Abstract
Purpose::
To determine the corneal endothelial sample profile in specular microscopy with reliability and reproducibility.
Methods::
A retrospective study was performed of 100 right eyes of patients who had undergone corneal specular microscopy, with different clinical situations, in March 2006 in a private clinic. The Bio-Optics specular microscopy was used with the Bambi 2000 Plus software. To compute the sample size we used the software package, CELLS ANALYZER - Corneal Endothelial Statistical LabPat Rec with planned reliability degree (RD) 95%. The sampling process of acquiring endothelial images and counting those cell images were stoped when the calculated relative error (RE) < 0.05. Endothelial samples were evaluated in relation to calculated sample size, considered number of endothelial images, and cell number of the specular microscopy exams.
Results::
The sex distribution was 64 (64%) of the eyes of female patients and 36(36%) of male patients. The minimum age was 16 years, the maximum 96 years and the average 51±19 years. The smallest computed sample size was 57 cells, the largest 315 cells and the average was 138±45 cells; The smallest endothelial sampling had 1 image, the largest 7 images and the average 2.35±1.05 images.
Conclusions::
To make corneal specular microscopy with reliability and reproducibility (RD 95% and RE<0.05) the sampling process needed 1 to 7 (2.35±1) different images and a large number of endothelial cells counted in different areas of corneal endothelial mosaics.
Clinical Trial::
BANPESQ - UFPR - BRAZIL
Keywords: cornea: endothelium • clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology • microscopy: confocal/tunneling