Factors associated with pulmonary infarction. A discriminant analysis study

To determine the relative importance of multiple interrelated factors that have been considered to contribute to pulmonary infarction, the authors performed a discriminant analysis on consecutively autopsied patients with pulmonary embolism. From the clinic records of 45 individuals, the authors tabulated the underlying illness, history of valvular or ischemic heart disease, right and left ventricular failure, sepsis, shock, malignancy, premortem functional status, and the clinician's suspicion of pulmonary embolism. At postmortem examination, the authors measured and recorded the extent of emphysema, pneumonia, neoplasia, pulmonary vascular atherosclerosis; thickness and dilatation of both cardiac ventricles; the presence of valvular heart disease; the number, diameter, and amount of occlusion of the pulmonary arteries that contained thromboemboli; the extension of the clot, the size of the infarct; the Reid-Index; and the thickness of pulmonary and bronchial arterial wall. The major determinants of infarction were as follows: poor premortem functional status, the number of lobes having emboli, left ventricular failure, and the presence of lung cancer. The authors then tested the equation generated from these patients on 21 additional patients. The discriminant function correctly classified 81% of first group and predicted the occurrence of infarction in new patients with 70% accuracy. The size of the infarct was most correlated with the use of vasodilators and the embolic burden.