Background: In December 2019, coronavirus disease 2019 (COVID-19) outbreak was reported from Wuhan, China. Information on the clinical course and prognosis of COVID-19 was not thoroughly described. We described the clinical courses and prognosis in COVID-19 patients. Methods: Retrospective case series of COVID-19 patients from Zhongnan Hospital of Wuhan University in Wuhan and Xishui Hospital, Hubei Province, China, up to February 10, 2020. Epidemiological, demographic, and clinical data were collected. The clinical course of survivors and non-survivors were compared. Risk factors for death were analyzed. Results: A total of 107 discharged patients with COVID-19 were enrolled. The clinical course of COVID-19 presented as a tri-phasic pattern. Week 1 after illness onset was characterized by fever, cough, dyspnea, lymphopenia, and radiological multi-lobar pulmonary infiltrates. In severe cases, thrombocytopenia, acute kidney injury, acute myocardial injury, and adult respiratory distress syndrome were observed. During week 2, in mild cases, fever, cough, and systemic symptoms began to resolve and platelet count rose to normal range, but lymphopenia persisted. In severe cases, leukocytosis, neutrophilia, and deteriorating multi-organ dysfunction were dominant. By week 3, mild cases had clinically resolved except for lymphopenia. However, severe cases showed persistent lymphopenia, severe acute respiratory dyspnea syndrome, refractory shock, anuric acute kidney injury, coagulopathy, thrombocytopenia, and death. Older age and male sex were independent risk factors for poor outcome of the illness. Conclusions: A period of 7–13 days after illness onset is the critical stage in the COVID-19 course. Age and male gender were independent risk factors for death of COVID-19.

Introduction This study aimed to develop and validate an interpretable machine-learning model based on clinical features for early predicting in-hospital mortality in critically ill patients with sepsis. Methods We enrolled all patients with sepsis in the Medical Information Mart for Intensive Care IV (MIMIC-IV, v.1.0) database from 2008 to 2019. Lasso regression was used for feature selection. Seven machine-learning methods were applied to develop the models. The best model was selected based on its accuracy and area under curve (AUC) in the validation cohort. Furthermore, we employed the SHapley Additive exPlanations (SHAP) method to illustrate the effects of the features attributed to the model, and to analyze how the individual features affect the output of the model, and to visualize the Shapley value for a single individual. Results In total, 8,817 patients with sepsis were eligible for participation, the median age was 66.8 years (IQR, 55.9–77.1 years), and 3361 of 8817 participants (38.1%) were women. After selection, 25 of a total 57 clinical parameters collected on day 1 after ICU admission remained associated with prognosis and were used for developing the machine-learning models. Among seven constructed models, the eXtreme Gradient Boosting (XGBoost) model achieved the best performance with an AUC of 0.884 and an accuracy of 89.5% in the validation cohort. Feature importance analysis showed that Glasgow Coma Scale (GCS) score, blood urea nitrogen, respiratory rate, urine output, and age were the top 5 features of the XGBoost model with the greatest impact. Furthermore, SHAP force analysis illustrated how the constructed model visualized the individualized prediction of death. Conclusions We have demonstrated the potential of machine-learning approaches for predicting outcome early in patients with sepsis. The SHAP method could improve the interpretability of machine-learning models and help clinicians better understand the reasoning behind the outcome. Graphical abstract

Objectives: Severe acute respiratory distress syndrome is complicated with coronavirus disease 2019 and extracorporeal membrane oxygenation support may be necessary in severe cases. This study is to summarize the clinical features, extracorporeal membrane oxygenation characteristics, and outcomes of patients with severe acute respiratory syndrome coronavirus 2 pneumonia received extracorporeal membrane oxygenation. Design: Descriptive study from two hospitals. Setting: The ICUs from university hospitals. Patients: Patients with severe acute respiratory syndrome coronavirus 2 pneumonia received mechanical ventilation, including those underwent extracorporeal membrane oxygenation from Zhongnan Hospital of Wuhan University and Wuhan Pulmonary Hospital from January 8, 2020, to March 31, 2020. Interventions: None. Measurements and Main Results: Clinical records, laboratory results, ventilator parameters, and extracorporeal membrane oxygenation-related data were abstracted from the medical records. One-hundred twenty-nine critically ill patients with severe acute respiratory syndrome coronavirus 2 pneumonia were admitted to ICU of the two referral hospitals. Fifty-nine patients received mechanical ventilation and 21 of them received extracorporeal membrane oxygenation support (fourteen from Zhongnan hospital and seven from Wuhan pulmonary hospital). Compared to mechanical ventilation patients without extracorporeal membrane oxygenation support, there was a tendency of decline in mortality but with no significant difference (no-extracorporeal membrane oxygenation group 24/38 [63.2%] vs extracorporeal membrane oxygenation group 12/21 [57.1%]; p = 0.782). For those patients with extracorporeal membrane oxygenation, 12 patients died and nine survived by April 7, 2020. Among extracorporeal membrane oxygenation patients, the Paco ₂ prior to extracorporeal membrane oxygenation was lower (54.40 mm Hg [29.20–57.50 mm Hg] vs 63.20 mm Hg [55.40–72.12 mm Hg]; p = 0.006), and pH prior to extracorporeal membrane oxygenation was higher (7.38 [7.28–7.48] vs 7.23 [7.16–7.33]; p = 0.023) in survivors than nonsurvivors. Conclusions: Extracorporeal membrane oxygenation might be an effective salvage treatment for patients with severe acute respiratory syndrome coronavirus 2 pneumonia associated with severe acute respiratory distress syndrome. Severe Co ₂ retention and acidosis prior to extracorporeal membrane oxygenation indicated a poor prognosis.

