Document resource
Background Acute gastrointestinal injury (AGI) is a frequent and deleterious complication in patients with sepsis, significantly impacting prognosis. Identifying patients at high risk is crucial for early intervention. This study aimed to analyze the risk factors for AGI development in septic patients and to construct a personalized risk prediction model.Methods A retrospective analysis was conducted using data from the MIMIC-IV database. A total of 247 septic patients with AGI were assigned to the AGI group, while 914 septic patients without AGI (NAGI) were included in the control (NAGI) group. Baseline demographics, clinical characteristics, and admission laboratory data (including Sequential Organ Failure Assessment (SOFA) score and serum lactate levels) were collected. Univariate and multivariate logistic regression analyses were performed to identify independent risk factors for AGI. These factors were then integrated to build a personalized risk prediction model. The discriminatory performance of individual factors and the composite model was assessed using the area under the receiver operating characteristic curve (AUC).Results Patients in the AGI group were significantly older and had higher admission SOFA scores and lactate levels compared to the NAGI group (P<0.05). Multivariate logistic regression identified four independent risk factors for AGI development in sepsis: advanced age, high admission SOFA score, elevated serum lactate level, and sepsis primarily caused by severe pneumonia (all P<0.05). Individually, age and severe pneumonia as the primary cause showed low to moderate predictive value (0.55 ≤ AUC ≤ 0.73, P<0.05). Admission SOFA score and lactate level demonstrated moderate predictive value (0.71 < AUC ≤ 0.86, P<0.001). Crucially, the composite risk model integrating all four factors yielded significantly higher predictive accuracy (AUC > 0.90, P<0.001).Conclusions Advanced age, elevated admission SOFA score, high serum lactate, and sepsis originating from severe pneumonia are independent risk factors for AGI in septic patients. The personalized risk prediction model constructed by combining these factors exhibits excellent predictive performance. This model provides a practical tool for clinicians to stratify sepsis patients based on their individual risk of developing AGI, thereby facilitating timely monitoring and proactive management to potentially improve outcomes.