I continue to be impressed with the output of published reports stemming
from the data of the Endophthalmitis Vitrectomy Study (EVS) Group. However,
I have concerns regarding some of the conclusions presented by Dr Doft and
coworkers in their most recent publication.1
The authors should accept the implications of the statistical analysis of
their results instead of making statements that are not supported by their
data. After reading their article, my interpretation is that when the data
support the notion that diabetics should have more problems, it is reported,
even when there is no "significant difference" to support this assumption.
When the diabetic group appears to have fewer problems than the nondiabetic
group, however, these subsets are reported as showing "no significant difference."
Enzenauer RW. Diabetes and the EVS: A Different Interpretation of the Results. Arch Ophthalmol. 2002;120(2):231-232. doi: