基于常规实验室检查的良恶性胸/腹水鉴别诊断模型研究

A model based on conventional laboratory parameters for differential diagnosis of malignant and benign pleural and peritoneal effusions

目的分析常规实验室检查结果对胸/腹水良恶性质的鉴别诊断价值,并建立简便的数学诊断模型。方法查阅胸/腹水患者的病案资料,核实病因诊断,收集血细胞分析、血生化检查、胸/腹水常规及生化检查、血清肿瘤标志物检测结果等。单因素Logistic回归分析各项检查单项对良恶性胸/腹水的鉴别诊断价值,多因素Logistic回归分析建立数学诊断模型,并评价其对恶性胸/腹水的诊断价值。结果常规实验检查的27项单项指标中有13项对胸/腹水性质有鉴别诊断价值(P=0.000-0.018)。血白细胞计数、血总胆红素、胸/腹水Rivalta试验和血CEA进入数学模型,该模型诊断恶性胸/腹水的敏感度为89.2%,准确度为85.8%,均高于细胞学和染色体检查,特异度为81.3%,低于后两者。结论部分常规实验室检查的单项指标对良恶性胸/腹水的鉴别诊断具有一定的价值,由常规实验室检查的4项单项指标建立的数学模型对良恶性胸/腹水的鉴别具有较大的价值。

Objective To analyze the values of conventional laboratory parameters in differential diagnosis of malignant and benign pleural and peritoneal effusions and establish a simple diagnostic model based on thems. Methods The medical records of each patient with pleural or peritoneal effusion was reviewed retrospectively and the etiological diagnosis was verified. The following data were collected： blood cell analysis, serum biochemistry, effusion routine and biochemistry, serum tumor biomarkers. Univariate Logistic regression was performed to analyze the diagnostic value of each laboratory parameter, and multivariate Logistic regression was performed to establish a mathematical model and its diagnostic performance was evaluated. Results Thirteen of 27 single parameters were significant in differentiating nmlignant effusions from benign ones （P = 0. 000- 0. 018）. The parameters entered the model were blood white cell count, total bilimbin, Rivalta test and carcinocmbryonic antigen. The sensitivity and accuracy of the model in malignant effusion diagnosis were 89.2% and 85.8 %, higher than cytology and chromosome analyses, and specificity 81.3 %, lower than them. Conclusion Some of conventional laboratory parameters are significant for differentiating malignant from benign effusions, and the mathematical model based on 4 parameters is simple and valuable in the differential diagnosis.