肝功能检验项目模型对初发甲状腺功能亢进的应用评价

发布时间：2018-05-06 06:02

本文选题：肝功能 + 诊断试验　；参考：《苏州大学》2015年硕士论文

【摘要】：目的:建立针对甲状腺功能亢进症初诊患者的肝功能检验项目筛查模型,为临床早期发现甲亢患者提供参考依据,也为临床判断甲亢性肝损害提供指导性建议。方法:(1)甲状腺功能亢进症初诊患者的肝功能常规项目的评价及筛查模型的建立与验证查阅甲状腺功能亢进症初诊患者的病例352份和正常人群体检结果434份,获得13项[总胆红素(TB)、直接胆红素(DB)、间接胆红素(FB)、总蛋白(TP)、白蛋白(ALB)、球蛋白(GLOB)、白蛋白与球蛋白的比值(A/G)、谷氨酸氨基转移酶(ALT)、天门冬氨酸氨基转移酶(AST)、γ-谷氨酰基转移酶(GGT)、总胆汁酸(TBA)、乳酸脱氢酶(LDH)、碱性磷酸酶(ALP)]结果,留取120例甲状腺功能亢进症初诊患者和146份正常体检人群的血清,测定以上13项指标。通过比对ROC曲线下面积(AUC),评价各单项指标对甲状腺功能亢进症初诊患者的诊断价值。以查阅病例组为训练组,运用Logistic回归统计方法,建立甲状腺功能亢进症初诊患者的肝功能常规检验项目模型,以留取的血清组为验证组,对该模型进行验证,评价模型的诊断价值。(2)尚未普及肝功能项目评价对留取120例甲状腺功能亢进症初诊患者和146份正常体检人群的血清,同时测定6项尚未广泛普及的肝功能检查项目[5,-核苷酸酶(5,-NT),α-岩藻糖苷酶(AFU),单胺氧化酶(MAO),亮氨酸氨基肽酶(LAP),天门冬氨酸氨基转移酶同工酶(m-AST),谷氨酸脱氢酶(GLDH)],通过比对ROC曲线下面积(AUC),评价6项指标对甲状腺功能亢进症初诊患者的诊断价值,同样运用Logistic回归统计方法,建立肝功能常规检验项目和6项未普及项目对初发甲亢的联合诊断模型,计算联合模型的AUC,将联合模型与肝功能常规模型进行比较,并探讨这6项指标与常规肝功能检验项目的相关性。为临床选择肝功能检验项目提供参考(3)亮氨酸氨基肽酶(LAP)作为甲状腺功能亢进的一项独立诊断指标的探讨将120例验证组的甲亢病例分为甲亢肝损害组和甲亢肝未损害组,比较LAP在这几组中的均值与方差,并研究LAP与甲状腺激素的相关性。结果:(1)肝功能常规项目检查中,筛查甲状腺功能亢进症初诊患者时ROC曲线下面积(AUC)结果前三位依次为ALP(0.884);ALT(0.780);GGT(0.702),初发甲亢患者肝功能辅助诊断模型由DB、TP、GLOB、ALT、LDH、ALP 6项指标组合构建,训练组AUC为0.937,验证组AUC为0.958,总体差异为2.3%;当诊断界值定位0.4时,训练组诊断准确率为90.2%,验证组诊断准确率为90.5%。模型诊断与临床诊断的一致性分析κ系数达到了0.807。(2)在尚未普及的肝功能检验项目中,诊断初发甲亢AUC最大的为LAP,达到了0.903,其余五项诊断价值较低。肝功能联合诊断模型由TB、FB、TP、ALT、AST、ALP、LAP和5,-NT八项构建,此时模型的AUC为0.982。相关系数最高的肝功能常规项目与未普及项目是:GLDH与AST(r=0.436,p0.01);m-AST与AST(r=0.742,p0.01);LAP与ALP(r=0.497,p0.01);MAO与GGT(r=0.245,p0.01);AFU与ALP(r=0.221,p0.01);5,-NT与GGT(r=0.537,p0.01)。(3)LAP在验证组甲亢整体组,验证组中甲亢伴肝损害组以及验证组中甲亢不伴肝损害组比较没有统计学差异(p0.05),但与正常组比较均具有统计学差异(p0.05),LAP与FT3,FT4的相关系数达到了0.628,0.640,与TSH的相关系数为-0.743。结论:(1)在诊断初发甲亢中,肝功能常规检验指标具有一定的诊断价值,其中以ALP为最高;而以肝功能常规检验指标中DB、TP、GLOB、ALT、LDH、ALP建立的回归模型具有很好的应用价值。(2)在尚未普及的肝功能检验项目中,只有LAP具有较高的诊断价值,因此在评判初发甲亢时,临床可不进行其余五项的检测,以节约成本和时间。与肝功能常规检验项目构建的诊断甲亢模型相比,肝功能联合模型虽然AUC有所提高,但综合评价并没有很大优势,因此肝功能常规项目构建的甲亢诊断模型完全能满足临床的筛查要求。(3)在初发甲亢中LAP的升高显示是由于甲亢性引起的,与肝脏受不受损没有关系,因此LAP建议作为诊断初发甲亢的一个独立危险指标,希望引起临床注意。
[Abstract]:Objective: to establish a screening model of liver function test for patients with hyperthyroidism and to provide reference for early detection of hyperthyroidism and to provide guidance for clinical diagnosis of hyperthyroidism. Methods: (1) the evaluation and screening model of routine liver function items in the first diagnosed patients with hyperthyroidism 352 cases of first diagnosed patients with hyperthyroidism were established and 434 were examined in normal people. 13 items of total bilirubin (TB), direct bilirubin (DB), indirect bilirubin (FB), total protein (TP), albumin (ALB), globulin (GLOB), albumin to globulin (A/G), glutamate aminotransferase (ALT), aspartate ammonia were obtained. The results of aminotransferase (AST), gamma glutamyl transferase (GGT), total bile acid (TBA), lactate dehydrogenase (LDH), alkaline phosphatase (ALP), were left to determine the above 13 indexes in 120 cases of primary hyperthyroidism and 146 normal physical examination people. By comparing the area under the ROC curve (AUC), the thyroid gland was evaluated for the thyroid gland. The diagnostic value of primary patients with hyperfunction of hyperfunction. In order to check case group as the training group, the Logistic regression method was used to establish a routine test model of liver function in the first diagnosed patients with hyperthyroidism, with the retained serum group as the validation group, the model was verified and the diagnostic value of the model was evaluated. (2) the liver function has not been popularized. The project evaluated the serum of 120 patients with primary hyperthyroidism and 146 normal people, and 6 items of liver function tests, [5, 5, -NT, alpha fucosidase (AFU), monoamine oxidase (MAO), leucine aminopeptidase (LAP), and