论文部分内容阅读
目的运用受试者工作特征(ROC)曲线评价脑脊液(CSF)及血清(Serum)中髓鞘碱性蛋白(MBP)早期诊断Guillain-Barré综合征(GBS)的价值。方法用ELISA法检测GBS组(36例)中MBP的水平并和对照组(33例)相比较。运用ROC曲线评价CSF及Serum中MBP诊断GBS的敏感性及特异性。结果GBS组CSF中MBP的水平明显高于对照组(P<0.01),Serum中MBP水平和对照组相比差异无统计学意义。CSF MBP诊断GBS的ROC曲线下面积(AUC)为0.804±0.056,最佳分界值为0.65pg/mL。以CSF MBP≥0.65pg/mL来预测GBS,敏感性为80.6%,特异性为78.8%。Serum MBP ROC曲线的AUC为0.548±0.070,最佳分界值为0.20pg/mL。以Serum MBP≥0.20pg/mL来预测GBS,敏感性为44.4%,特异性为69.7%。两条曲线AUC的差异有统计学意义(P<0.01)。结论CSF MBP水平对诊断早期GBS有一定的准确性,可以作为早期诊断GBS较为敏感的指标。
Objective To assess the value of early diagnosis of Guillain-Barré syndrome (GBS) in cerebrospinal fluid (CSF) and serum myelin basic protein (MBP) using receiver operating characteristic (ROC) curves. Methods The level of MBP in GBS group (36 cases) was detected by ELISA and compared with that in control group (33 cases). The ROC curve was used to evaluate the sensitivity and specificity of MBP in CSF and Serum in the diagnosis of GBS. Results The level of MBP in CSF of GBS group was significantly higher than that of control group (P <0.01). The MBP level in Serum was not significantly different from that in control group. The area under the ROC curve (AUC) for CSF MBP diagnosis GBS was 0.804 ± 0.056, the best cut-off value was 0.65pg / mL. GBS was predicted with CSF MBP> 0.65 pg / mL with a sensitivity of 80.6% and a specificity of 78.8%. The AUC of the Serum MBP ROC curve was 0.548 ± 0.070 with an optimal cutoff of 0.20 pg / mL. GBS was predicted with Serum MBP ≧ 0.20 pg / mL with a sensitivity of 44.4% and a specificity of 69.7%. The difference of AUC between the two curves was statistically significant (P <0.01). Conclusion The CSF MBP level has certain accuracy in the diagnosis of early GBS and can be used as a sensitive indicator of early diagnosis of GBS.