Female patients with sepsis have better clinical outcomes than male patients in terms of mortality and length of hospitalization and ICU stay.This is an open access article distributed under the terms of the Creative Commons Attribution-Non Commercial-No Derivatives License 4.0 (CCBY-NC-ND), where it is permissible to download and share the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal. http://creativecommons.org/licenses/by-nc-nd/4.0.

Objectives: We investigated the impact of obesity (proxied as body mass index (BMI)), on short-and long-term mortality in sepsis patients. Methods: We conducted a retrospective analysis with adult sepsis ICU patients in a US medical institution from 2001 to 2012 in the MIMIC-III database. The WHO BMI categories were used. Multivariate logistic regression assessed the relationships between BMI and 30-day and 1-year mortality. Results: In total, 5563 patients were enrolled. Obese patients tended to be younger (P < 0.001), to be female (P < 0.001), to acquire worse SOFA scores (P < 0.001), and to receive more aggressive treatment compared with their normal weight counterparts. Obese patients had notably longer mechanical ventilation periods and ICU and hospital lengths of stay (LOSs). In the final model, overweight and obese patients had lower 30-day (OR 0.77, 95% CI 0.66-0.91; OR 0.65, 95% CI 0.56-0.77, respectively) and 1-year (OR 0.83, 95% CI 0.71-0.96; OR 0.70, 95% CI 0.60-0.81, respectively) mortality risks than normal weight patients. In contrast, underweight patients had worse 30-day and 1-year outcomes compared with normal weight patients (P = 0.01, P < 0.001, respectively). In morbidly obese, severe sepsis and septic shock patients, obesity remained protective. Conclusions: Obesity was correlated with short-and long-term survival advantages in sepsis patients.

Background: Risk stratification plays an essential role in the decision making for sepsis management, as existing approaches can hardly satisfy the need to assess this heterogeneous population. We aimed to develop and validate a machine learning model to predict in-hospital mortality in critically ill patients with sepsis. Methods: Adult patients fulfilling the definition of Sepsis-3 were included at a large tertiary medical center. Relevant clinical features were extracted within the first 24 h in ICU, re-classified into different genres, and utilized for model development under three strategies: “Basic + Lab”, “Basic + Intervention”, and “Whole” feature sets. Extreme gradient boosting (XGBoost) was compared with logistic regression (LR) and established severity scores. Temporal validation was conducted using admissions from 2017 to 2019. Results: The final cohort included 24,272 patients, of which 4013 patients formed the test cohort for temporal validation. The trained and fine-tuned XGBoost model with the whole feature set showed the best discriminatory ability in the test cohort with AUROC as 0.85, significantly higher than the XGBoost “Basic + Lab” model (0.83), the LR “Whole” model (0.82), SOFA (0.63), SAPS-II (0.73), and LODS score (0.74). The performance in varying subgroups remained robust, and predictors, such as increased urine output and supplemental oxygen therapy, were crucially correlated with improved survival when interpretability was explored. Conclusions: We developed and validated a novel XGBoost-based model and demonstrated significantly improved performance to LR and other scores in predicting the mortality risks of sepsis patients in the hospital using features in the first 24 h.

Paraphrase generation plays key roles in NLP tasks such as question answering, machine translation, and information retrieval. In this paper, we propose a novel framework for paraphrase generation. It simultaneously decodes the output sentence using a pretrained wordset-to-sequence model and a round-trip translation model. We evaluate this framework on Quora, WikiAnswers, MSCOCO and Twitter, and show its advantage over previous state-of-the-art unsupervised methods and distantly-supervised methods by significant margins on all datasets. For Quora and WikiAnswers, our framework even performs better than some strongly supervised methods with domain adaptation. Further, we show that the generated paraphrases can be used to augment the training data for machine translation to achieve substantial improvements.

Background Nutritional therapy is essential for patients in the intensive care unit (ICU), when optimal caloric goals remain controversial, especially for the late acute phase (after day 3). This study aimed to investigate the impact of hypocaloric versus normocaloric feeding in this period. Methods We conducted a retrospective analysis within an up-to-date large database, including adult patients who were admitted to a tertiary hospital from 2008 to 2019, and stayed in the ICU for over one week. Individuals were divided according to the mean daily caloric intake from day 4 to day 7 (< 20 Kcal/kg/d; >= 20 Kcal/kg/d). The restricted cubic spline for a cox proportional hazards model was employed to assess the association between mean caloric intake divided by predicted energy expenditure (EE) and 1-year mortality. Results 3,545 eligible patients formed the study population. Most patients received progressively-elevating nutrition, achieving median values of 18.4 Kcal/kg/d and 0.71 g/kg/d in caloric and protein intake from day 4 onwards. Hypocaloric feeding was correlated with reduced nosocomial infection (41.7% vs 46.7%, P = 0.003), hyperglycemia episodes (63.6% vs 67.8%, P = 0.008), ventilation durations (3.6 vs 4.1 days, P = 0.001), and ICU length of stay (LOS) (11.6 vs 13.4 days, P < 0.001) compared to normal energy provision. In-hospital and 1-year mortality risks displayed no significant changes between the two strategies (ORs [95% CIs]: 0.80 [0.61–1.04], P = 0.10; 0.81 [0.64–1.02], P = 0.08, respectively). Achieving a calorie/EE of 40 ~ 70% showed a great 1-year survival benefit when predicted equations were used. Conclusion Compared to normocaloric feeding after day 3, hypocaloric feeding significantly reduced nosocomial infection rate, hyperglycemia episodes, ventilation days, and ICU LOS, when no significant changes were observed in the hospital and 1-year mortality risks. The optimal goal might be set at 40 ~ 70% of EE when predicted equations were used to prevent overfeeding.