aspartate aminotransferase isoenzyme (m-AST). Glutamate dehydrogenase (GLDH)), by comparing the area under the ROC curve (AUC), evaluate the diagnostic value of 6 indexes for the first diagnosed patients with hyperthyroidism, and also use the Logistic regression method to establish a joint diagnosis model of the liver function routine test and 6 unpopular items for primary hyperthyroidism, and calculate the AUC of the joint model. The combined model was compared with the routine liver function model, and the correlation between the 6 indexes and the routine liver function test project was discussed. The reference (3) LAP as an independent diagnostic indicator of hyperthyroidism was provided for the clinical selection of the liver function test. 120 cases of hyperthyroidism were divided into hyperthyroidism liver. The mean and variance of LAP in these groups were compared and the correlation between LAP and thyroid hormones was compared in the group of damage and hyperthyroidism. Results: (1) in the routine examination of liver function, the first three cases of ROC curve area (AUC) in the first diagnosis of hyperthyroidism were ALP (0.884); ALT (0.780); GGT (0.702), primary hyperthyroidism. The model of liver function assisted diagnosis was composed of 6 indexes, DB, TP, GLOB, ALT, LDH, ALP. The training group AUC was 0.937, the verification group AUC was 0.958, the total difference was 2.3%. The diagnostic accuracy of the training group was 90.2% when the diagnostic value was 0.4, and the diagnostic accuracy of the verification group was the consistency analysis kappa coefficient of the 90.5%. model diagnosis and the clinical diagnosis. 0.807. (2) in the unpopularized liver function test project, the largest AUC in the diagnosis of primary hyperthyroidism was LAP, reached 0.903, and the other five had lower diagnostic value. The joint diagnosis model of liver function was constructed by TB, FB, TP, ALT, AST, ALP, LAP and 5, -NT eight. The model AUC was the normal and unpopularized liver function item with the highest correlation coefficient. GLDH and AST (r=0.436, P0.01), m-AST and AST (r=0.742, P0.01), LAP and ALP (r=0.497, P0.01); 3) in the whole group of hyperthyroidism, there was no statistical difference between hyperthyroidism and liver damage group in the test group and the group of hyperthyroidism without liver damage. P0.05), but compared with the normal group, there were statistical differences (P0.05), LAP and FT3, FT4 correlation coefficient reached 0.628,0.640, and the correlation coefficient of TSH was -0.743. conclusion: (1) in the diagnosis of primary hyperthyroidism, the routine test index of liver function has certain diagnostic value, among which ALP is the highest, and DB, TP, GLOB, in the routine test of liver function. The regression model established by ALT, LDH and ALP has good application value. (2) only LAP has high diagnostic value in the liver function test project which has not been popularized. Therefore, in the evaluation of primary hyperthyroidism, the other five items can not be tested in order to save cost and time. The combined model of liver function has improved AUC, but the comprehensive evaluation does not have great advantages. Therefore, the hyperthyroidism diagnostic model constructed by the routine liver function project can fully meet the clinical screening requirements. (3) the elevation of LAP in the primary hyperthyroidism is due to hyperthyroidism and is not related to the liver damage. Therefore, LAP is recommended as a result. It is an independent risk factor for diagnosis of primary hyperthyroidism, and it hopes to attract clinical attention.

【学位授予单位】：苏州大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：R581.1

【参考文献】