【作者】 刘鑫；

【导师】 曹春霞；

【作者基本信息】 天津大学 ， 生物医学工程， 2023， 硕士

【摘要】 目的:本研究旨在分析创伤患者心血管疾病(Cardiovascular disease,CVD)的发病特征,探究创伤与CVD发病风险关系,构建创伤患者CVD风险预测模型,为创伤患者CVD预防和干预提供理论参考。方法:依托开滦研究,研究对象包括创伤患者和未受伤参与者(对照)。创伤患者为2006年5月至2018年12月因创伤入院治疗的患者,为每一例创伤患者根据性别、年龄(±3岁)从未受伤参与者中匹配4名对照。创伤患者从创伤入院开始随访(对照从同一天开始),直至CVD发生、死亡、或2020年12月31日研究截止。采用描述性分析和组间比较,分析创伤患者CVD发病特征;使用Cox回归模型探究创伤与CVD发病风险关系;将创伤患者随机分为训练集(70%)和验证集(30%),首先用训练集筛选、处理预测变量,并用Cox回归模型拟合创伤患者CVD风险预测模型,而后于验证集中通过C-index和校准曲线评价模型效能。结果:研究对象共计包括3 986名创伤患者和15 944名对照。在8.67(4.67-12.78)年中位随访时间内,创伤患者新发CVD事件402例,累计发病率为10.09%,发病密度为13.07/1 000人年;创伤患者CVD多发生在伤后5.77(3.24-8.71)年内,此后呈逐渐下降趋势。与对照组相比,创伤患者CVD发病风险增加35%(Hazard Ratio,HR=1.35,95%CI:1.21-1.52),其中,中重度伤、创伤年龄<35岁、女性、中心性肥胖患者的CVD发病风险,其HR(95%CI)分别为1.44(1.21-1.71)、5.36(2.71-10.64)、5.10(2.65-9.82)和1.63(1.40-1.90)。创伤患者CVD风险预测模型的C-index在训练集中为0.73(0.70-0.76),在验证集中为0.70(0.65-0.75),表明模型具有良好的区分度;校准曲线显示创伤患者CVD发病预测概率与实际发生率具有良好的一致性。结论:对2006-2018年开滦研究队列中的创伤患者进行分析发现,创伤患者CVD事件多发生在创伤后前5.77年,且创伤患者CVD发病风险增加,尤其是中重度伤、创伤年龄较小、女性、中心性肥胖患者CVD发病风险更高。由年龄、性别、吸烟、降糖药服用状况、收缩压、腰围、空腹血糖、高密度脂蛋白胆固醇、高敏C反应蛋白9个变量组成的CVD风险预测模型,能够对创伤患者CVD发病风险进行有效预测。更多还原

【Abstract】 Objective:The objective of this study was to analyze the characteristics of cardiovascular disease(CVD)onset in trauma patients,explore the association between trauma and the risk of CVD onset,and construct a CVD risk prediction model for trauma patients,so as to provide a theoretical reference for CVD prevention and intervention in trauma patients.Methods:Based on the Kailuan Study,the study population consisted of trauma patients and common participants(controls).Trauma patients are those admitted for trauma between May 2006 and December 2018,and four controls were matched for each trauma patient based on sex and age(±3 years)from common participants.Trauma patients were followed from the time of trauma admission(controls from the same date)until the onset of CVD,death,or the end of the study on December 31,2020.Descriptive analysis and between-group comparisons were used to analyze the characteristics of CVD onset in trauma patients;The association between trauma and the risk of CVD was investigated by Cox regression models;70%of trauma patients were randomly selected as the training set and the remaining 30%as the validation set,and the training set was first used to screen and process predictor variables and fit a prediction model of CVD risk in trauma patients with Cox regression models,while the C-index and calibration curves were used to evaluate the model in the validation set.Results:The study population included 3 986 trauma patients and 15 944 controls.During a median follow-up time of 8.67(4.67-12.78)years,402 CVD events were identified in trauma patients,and the cumulative incidence of CVD in trauma patients was 10.09%,with an incidence density of 13.07 per 1,000 person-years.CVD in trauma patients occurred mostly within 5.77(3.24-8.71)years after the trauma.Compared to controls,the risk of CVD increased by 35%in trauma patients(Hazard Ratio,HR=1.35,95%CI:1.21-1.52),with a higher risk of CVD in patients with moderately severe injuries,age at trauma<35 years,female,and central obesity,with HRs(95%CI)of1.44(1.21-1.71),5.36(2.71-10.64),5.10(2.65-9.82),and 1.63(1.40-1.90),respectively.The C-index of the CVD risk prediction model for trauma patients was 0.73(0.70-0.76)in the training set and 0.70(0.65-0.75)in the validation set,indicating good discrimination of the model;the calibration curve showed good agreement between the predicted probability and the actual incidence of CVD in trauma patients.Conclusions:Analysis of trauma patients in the Kailuan Study cohort from 2006to 2018 revealed that the majority of CVD events in trauma patients occurred in the first 5.77 years after trauma and that the risk of CVD was increased in trauma patients,especially in patients with moderate to severe trauma,younger age at trauma,female,and central obesity.A CVD risk prediction model consisting of nine variables,including age,gender,smoking,glucose-lowering drug-taking status,systolic blood pressure,waist circumference,fasting glucose,high-density lipoprotein cholesterol,and high-sensitivity C-reactive protein,can effectively predict the risk of CVD development in trauma patients.更多还原