Development and evaluation of a prognostic model for patients with sepsis based on laboratory indicators

Early risk stratification in sepsis remains challenging due to disease heterogeneity. We developed a nomogram based on admission laboratory results to facilitate early triage and personalized care.

We analyzed 162 patients with sepsis admitted between January 2023 and December 2024 (48 with a poor prognosis, 114 who improved). Demographic data, clinical scores (Sequential Organ Failure Assessment, Glasgow Coma Scale), and 24‑hour admission laboratory parameters were collected. Least absolute shrinkage and selection operator (LASSO) regression was applied for feature selection, followed by nomogram construction. Model performance was assessed by receiver operating characteristic curves, calibration plots, and concordance index.

LASSO regression identified the Sequential Organ Failure Assessment score (odds ratio [OR], 1.271 [95% CI, 1.153-1.402), Glasgow Coma Scale score (OR, 0.807 [95% CI, 0.745-0.875]), serum urea nitrogen (OR, 1.097 [95% CI, 1.054-1.142]), and interleukin 6 (OR, 1.001 [95% CI, 1.000-1.001]) as independent predictors of poor prognosis in patients with sepsis. The nomogram showed concordance indices of 0.806 (95% CI, 0.716-0.896) in the training set and 0.889 (95% CI, 0.766-1.000) in the testing set. Sensitivity and specificity were 85.0% and 68.8% (training) and 87.9% and 86.7% (testing), respectively. Calibration curves demonstrated good agreement between predicted and observed outcomes.

The LASSO‑based nomogram exhibited discriminative ability for sepsis prognosis. It may serve as a practical tool for early risk stratification, supporting timely optimization of therapeutic strategies and improving patient outcomes.