Introduction Septic patients requiring intensive care unit (ICU) readmission are at high risk of mortality, but research focusing on the association of ICU readmission due to sepsis and mortality is limited. The aim of this study was to develop and validate a machine learning (ML) model for predicting in-hospital mortality in septic patients readmitted to the ICU using routinely available clinical data. Methods The data used in this study were obtained from the Medical Information Mart for Intensive Care (MIMIC-IV, v1.0) database, between 2008 and 2019. The study cohort included patients with sepsis requiring ICU readmission. The data were randomly split into a training (75%) data set and a validation (25%) data set. Nine popular ML models were developed to predict mortality in septic patients readmitted to the ICU. The model with the best accuracy and area under the curve (A.C.) in the validation cohort was defined as the optimal model and was selected for further prediction studies. The SHAPELY Additive explanations (SHAP) values and Local Interpretable Model-Agnostic Explanation (LIME) methods were used to improve the interpretability of the optimal model. Results A total of 1117 septic patients who had required ICU readmission during the study period were enrolled in the study. Of these participants, 434 (38.9%) were female, and the median (interquartile range [IQR]) age was 68.6 (58.4–79.2) years. The median (IQR) ICU interval duration was 2.60 (0.64–5.78) days. After feature selection, 31 of 47 clinical factors were ultimately chosen for use in model construction. Of the nine ML models tested, the best performance was achieved with the random forest (RF) model, with an A.C. of 0.81, an accuracy of 85% and a precision of 62% in the validation cohort. The SHAP summary analysis revealed that Glasgow Coma Scale score, urine output, blood urea nitrogen, lactate, platelet count and systolic blood pressure were the top six most important factors contributing to the RF model. Additionally, the LIME method demonstrated how the RF model works in terms of explaining risk of death prediction in septic patients requiring ICU readmission. Conclusion The ML models reported here showed a good prognostic prediction ability in septic patients requiring ICU readmission. Of the features selected, the parameters related to organ perfusion made the largest contribution to outcome prediction during ICU readmission in septic patients.

Background: The outbreak of coronavirus disease 2019 (COVID-19) has led to a large and increasing number of patients requiring prolonged mechanical ventilation and tracheostomy. The indication and optimal timing of tracheostomy in COVID-19 patients are still unclear, and the outcomes about tracheostomy have not been extensively reported. We aimed to describe the clinical characteristics and outcomes of patients with confirmed severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pneumonia who underwent elective tracheostomies.Methods: The multi-center, retrospective, observational study investigated all the COVID-19 patients who underwent elective tracheostomies in intensive care units (ICUs) of 23 hospitals in Hubei province, China, from January 8, 2020 to March 25, 2020. Demographic information, clinical characteristics, treatment, details of the tracheostomy procedure, successful weaning after tracheostomy, and living status were collected and analyzed. Data were compared between early tracheostomy patients (tracheostomy performed within 14 days of intubation) and late tracheostomy patients (tracheostomy performed after 14 days).Results: A total of 80 patients were included. The median duration from endotracheal intubation to tracheostomy was 17.5 [IQR 11.3–27.0] days. Most tracheotomies were performed by ICU physician [62 (77.5%)], and using percutaneous techniques [63 (78.8%)] at the ICU bedside [76 (95.0%)]. The most common complication was tracheostoma bleeding [14 (17.5%)], and major bleeding occurred in 4 (5.0%) patients. At 60 days after intubation, 31 (38.8%) patients experienced successful weaning from ventilator, 17 (21.2%) patients discharged from ICU, and 43 (53.8%) patients had died. Higher 60 day mortality [22 (73.3%) vs. 21 (42.0%)] were identified in patients who underwent early tracheostomy.Conclusions: In patients with SARS-CoV-2 pneumonia, tracheostomies were feasible to conduct by ICU physician at bedside with few major complications. Compared with tracheostomies conducted after 14 days of intubation, tracheostomies within 14 days were associated with an increased mortality rate.

Objectives: This observational study was conducted to investigate capillary refill time (CRT) during the early phase of ICU admission in relationship with microvascular flow alteration and outcome in critically ill patients. Design: Prospective, observational, pilot study. Setting: ICU in a university hospital. Patients: Two hundred eighty-two critically ill adult patients admitted to the ICU. Interventions: None. Measurements and Main Results: All patients underwent simultaneous measurements by CRT and sidestream dark field imaging within 24 hours of ICU admission. Other clinical data such as demographic characteristics, hemodynamics, laboratory values, treatment, and physiologic parameters were also included simultaneously. Microcirculatory measurements were performed at 10.2 ± 5.7 hours after ICU admission. Of the 282 included patients, 106 (37.6%) were female, the median (interquartile range) age was 63 years (53–74 yr), and the median Sequential Organ Failure Assessment (SOFA) score was 5 (2–7). The primary finding was the association between CRT and simultaneous the condition of peripheral circulation (microvascular flow index [MFI]: r = –0.4430, p < 0.001; proportion of perfused vessels: r = –0.3708, p < 0.001; heterogeneity index: r = 0.4378, p < 0.001; perfused vessel density: r = –0.1835, p = 0.0020; except total vessel density: p = 0.9641; and De Backer score: p = 0.5202) in critically ill patients. In addition, this relationship was also maintained in subgroups. Microcirculatory flow abnormalities, 28-day mortality, and SOFA score appeared to be more severe for increasing CRT. In a multivariable analysis, prolonged CRT was independently associated with microvascular flow abnormalities (MFI < 2.6; odds ratio [OR], 1.608; 95% CI, 2.1–10.2; p < 0.001). Similarly, multivariable analysis identified CRT as an independent predictor of 28-day mortality (OR, 1.296; 95% CI, 1.078–1.558; p = 0.006). Conclusions: In our ICU population, a single-spot prolonged CRT was independently associated with abnormal microcirculation and increased mortality.