论文部分内容阅读
目的应用决策树模型(CART),纳入总淋巴细胞(TLC)计数、血红蛋白(HGB)等多个血常规指标,研究未抗病毒治疗HIV感染者TLC与CD4~+T淋巴细胞(简称CD4细胞)计数的相关性,并与受试者工作特征曲线(简称ROC曲线)常规分类方法比较。方法选取昆山市未抗病毒治疗HIV感染者297例,采集433份血样标本,检测血常规指标和CD4细胞计数。将血常规指标与CD4细胞计数进行相关性分析,以CD4细胞计数≤350个/uL、≤500个/uL为临界点,分别计算CART、常规ROC曲线分类方法的灵敏度、特异度、阳性预测值(PPV)、阴性预测值(NPV)和约登指数。结果 TLC、HGB、白细胞(WBC)计数、红细胞(RBC)计数、血小板(PLT)计数、中性粒细胞(WLCC)计数与CD4细胞计数均显著相关(P<0.05),纳入TLC、HGB、RBC、WBC指标,CART模型分类CD4细胞计数≤350个/uL的灵敏度、特异度、PPV、NPV和约登指数分别为36.4%、95.3%、84.2%、68.6%和0.317;纳入TLC、HGB、RBC指标,CART模型分类CD4细胞计数≤500个/uL的灵敏度、特异度、PPV、NPV和约登指数分别为92.0%、40.9%、78.0%、69.2%和0.329;两个CART模型均略优于ROC曲线分类方法。结论应用CART模型纳入TLC、HGB等多个血常规指标能有效预测CD4细胞计数,可在资源有限地区用于未治疗HIV感染者疾病进展监测。
Objective To investigate the relationship between TLC and CD4 ~ + T lymphocytes (CD4 cells) in HIV-infected patients without any antiviral therapy by using the CART model, including the total lymphocyte count (TLC) and hemoglobin (HGB) The correlations were compared with the conventional classification of receiver operating characteristic curves (ROC curves for short). Methods A total of 297 HIV-infected patients without anti-virus in Kunshan City were enrolled. 433 blood samples were collected and blood routine indexes and CD4 cell counts were measured. The correlation between blood routine and CD4 count was analyzed. The sensitivity, specificity and positive predictive value of CART and routine ROC curve were calculated respectively with CD4 count ≤350 cells / uL and ≤500 cells / uL as the critical point (PPV), negative predictive value (NPV) and Youden index. Results TLC, HGB, WBC count, RBC count, PLT count and WLCC count were all significantly correlated with CD4 count (P <0.05). TLC, HGB, RBC The sensitivity, specificity, PPV, NPV and Youden index of CD4 cell count of CD4 cell count of ≤350 cells / WBC in CART model were 36.4%, 95.3%, 84.2%, 68.6% and 0.317, respectively. The sensitivity, specificity, PPV, NPV and Youden index of CD4 cell count ≤ 500 cells / uL for CART model classification were 92.0%, 40.9%, 78.0%, 69.2% and 0.329, respectively; both CART models were slightly better than the ROC curve Classification. Conclusion The inclusion of multiple blood parameters such as TLC and HGB in CART model can effectively predict CD4 count and can be used to monitor the progression of diseases in untreated HIV-infected patients in limited resources.