Construction of a prognosis prediction model for traumatic spinal cord injury based on Logical-Nomogram

Objective:To analyze the factors associated with poor prognosis of traumatic spinal cord injury(TSCI) based on multivariate Logistic regression analysis, and to construct and verify a Nomogram prediction model.Methods:A total of 250 TSCI patients admitted to the 920th Hospital of Joint Logistic Support Force of the Chinese People’s Liberation Army from March 2020 to September 2022 were selected as research objects, and randomly divided into training group (n=175) and verification group (n=75) according to a ratio of 7:3.The Japanese Orthopaedic Society (JOA) score was used to evaluate the prognosis of patients before treatment and 6 months after treatment.A JOA score improvement rate ≥60% was considered as good prognosis group, and a JOA score improvement rate < 60%was considered as poor prognosis group.The prognosis of the two groups at 6 months after treatment was statistically compared.Univariate and multivariate Logistic regression analysis were used to analyze the influencing factors of poor prognosis, and a Nomogram prediction model was established based on the influencing factors to verify the prediction efficacy and clinical efficacy of this model.Results:The spinal canal encroachment rate ≥50%, injury severity (complete injury), injury to treatment time ≥8 h, decreased peripheral blood fibrinogen (FIB) level, serum high mobility group protein B1(HMGB1) and nuclear factor κB (NF-κB) levels, and increased peripheral blood neutrophil count to lymphocyte count ratio (NLR) level were independent risk factors for poor prognosis (P< 0.05).The area under the curve(AUC)of the Nomogram prediction model for predicting poor prognosis was 0.944, and it had a positive net benefit.Conclusion:The spinal canal encroachment rate ≥50%, injury degree, treatment time, and peripheral blood HMGB1, NF-κB, NLR and FIB levels are risk factors for poor prognosis in patients with TSCI.The Nomogram model based on the above factors has good predictive efficacy for poor prognosis, which is helpful for clinical screening of high-risk groups and making treatment plans.