三碘甲状腺原氨酸在脑外伤预后模型中的研究价值

The value of triiodothyronine in the prognostic model of traumatic brain injury

目的探究三碘甲状腺原氨酸纳入脑外伤预后模型中的可行性。方法回顾性分析2018年6月至2021年5月徐州医科大学附属连云港医院收治319例神经外科脑外伤病人临床资料,建立预后模型,检测模型性能。结果模型组中,预后良好组共187人,预后不良组共132人。预后良好组年龄(53.18±14.81)岁,预后不良组年龄(59.54±13.69)岁。预后良好组格拉斯哥昏迷量表(GCS)评分(10.44±2.43)分,预后不良组GCS评分(5.96±2.78)分。预后良好组瞳孔对光反射双侧阳性83.96%(157/187例),一侧阴性10.70%(20/187例),双侧阴性5.35%(10/187例),预后不良组瞳孔对光反射双侧阳性29.55%(39/132例),一侧阴性17.42%(23/132例),双侧阴性53.03%(70/132例)。预后良好组赫尔辛基CT(HCT)评分(3.17±2.48)分,预后不良组HCT评分(7.08±3.11)分。预后良好组抗凝、抗血小板药物服用史5.88%(11/187例),预后不良组抗凝、抗血小板药物服用史13.64%(14/132例)。预后良好组白细胞(13.58±5.76)×10^(9)/L,预后不良组白细胞(16.92±6.45)×10^(9)/L。预后良好组T3水平(1.20±0.32)mmol/L,预后不良组T3水平(0.91±0.03)mmol/L。逐步回归及多因素分析后,年龄,GCS评分,瞳孔对光反射双侧阴性,HCT评分,抗凝、抗血小板药物服用史,白细胞计数及T3水平均与预后不良相关。模型A包括年龄,GCS评分,瞳孔对光反射;模型B在模型A的基础上添加HCT评分及抗凝、抗血小板药物服用史;模型C在模型B的基础上添加白细胞计数及T3水平。随着纳入因素的增多,模型的性能逐渐增强。模型C是最佳预测模型,经内部验证及外部验证依然有较强性能。结论T3水平与预后不良成负相关,做为实验室指标与白细胞计数共同纳入脑外伤预后模型可以提高模型的性能。

Objective To explore the feasibility of taking triiodothyronine into the prognostic model of brain injury.Methods The clinical data of 319 cases with neurosurgical trauma treated in Lianyungang Hospital Affiliated to Xuzhou Medical University from June 2018 to May 2021 were analyzed retrospectively,a prognostic model was established,and the performance of the model was tested.Re⁃sults In 319 cases,there were 187 patients with good prognosis and 132 patients with poor prognosis.Patients in the good prognosis group aged(53.18±14.81)years old,and those in the poor prognosis group(59.54±13.69)years old.Glasgow Coma Scale(GCS)score of the good prognosis group was(10.44±2.43),and that of the poor prognosis group(5.96±2.78).In the good prognosis group,83.96%(157/187)of the patients had positive pupillary light reflex on both sides,10.70%(20/187)of the patients had negative pupillary light reflex on one side,and 5.35%(10/187)had negative pupillary light reflex on both sides.In the poor prognosis group,29.55%(39/132)of the patients had positive pupillary light reflex on both sides,17.42%(23/132)of the patients had negative pupillary light reflex on one side and 53.03%(70/132)had negative pupillary light reflex on both sides.Helsinki CT(HCT)score of the good prognosis group was(3.17±2.48),and that of the poor prognosis group was(7.08±3.11).The history of anticoagulant and antiplatelet drug intake in the good prognosis group was 5.88%(11/187 cases),and that in the poor prognosis group was 13.64%(14/132 cases).The white blood cell counts in the good and poor prognosis groups were(13.58±5.76)×10^(9)/L,and(16.92±6.45)×10^(9)/L,respectively.The T3 levels in the good and poor prognosis groups were(1.20±0.32)mmol/L,and(0.91±0.03)mmol/L.Stepwise regression and multivariate analysis results showed that age,GCS score,negative bilateral pupil reflex,HCT score,history of anticoagulant and antiplatelet medication,white blood cell count and T3 level were all associated with poor prognosis.Model A included age,GCS score,and pupil reflex to light,model B was added with HCT score and anticoagulant and antiplatelet medication history on the basis of model A,and model C added with white blood cell count and T3 level on the basis of model B.With the increase in included factors,the performance of the models was gradually enhanced.Model C was the best prediction model,which had strong performance after internal and external verification.Conclusion T3 level is negatively correlated with poor prognosis,whose inclusion with white blood cell count in the prognostic model can improve the performance of the model.