Methodological problems in placental morphometry: apologia for the use of stereology based on sound sampling practice.

Several problem areas in morphometry of human and animal placentae are reviewed. Attention is given to methods of tissue processing (handling, mode of fixation, embedding, shrinkage) and sampling (of organs, tissue blocks, sections, micrographs). Principal sources of bias and sampling variability are identified and the crucial importance of randomized sampling is emphasized. Methods for obtaining structural quantities from sections are compared. The case is made for estimating absolute values (volumes, surface areas, lengths, numbers, thicknesses) using stereological principles rather than relying on planar data (profile areas, perimeter lengths, numbers, apparent thicknesses). Absolute values may be obtained simply and efficiently without resort to expensive measuring devices. Finally, morphological descriptors suitable for correlating with functional data or for comparing normal and diseased organs are surveyed.