[Value and rank of diagnostic laboratory parameters in cardiovascular diseases. Results of multivariant and discriminant analyses of the Berlin EBMO-Cor population study].

Based on present clinical-chemical parameters a representative population were studied; importance and order of rank of 10 data chosen from a total data pool are determined by means of multivariant and discriminant analysis. As a result the validity of characters with and without apriori probabilities for 5 pairs of classes (persons suffering from a heart disease in comparison to healthy persons divided in sexes and overall, as well as examination of separate sexes for healthy persons and those suffering from a heart disease) are examined (alpha less than or equal to 0.05). This calculation tested the importance of the parameters hemoglobin and creatinine in being different in males and females. To diagnose persons suffering from a heart disease (X-ray-morphologically defined suspects of heart and vessel diseases) from healthy persons, the optimized number of characteristics were determined in cholesterol, erythrocyte sedimentation rate, ASAT, blood glucose (independent of sex). By means of these results it is possible to identify persons suffering from a heart disease from healthy persons and to call for illustrative laboratory examinations using already mentioned parameters and the function of discriminance W = p sigma i = 1 ai X Xi. The determination of sensitivity and specification yielded a value of up to 98 per cent resp. 94 per cent depending on whether with or without apriori probabilities. This enables use to be modified to suit different purposes.