Evaluating machine learning models for sepsis prediction: A systematic review of methodologies.

Studies for sepsis prediction using machine learning are developing rapidly in medical science recently. In this review, we propose a set of new evaluation criteria and reporting standards to assess 21 qualified machine learning models for quality analysis based on PRISMA. Our assessment shows that (1.) the definition of sepsis is not consistent among the studies; (2.) data sources and data preprocessing methods, machine learning models, feature engineering, and inclusion types vary widely among the studies; (3.) the closer to the onset of sepsis, the higher the value of AUROC is; (4.) the improvement in AUROC is primarily due to using machine learning as a feature engineering tool; (5.) deep neural networks coupled with Sepsis-3 diagnostic criteria tend to yield better results on the time series data collected from patients with sepsis. The new evaluation criteria and reporting standards will facilitate the development of improved machine learning models for clinical applications.

Establishing a new body mass index cutoff value for malnutrition diagnosis using the Global Leadership Initiative on Malnutrition (GLIM) tool in Chinese older adults.

BACKGROUND: The body weight of the Chinese population rose rapidly over the past two decades. The old 2001 body mass index (BMI) cutoff value for malnutrition may underestimate malnutrition diagnosis. We explored the BMI cutoff value for malnutrition diagnosis based on national BMI data from the past 30 years and applied it to the Global Leadership Initiative on Malnutrition (GLIM) criteria when investigating malnutrition in hospitalized older patients. METHODS: To explore the BMI cutoff value for malnutrition, we established a linear stepwise model to predict the annual increasing BMI trend based on data from the national BMI data set. The new cutoff value was applied to a large-scale data set from a cross-sectional study pertaining to older hospitalized patients recruited from 30 large hospitals in China. RESULTS: The BMI increased from 21.8 to 23 kg/m2 in two decades. We calculated that the net BMI increase will be 1.49 kg/m2 from 1999 to 2019. We subsequently proposed that the BMI cutoff value for malnutrition should rise to 20 kg/m2 . This cutoff value was applied to the validation data set, containing 8725 patients, and the GLIM-determined malnutrition rate was 24.58% (using the Nutrition Risk Screening 2002 [NRS-2002]) and 23.32% (using the Mini Nutritional Assessment-Short Form [MNA-SF]). The results significantly differed from those obtained using the 2001 Chinese BMI criteria. CONCLUSION: The GLIM tool has good applicability in Asian populations, especially in Chinese older patients. The BMI cutoff value for malnutrition should be adjusted to 20 kg/m2 for Chinese adults.

Efficacy of intravenous vitamin C intervention for septic patients: A systematic review and meta-analysis based on randomized controlled trials.

BACKGROUND: The role of vitamin C in sepsis is still controversial, we aimed to systematically review the efficacy of intravenous vitamin C supplementation in the treatment of sepsis. METHODS: MEDLINE, EmBase, Web of Science, WanFang Data and CNKI were comprehensively searched to collect randomized controlled trails (RCTs) of vitamin C supplementation for patients with sepsis or sepsis shock from January 2000 to March 2021. Two researchers independently screened the literature, extracted the data and accessed the risk of bias in the included studies; meta-analysis was then performed by using Revman 5.4 software. RESULTS: A total of 10 RCTs involving 1400 participants were included. The results of meta-analysis showed that intravenous vitamin C supplementation can improve SOFA (ΔSOFA) within 72 h [RR = 1.32,95% CI(0.80,1.85), P < 0.0001] of septic patients. There were no difference on short term mortality (28-30d)[RR = 0.83,95% CI(0.65,1.05), P = 0.11], long term mortality (90d) [RR = 1.16,95% CI(0.82,1.66), P = 0.40], hospital LOS[RR = 0.15,95% CI(-0.73,1.03), P = 0.55], ventilator-free days[RR = 0.09,95% CI(-0.24,0.42), P = 0.60], ICU-LOS[RR = 0.22,95% CI(-0.13,0.57), P = 0.22], between two groups. The results of Subgroup analysis showed that intravenous vitamin C alone can reduce the risk of short term mortality (28-30d) [RR = 0.61,95% CI(0.47,0.79), P = 0.0002]of sepsis patients. CONCLUSION: Based on current RCTs, our work indicated that mono-intravenous vitamin C therapy may reduce short-term mortality of sepsis patients, and it may protect organ functions. Due to the limitation of the quantity and quality of included studies, the above conclusions need to be verified by more large scale and high quality randomized control trials.