Acute Kidney Injury in Critically Ill Patients

Acute kidney injury (AKI) in critically ill patients is characterized by high incidence, delayed diagnosis and treatment, and high mortality. Early identification and precision management are key to improving prognosis. Currently, in China, the population with severe AKI faces prominent challenges, including a lack of standardized, localized specialized data, insufficient early warning and subtyping capabilities, and a shortage of high-quality evidence-based guidance for clinical decision-making. These issues constrain the application of artificial intelligence (AI) technologies in the precision diagnosis and treatment of AKI. Leveraging the Critical Care Medicine Specialty Alliance, which has been approved by the Beijing Hospital Management Center and consists of 19 tertiary hospital ICUs nationwide, this project will conduct a three-year prospective, observational registry study. The investigators plan to consecutively enroll 23,600 adult critically ill patients (with an anticipated \>3,000 AKI patients). The study will systematically collect clinical characteristics, time-series monitoring data, laboratory parameters, renal ultrasound imaging, biomarkers, and omics data, while concurrently retaining biological samples, to establish the largest multi-modal specialized disease dataset and biobank for severe AKI in China. Focusing on the entire AKI continuum of "early warning - diagnosis - phenotyping - treatment - prognosis," the study aims to: ① characterize the epidemiological features and disease burden of ICU-AKI in China; ② develop an early warning system for AKI; ③ identify AKI sub-phenotypes using machine learning and establish a precision management framework; ④ develop an intelligent decision support system for renal replacement therapy; ⑤ evaluate prognosis; and ⑥ promote medical-engineering collaborative translation. Expected outcomes include 3-5 early warning/prognostic models and one intelligent decision support system, along with applications for 3-