Unlimited volumes of laboratory data: a confusing and diagnostically deceptive product of modern technology

The large volumes of laboratory data currently available in clinical practice can lead to erroneous conclusions. Our current statistical interpretation of these data is univariate (one variable at a time) and often not age- and sex-corrected. Using an optimal technique of multivariate analysis, a SMAC profile of 19 tests performed on normal subjects resulted in over a 500% improvement in defining the reference range. Using physiologic subsets of the SMAC profile for patients, improvements in interpretation of between 100% and 300% are possible. Results indicate a serious clinical problem that will require modification of laboratory reports using modern technology as an adjunct for diagnostic medicine.