Di Bello V, Gori E, Bellina C R, Parodi O, Molea N, Santoro G, Mariani G, Conti U, Magagnini E, Marzullo P
Second Institute of Clinical Medicine, University of Pisa, Italy.
J Nucl Cardiol. 1994 Jul-Aug;1(4):372-81. doi: 10.1007/BF02939958.
The incremental diagnostic information of two noninvasive tests for the detection of coronary artery disease (CAD), dipyridamole echocardiography, and exercise 201Tl myocardial scintigraphy was assessed in a series of 102 patients with ordered logistic regression and receiver-operating characteristic curves.
Patients were selected from those referred to our cardiovascular centers with the clinical suspicion of CAD. After clinical evaluation, all patients underwent both noninvasive tests during hospitalization 2 weeks before coronary arteriography. The coronary arteriogram was used as a gold standard: CAD was defined as the presence of one or more vessels with 50% or greater narrowing of the luminal diameter. Clinical data were 73.0% +/- 5.7% accurate in the prediction of CAD. The addition of dipyridamole echocardiographic data to the clinical model yielded a diagnostic accuracy of 88.3% +/- 4.3% (p < 0.00001), whereas the addition of thallium scintigraphic parameters to the clinical model improved diagnostic accuracy to 93.8% +/- 2.6% (p < 0.00001). A significant increase in accuracy to 97.2% +/- 1.4% was achieved when thallium scintigraphic data were added to the clinical and dipyridamole-echocardiographic model (p < 0.00001).
Both noninvasive methods for detection of CAD, DET, and ETS showed a good diagnostic accuracy especially when tests-derived parameters were combined with clinical data by means of relative logistic models; nevertheless the ETS model showed a higher sensitivity in comparison with the DET model, essentially in presence of a lower extent of CAD.
在102例患者中，采用有序逻辑回归和受试者操作特征曲线评估了双嘧达莫超声心动图和运动201Tl心肌显像这两种用于检测冠状动脉疾病（CAD）的非侵入性检查的增量诊断信息。
患者选自因临床怀疑CAD而转诊至我们心血管中心的人群。经过临床评估后，所有患者在冠状动脉造影前2周住院期间均接受了这两种非侵入性检查。冠状动脉造影用作金标准：CAD定义为存在一条或多条管腔直径狭窄50%或以上的血管。临床数据对CAD预测的准确率为73.0%±5.7%。将双嘧达莫超声心动图数据添加到临床模型中，诊断准确率为88.3%±4.3%（p<0.00001），而将铊显像参数添加到临床模型中，诊断准确率提高到93.8%±2.6%（p<0.00001）。当将铊显像数据添加到临床和双嘧达莫超声心动图模型中时，准确率显著提高到97.2%±1.4%（p<0.00001）。
检测CAD的两种非侵入性方法，即双嘧达莫超声心动图（DET）和运动铊心肌显像（ETS），均显示出良好的诊断准确率，尤其是当通过相对逻辑模型将检查衍生参数与临床数据相结合时；然而，ETS模型与DET模型相比显示出更高的敏感性，尤其是在CAD程度较低的情况下。
