Early prognosis prediction for non-variceal upper gastrointestinal bleeding in the intensive care unit: based on interpretable machine learning.

INTRODUCTION: This study aims to construct a mortality prediction model for patients with non-variceal upper gastrointestinal bleeding (NVUGIB) in the intensive care unit (ICU), employing advanced machine learning algorithms. The goal is to identify high-risk populations early, contributing to a deeper understanding of patients with NVUGIB in the ICU. METHODS: We extracted NVUGIB data from the Medical Information Mart for Intensive Care IV (MIMIC-IV, v.2.2) database spanning from 2008 to 2019. Feature selection was conducted through LASSO regression, followed by training models using 11 machine learning methods. The best model was chosen based on the area under the curve (AUC). Subsequently, Shapley additive explanations (SHAP) was employed to elucidate how each factor influenced the model. Finally, a case was randomly selected, and the model was utilized to predict its mortality, demonstrating the practical application of the developed model. RESULTS: In total, 2716 patients with NVUGIB were deemed eligible for participation. Following selection, 30 out of a total of 64 clinical parameters collected on day 1 after ICU admission remained associated with prognosis and were utilized for developing machine learning models. Among the 11 constructed models, the Gradient Boosting Decision Tree (GBDT) model demonstrated the best performance, achieving an AUC of 0.853 and an accuracy of 0.839 in the validation cohort. Feature importance analysis highlighted that shock, Glasgow Coma Scale (GCS), renal disease, age, albumin, and alanine aminotransferase (ALP) were the top six features of the GBDT model with the most significant impact. Furthermore, SHAP force analysis illustrated how the constructed model visualized the individualized prediction of death. CONCLUSIONS: Patient data from the MIMIC database were leveraged to develop a robust prognostic model for patients with NVUGIB in the ICU. The analysis using SHAP also assisted clinicians in gaining a deeper understanding of the disease.

Environmental and economic assessments of industry-level medical waste disposal technologies - A case study of ten Chinese megacities.

Medical waste (MW) is exploding due to the COVID-19 pandemic, posing a significant environmental threat, and leading to the urgent requirement for affordable and environmentally friendly MW disposal technologies. Prior research on individual MW disposal plants is region-specific, applying these results to other regions may introduce bias. In this study, major MW disposal technologies in China, i.e., incineration technologies (pyrolysis incineration and rotary kiln incineration), and sterilization technologies (steam sterilization, microwave sterilization, and chemical disinfection) with residue landfill or incineration were analyzed from an industry-level perspective via life cycle assessment (LCA), life cycle costing (LCC) and net present value (NPV) methods. Life cycle inventories and economic cost data for 4-5 typical companies were selected from 128 distinct enterprises and academic sources for each technology. LCA results show that microwave sterilization with residue incineration has the lowest environmental impact, emitting only 480 kg CO2 eq. LCC and NPV analyses indicate that steam sterilization with landfilling is the most economical, yielding revenues of 1,210 CNY/t and breaking even in the first year. Conversely, pyrolysis and rotary kiln incineration break even between the 4th and 5th years. Greenhouse gas emissions from the MW disposal in ten cities with the largest MW production in 2020 increased by 7% over 2019 to 43,800 tons and other pollutants increased by 6% to 12%. Economically, Shanghai exhibits the highest cost-effectiveness, while Nanjing delivers the lowest. It can be observed that the adoption of optimal environmental technologies has resulted in a diminution of greenhouse gas emissions by 279,000 tons and energy conservation of 1.76 billion MJ.