Are anomalies in endothelial cell density measurements of donor eyes due to systematic bias?
OCT 24, 2022
Eye bank technicians measured endothelial cell density in a large study from one eye bank to evaluate possible anomalies in measurements. Heat maps revealed a discontinuity at and above 2500 cells/mm2.
This cross-sectional study from one eye bank included cornea endothelial cell density data for 44,479 donor eyes from June 2012 to June 2016. Endothelial cell density was measured by eye bank technicians using the variable frame method. A heat map was created to analyze endothelial cell density measurements to evaluate possible anomalies in measurement.
Mean donor age was 58 years, and mean corneal endothelial cell density was 2717 cells/mm2. Heat maps demonstrated a discontinuity at 2500 cells/mm2; there was an accentuated endothelial cell density at and immediately above 2500 cells/mm2 with a gap below this value.
The data is from one eye bank and therefore may not be generalizable across all donor corneal tissue. The cause of the anomaly at 2500 cells/mm2 is unexplained and unable to be confirmed using only the data available from this study.
As the algorithm used to produce endothelial cell measurements is unlikely to have a defect at 2500 cells/mm2, the findings are thought to represent a systematic human bias toward increasing cell counts above the 2500 cells/mm2 threshold. Both the Cornea Donor Study and Cornea Preservation Time Study, two large studies evaluating the relationship between donor characteristics and corneal transplant outcomes, used endothelial cell density cutoffs under 2500 cells/mm2, so this is not a pre-determined threshold. However, surgeon preference is a factor in tissue acceptance, and it is unclear how much endothelial cell density factors into tissue acceptance for surgeons and what the subjective threshold is for tissue acceptance.