Analysis on risk factors for prognosis of traumatic brain injury in adults and establishment of the prediction model

包明冬,葛俊苗,杨秋子,孙季冬,武秀权,蒋晓帆,罗鹏
DOI: https://doi.org/10.3760/cma.j.cn501098-20221213-00794
2023-01-01
Abstract:Objective:To analyze risk factors for prognosis of adult patients with traumatic brain injury (TBI), construct the prognostic model of TBI and evaluate its predictive value.Methods:A case-control study was used to analyze the clinical data of 522 patients with TBI admitted to Xijing Hospital of Air Force Medical University from March 2011 to September 2019, including 438 males and 84 females; aged 18-75 years [(44.9±15.0)years]. According to the Glasgow outcome score (GOS) at discharge, the patients were divided into good prognosis group (GOS 4-5 points, n=165) and poor prognosis group (GOS 1-3 points, n=357). The two groups were compared with regards to qualitative data such as sex, underlying diseases, causes of injury, multiple injuries, open injuries, intracranial foreign bodies, cerebral herniation, consciousness status on admission and at discharge, surgery, lung infection on admission, tracheostomy, ventilator-assisted ventilation, hospital-acquired pneumonia/pathogenic bacteria and intracranial infection, and quantitative data such as Glasgow coma score (GCS) on admission and at discharge, age, measurements on admission [systolic blood pressure, diastolic blood pressure, mean arterial pressure, temperature, heart rate, creatinine, urea nitrogen, blood sodium, blood potassium, blood glucose, prothrombin time (PT), activated partial thromboplastin time (APTT), platelets, international normalized ratio (INR), pupil size of both eyes] and length of hospital stay. Univariate analysis and Lasso regression analysis were used to screen the risk factors affecting the prognosis of TBI patients, and the selected influencing factors were included in multivariate Logistic regression analysis to identify independent risk factors and construct regression equations. R was used to draw a visual nomogram based on regression equation for predicting the prognosis of TBI patients. The prognostic predictive value of the nomogram was evaluated by using the receiver operating characteristic (ROC) curve, and the area under the curve (AUC), Youden index, sensitivity, specificity and consistency index (C index) were calculated. Results:Univariate analysis showed that there were significant differences between the two groups in underlying diseases, open injuries, cerebral herniation, consciousness status on admission and at discharge, lung infection on admission, tracheostomy, ventilator-assisted ventilation, hospital-acquired pneumonia/pathogenic bacteria, GCS on admission and at discharge, age, and measurements on admission (systolic blood pressure, mean arterial pressure, body temperature, heart rate, creatinine, urea nitrogen, blood potassium, blood glucose, PT, INR, pupil size of right eye) (all P<0.05 or 0.01). There were no significant differences between the two groups in gender, causes of injury, multiple injuries, intracranial foreign bodies, surgery, intracranial infection, measurements on admission (diastolic blood pressure, blood sodium, APTT, platelets, pupil size of left eye) and length of hospital stay (all P>0.05). After screening by Lasso regression model, the results of multivariate Logistic regression analysis showed that GCS on admission ( OR=0.67, 95% CI 0.62, 0.73, P<0.01), age ( OR=1.03, 95% CI 1.01, 1.04, P<0.01), blood glucose on admission ( OR=1.17, 95% CI 1.06, 1.30, P<0.01) and INR on admission ( OR=17.08, 95% CI 2.12, 137.89, P<0.01) could be used as the main risk factors to construct the prediction model, and the regression equation was constructed: Logit [ P/(1- P)]=-0.398× "GCS on admission"+0.024× "age"+0.158×"blood glucose on admission"+2.838×"INR on admission"-1.693. The AUC for the prognosis prediction in adult patients with TBI using R based on a visual nomogram model was 0.87 (95% CI 0.83, 0.89, P<0.01). The Youden index for the predicted probability was 0.60 (sensitivity of 85.2% and specificity of 75.2%), with the C index of 0.87. Conclusion:Age, GCS on admission, blood glucose on admission and INR on admission are the main risk factors affecting the prognosis of TBI in adults, and the nomogram drawn by these parameters can better predict their clinical outcome.
What problem does this paper attempt to address?