目的通过临床表现及辅助检查,建立结核性脑膜炎和病毒性脑膜炎的鉴别诊断评分模型。方法选取连续住院的结核性脑膜炎和病毒性脑膜炎患者各41例,统计、分析33项临床特征及辅助检查数据。采用logistic回归方法建立评分模型并利用ROC曲线找到诊断切点。结果进入鉴别诊断模型的指标为:CSF糖/血糖<0.5、颅外结核、脑积水、CSF细胞数>37.5×106/L、发病天数>13 d、结核菌特异性抗原阳性。评分≮8分的结核脑可能性大。结论该鉴别诊断评分模型具有较高的诊断敏感度和特异性,可以应用于2种脑膜炎的鉴别诊断。 Objective To establish a differential diagnosis score model of tuberculous meningitis and viral meningitis through clinical manifestations and laboratory examinations. Methods A total of 41 consecutive patients with tuberculous meningitis and viral meningitis were enrolled in this study. A total of 33 clinical features and auxiliary examinations were analyzed statistically. The logistic regression method was used to establish the scoring model and the ROC curve was used to find the diagnostic point. Results The indexes of entering differential diagnosis model were as follows: CSF glucose / blood glucose <0.5, extracranial tuberculosis, hydrocephalus, CSF cell count> 37.5 × 106 / L, onset time> 13 days. Score ≮ 8 points TB brain possibility. Conclusion The differential diagnostic scoring model has high diagnostic sensitivity and specificity and can be applied to the differential diagnosis of two kinds of meningitis.