Related Experiment Video
Updated: Sep 11, 2025
Establishing a Competing Risk Regression Nomogram Model for Survival Data
Published on: October 23, 2020
Development and validation of a nomogram for predicting mortality for ICU patients with severe thoracic trauma: data
Ziming Huang1, Hengfa Ge1, Ying Sun2
1Department of Trauma Center, The Affiliated Huaian No.1 People's Hospital of Nanjing Medical University, Beijing West Road 1, Huaian, Jiangsu 230032, China.
Background:
Severe thoracic trauma is a leading contributor to mortality in critically injured patients, particularly when complicated by concomitant severe traumatic brain injury (TBI), which may independently impair neurological and respiratory function. Accurate assessment and timely intervention play a crucial role in these patients. However, risk factors for severe thoracic trauma remain unclear, and a prediction rule remains to be established. We developed and internally validated a nomogram that allows clinicians to quantify the risk of severe thoracic trauma.
Methods:
Clinical data from the MIMIC-IV database were retrospectively searched to identify a study cohort comprising patients with severe thoracic trauma. Using LASSO regression analysis, We screened out independent risk factors associated with 28-day mortality and incorporated them into nomogram model. The performance of each model was assessed by calculating receiver operating characteristic (ROC) curves, calibration plots and decision curve analysis (DCA).
Results:
The final analysis incorporated 2159 patients, with 192 deaths (8.9 %) occurring within 28-day of ICU admission. we constructed a nomogram that incorporates risk factors including heart rate (HR), traumatic brain injury (TBI), oxygen saturation (SpO2), systolic blood pressure (SBP), ventilation, and Sequential Organ Failure Assessment (SOFA) score on the first day of admission to ICU. The nomogram outperformed SOFA and Model 1 (risk factors including SBP, SpO2, TBI and ventilation) with an area under the receiver operating characteristic curve (ROC) of 0.854 (95 %CI 0.736-0.791, P < 0.001) in the training cohort and 0.859 (95 %CI 0.713-0.794, P < 0.001) in the validation cohort. The analysis of the calibration curve demonstrated that the nomogram exhibited a strong alignment with the observed 28-day mortality rates in severe thoracic trauma patients.
Conclusions:
The study identified independent risk factors associated with the 28-day mortality risk and developed predictive nomogram models for ICU patients suffering from severe thoracic trauma. The nomogram shows promise in guiding strategies aimed at improving prognosis for patients with such injuries.
