Carrión-Valero F, Perpiñá-Tordera M, Sanchis-Aldás J
Service of Pneumology, Hospital Clinico Universitario, Valencia, Spain.
Int J Tuberc Lung Dis. 2003 Sep;7(9):892-8.
To assess the value of discriminant analysis as a method of optimizing the discriminant power of routine parameters in differentiating between malignant and non-malignant pleural effusions.
Retrospective review of the medical records of 245 patients with exudative pleural effusion.
The most powerful predictor of the malignant etiology of pleural effusion was a function that consisted of seven variables: age (years); effusion volume (coded as up to one third = 1, up to two thirds = 2, massive = 3); sedimentation rate (mm/h); monocyte count in the peripheral blood (cells/mm3); bloodstained exudate (coded as yes = 1, no = 2); and glucose (mg/dL) and iron (microg/dL) concentration in pleural fluid. This function showed a sensitivity of 77%, specificity of 85%, positive predictive value (PPV) of 76%, negative predictive value (NPV) of 86%, and was able to give an 82% rate of correct classification. In patients aged 50 years or younger, the NPV ranged between 91 and 98%, whereas in those older than 60 years, the PPV was 89%.
The calculated discriminant function is a simple, rapid, and inexpensive method for screening patients with pleural effusion for malignant etiology.
评估判别分析作为一种优化常规参数判别能力以区分恶性和非恶性胸腔积液的方法的价值。
回顾性分析245例渗出性胸腔积液患者的病历。
胸腔积液恶性病因的最强预测指标是一个由七个变量组成的函数：年龄（岁）；积液量（编码为至多三分之一 = 1，至多三分之二 = 2，大量 = 3）；血沉（mm/h）；外周血单核细胞计数（细胞/mm³）；血性渗出液（编码为是 = 1，否 = 2）；以及胸腔积液中的葡萄糖（mg/dL）和铁（μg/dL）浓度。该函数的敏感性为77%，特异性为85%，阳性预测值（PPV）为76%，阴性预测值（NPV）为86%，正确分类率为82%。在50岁及以下的患者中，NPV在91%至98%之间，而在60岁以上的患者中，PPV为89%。
计算得到的判别函数是一种简单、快速且廉价的方法，用于筛查胸腔积液患者的恶性病因。
