Construction of a risk model for short-term mortality in ICU sepsis patients based on conventional indicators: A prospective cohort study

Sepsis, a life-threatening systemic inflammatory response syndrome triggered by infection, is associated with high morbidity and substantial short-term mortality. This study aimed to develop and validate a nomogram for predicting 28-day mortality risk in sepsis patients admitted to the intensive care unit (ICU). The study included 338 sepsis patients who met the inclusion criteria. Of these, 229 patients from the ICU with sepsis from January 2018 to December 2021 were assigned to the training set, and 109 ICU sepsis patients from January 2022 to December 2023 made up the validation set. Following a 28-day ICU stay, the training set was categorized into survivors (169 patients) and non-survivors (60 patients) based on their outcomes. Univariate and multivariate logistic regression analyses were conducted to identify independent risk factors for mortality in sepsis patients. A nomogram was then constructed using these risk factors to estimate the 28-day mortality risk for ICU patients diagnosed with sepsis. The nomogram's predictive accuracy was assessed using the area under the curve from the receiver operating characteristic curve, along with calibration and decision curve analysis. Logistic regression analysis indicated that sequential organ failure assessment (OR = 1.368, 95% CI = 1.175-1.773), disseminated intravascular coagulation (OR = 1.225, 95% CI = 1.195-1.924), white blood cell (OR = 1.181, 95% CI = 1.012-1.378), and interleukin-6 (OR = 1.063, 95% CI = 0.920-1.208) are independent factors influencing mortality in ICU patients with sepsis within a 28-day period. The receiver operating characteristic curve analysis revealed area under the curves of 0.913 (95% CI = 0.863-0.963) for the training set and 0.799 (95% CI = 0.669-0.929) for the validation set. Additionally, the calibration curves for both datasets closely correspond with the diagonal line, suggesting a strong fit. The decision curve analysis curve illustrates that the nomogram's training set along with the validation set offer an improved clinical net benefit. This research identified sequential organ failure assessment, disseminated intravascular coagulation, white blood cell, and interleukin-6 as independent factors affecting 28-day mortality in ICU sepsis patients and developed a predictive nomogram. This nomogram facilitates individualized prognostic assessment and offers a scientific basis for the widespread application of this risk-stratification tool in ICUs.