Noninvasive assessment of tissue iron stores

Multivariate regression methods were used to create two equations for calculation of the "iron index," a predictor of bone marrow iron scores. With the use of the most efficient cut-off point, approximately 90% of 210 patients could be diagnosed correctly. Selection of a criterion that minimizes false negatives (missed iron deficiency) would increase the number of unnecessary bone marrow aspirations, but, even then, more than a third of such tests could be avoided. Rapid turnaround time would be desirable for all measurements used to calculate the iron index, although it is recognized that for most hospitalized patients the length of hospital stay is determined by severity of an underlying disease, not the presence of anemia.