Objective:To construct a risk prediction model for early post-injury respiratory failure in patients with traumatic cervical spinal cord injury (TCSCI) and validate its efficacy.Methods:A retrospective cohort study was conducted to analyze the clinical data of 393 TCSCI patients admitted to Zhengzhou Central Hospital Affiliated to Zhengzhou University from January 2020 to October 2024, including 294 males and 99 females, aged 18-82 years [59(45, 72)years]. Among them, 76 patients had respiratory failure (19.3%). The patients were randomly divided into the training set ( n=275) and validation set ( n=118) at a ratio of 7∶3. According to the presence of respiratory failure within one week after admission, 275 patients in the training set were divided into respiratory failure group ( n=53) and non-respiratory failure group ( n=222). The demographic data, injury characteristics, laboratory test results, and imaging findings of the patients were collected. Risk factors were determined through univariate analysis and multivariate Logistic regression analysis and a nomogram prediction model was constructed. The area under the receiver operating characteristic (ROC) curve (AUC) and Hosmer-Lemeshow test were used to evaluate the discrimination and calibration of the model. Decision curve analysis (DCA) was plotted to evaluate the clinical effectiveness of the prediction model. Results:The results of the univariate analysis showed that there were significant differences in history of respiratory diseases, causes of injury, Glasgow coma scale (GCS), American Spinal Injury Association (ASIA) classification, ASIA-motor score (AMS), injury severity score (ISS), clinical pulmonary infection score (CPIS), hypoproteinemia and cervical vertebra fracture and dislocation between the respiratory failure group and non-respiratory failure group in the training set ( P<0.05). The results of multivariate Logistic regression analysis indicated that GCS, ASIA classification, CPIS, and hypoproteinemia were independent risk factors for early post-injury respiratory failure in TCSCI patients ( P<0.05). Based on the above four variables, a Logistic regression equation was constructed: Logit( P)=2.361-0.675×ASIA classification+0.419×CPIS-0.358×GCS+0.854×hypoproteinemia. In the prediction model established based on this equation, the AUC was 0.96 (95% CI 0.94, 0.99) in the training set and 0.89 (95% CI 0.82, 0.96) in the validation set. In the calibration curves of the training set and validation set, the prediction curve and reference curve were approximately overlapping, with the average absolute errors of 0.04 and 0.03. DCA results demonstrated that both the training and validation sets exhibited positive net benefits when threshold probabilities fell within ranges of 0%-78% and 0%-87%, respectively. Conclusion:The risk prediction model for early post-injury respiratory failure in TCSCI patients based on GCS, ASIA classification, CPIS and hypoproteinemia has good predictive efficacy and clinical practicability.

Objective:To construct a risk prediction model for early post-injury respiratory failure in patients with traumatic cervical spinal cord injury (TCSCI) and validate its efficacy.Methods:A retrospective cohort study was conducted to analyze the clinical data of 393 TCSCI patients admitted to Zhengzhou Central Hospital Affiliated to Zhengzhou University from January 2020 to October 2024, including 294 males and 99 females, aged 18-82 years [59(45, 72)years]. Among them, 76 patients had respiratory failure (19.3%). The patients were randomly divided into the training set ( n=275) and validation set ( n=118) at a ratio of 7∶3. According to the presence of respiratory failure within one week after admission, 275 patients in the training set were divided into respiratory failure group ( n=53) and non-respiratory failure group ( n=222). The demographic data, injury characteristics, laboratory test results, and imaging findings of the patients were collected. Risk factors were determined through univariate analysis and multivariate Logistic regression analysis and a nomogram prediction model was constructed. The area under the receiver operating characteristic (ROC) curve (AUC) and Hosmer-Lemeshow test were used to evaluate the discrimination and calibration of the model. Decision curve analysis (DCA) was plotted to evaluate the clinical effectiveness of the prediction model. Results:The results of the univariate analysis showed that there were significant differences in history of respiratory diseases, causes of injury, Glasgow coma scale (GCS), American Spinal Injury Association (ASIA) classification, ASIA-motor score (AMS), injury severity score (ISS), clinical pulmonary infection score (CPIS), hypoproteinemia and cervical vertebra fracture and dislocation between the respiratory failure group and non-respiratory failure group in the training set ( P<0.05). The results of multivariate Logistic regression analysis indicated that GCS, ASIA classification, CPIS, and hypoproteinemia were independent risk factors for early post-injury respiratory failure in TCSCI patients ( P<0.05). Based on the above four variables, a Logistic regression equation was constructed: Logit( P)=2.361-0.675×ASIA classification+0.419×CPIS-0.358×GCS+0.854×hypoproteinemia. In the prediction model established based on this equation, the AUC was 0.96 (95% CI 0.94, 0.99) in the training set and 0.89 (95% CI 0.82, 0.96) in the validation set. In the calibration curves of the training set and validation set, the prediction curve and reference curve were approximately overlapping, with the average absolute errors of 0.04 and 0.03. DCA results demonstrated that both the training and validation sets exhibited positive net benefits when threshold probabilities fell within ranges of 0%-78% and 0%-87%, respectively. Conclusion:The risk prediction model for early post-injury respiratory failure in TCSCI patients based on GCS, ASIA classification, CPIS and hypoproteinemia has good predictive efficacy and clinical practicability.