxygen treatment and to identify the variables predicting high-flow nasal oxygen treatment failure in coronavirus disease 2019 patients with acute hypoxemic respiratory failure. Design: A multicenter, retrospective cohort study. Setting: Three tertiary hospitals in Wuhan, China. Patients: Forty-three confirmed coronavirus disease 2019 adult patients with acute hypoxemic respiratory failure treated with high-flow nasal oxygen. Interventions: None. Measurements and Main Results: Mean age of the enrolled patients was 63.0 ± 9.7 years; female patients accounted for 41.9%. High-flow nasal oxygen failure (defined as upgrading respiratory support to positive pressure ventilation or death) was observed in 20 patients (46.5%), of which 13 (30.2%) required endotracheal intubation. Patients with high-flow nasal oxygen success had a higher median oxygen saturation (96.0% vs 93.0%; p < 0.001) at admission than those with high-flow nasal oxygen failure. High-flow nasal oxygen failure was more likely in patients who were older (p = 0.030) and male (p = 0.037), had a significant increase in respiratory rate and a significant decrease in the ratio of oxygen saturation/Fio ₂ to respiratory rate index within 3 days of high-flow nasal oxygen treatment. In a multivariate logistic regression analysis model, male and lower oxygen saturation at admission remained independent predictors of high-flow nasal oxygen failure. The hospital mortality rate of the cohort was 32.5%; however, the hospital mortality rate in patients with high-flow nasal oxygen failure was 65%. Conclusions: High-flow nasal oxygen may be effective for treating coronavirus disease 2019 patients with mild to moderate acute hypoxemic respiratory failure. However, high-flow nasal oxygen failure was associated with a poor prognosis. Male and lower oxygenation at admission were the two strong predictors of high-flow nasal oxygen failure.

Background: An early diagnosis model with clinical phenotype classification is key for the early identification and precise treatment of sepsis-associated acute respiratory distress syndrome (ARDS). This study aimed to: 1) build a machine learning diagnostic model for patients with sepsis-associated ARDS using easily accessible early clinical indicators, 2) conduct rapid classification of clinical phenotypes in this population, and 3) explore the differences in clinical characteristics, outcomes, and treatment responses of different phenotypes.Methods: This study is based on data from the Telehealth Intensive Care Unit (eICU) and Medical Information Mart for Intensive Care IV (MIMIC-IV). We trained and tested the early diagnostic model of sepsis-associated ARDS patients in the eICU. We used key predictive indicators to cluster sepsis-associated ARDS patients and determine the characteristics and clinical outcomes of different phenotypes, as well to explore the differences of in-hospital mortality among different the positive end-expiratory pressure (PEEP) levels in different phenotypes. These results are verified in MIMIC-IV to evaluate whether they are repeatable.Results: Among the diagnostic models constructed in 19,249 sepsis patients and 5,947 sepsis-associated ARDS patients, the AdaBoost (Decision Tree) model achieved the best performance with an area under the receiver operating characteristic curve (AUC) of 0.895, which is higher than that of the traditional Logistic Regression model (Z = −2.40,p = 0.013), and the accuracy of 70.06%, sensitivity of 78.11% and specificity of 78.74%. We simultaneously identified three sepsis-associated ARDS phenotypes. Cluster 0 (n = 3,669) had the lowest in-hospital mortality rate (6.51%) and fewer laboratory abnormalities (lower WBC (median:15.000 K/mcL), lower blood glucose (median:158.000 mg/dl), lower creatinine (median:1.200 mg/dl), lower lactic acid (median:3.000 mmol/L); p < 0.001). Cluster 1 (n = 1,554) had the highest in-hospital mortality rate (75.29%) and the most laboratory abnormalities (higher WBC (median:18.300 K/mcL), higher blood glucose (median:188.000 mg/dl), higher creatinine (median:2.300 mg/dl), higher lactic acid (median:3.900 mmol/L); p < 0.001). Cluster 2 (n = 724) had the most complex condition, with a moderate in-hospital mortality rate (29.7%) and the longest intensive care unit stay. In Clusters 0 and 1, patients with high PEEP had higher in-hospital mortality rate than those with low PEEP, but the opposite trend was seen in Cluster 2. These results were repeatable in 11,935 patients with sepsis and 2,699 patients with sepsis-associated ARDS patients in the MIMIC-IV cohort.Conclusion: A machine learning diagnostic model of sepsis-associated ARDS patients was established. Three phenotypes with different clinical features and outcomes were clustered, and these had different therapeutic responses. These findings are helpful for the early and rapid identification of sepsis-associated ARDS patients and their precise individualized treatment.