Abstract
The aim of our prospective multicentric study was to develop a screening method for pulmonary hypertension in patients with chronic lung diseases. We investigated 710 patients in 10 hospitals: 315 males and 109 females with chronic obstructive pulmonary disease, and 286 males with silicosis. Manifest pulmonary hypertension was defined as pulmonary artery pressure greater than 20 mmHg (2.7 kPa) at rest. The multivariate statistical method used was a stepwise discriminant analysis. In males with chronic obstructive pulmonary disease, the diameter of the right descending pulmonary artery, forced expiratory volume in one second (FEV1) arterial oxygen tension (PaO2) at rest, and age turned out to be relevant for discrimination of groups with and without manifest pulmonary hypertension. For females the FEV1/FVC (forced vital capacity) ratio replaced the absolute value of FEV1 in the calculated discriminant function. In females, sensitivity and specificity were below 80%. In males, both were distinctly above 80%. In silicosis, the diameter of the right descending pulmonary artery was much less important, since it could frequently not be measured precisely. In these cases, precision of the prediction of about 80% could only be obtained by combined evaluation of spirometry, PaO2 during exercise, and body plethysmography. The calculated discriminant functions are appropriate for screening patients with risk of pulmonary hypertension. For different chronic lung diseases, and for both sexes, different combinations of parameters are relevant. The method is recommended to select patients who should undergo an invasive examination of pulmonary haemodynamics.