THE REPORT by Klein et al1 in this month's ARCHIVES, presenting further analyses of data from the Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR), raises interesting issues about the Early Treatment Diabetic Retinopathy Study2(ETDRS) retinopathy severity scale and its use in epidemiologic studies and clinical trials. After adding 2 steps in the mild nonproliferative range of the ETDRS retinopathy scale, the authors assessed whether a 1 or more or 2 or more step progression along the ETDRS retinopathy severity scale during a 4-year period is predictive of subsequent progression to proliferative retinopathy or clinically significant macular edema during the next 6 years. The results of the analyses of this well-designed and well-conducted population-based study, the WESDR, are displayed in Figure 3 of their article.1 The data show that patients who had experienced a 1-step progression of retinopathy during the first 4 years of follow-up were 5.8 times more likely to develop proliferative diabetic retinopathy and 3.8 times more likely to develop clinically significant macular edema in the subsequent 6 years of follow-up than those patients who did not have a 1-step progression at 4 years. An additional issue raised by the authors is the use of a 1-step progression as a plausible outcome measure for clinical trials of diabetic retinopathy, based on the results of this epidemiologic study.
Chew EY. Surrogate Outcomes for Clinical TrialsUse With Caution. Arch Ophthalmol. 2001;119(4):576-578. doi:10.1001/archopht.119.4.576