Systematic Assessment of Clinical Methods to Diagnose and Monitor Diabetic Retinal Neuropathy

Table 3

Number of studies with imaging investigations demonstrating evidence of changes before traditionally graded retinopathy.

Imaging	Compelling	Moderate	Weak	No evidence despite investigation
OCT	23	19	4	1
Retinal thickness	13	16	1	2
ONH shape	13	0	1	0
Vessel calibre changes	12	3	0	0
Capillary density in perifoveal intercapillary area	11	1	0	0
Retinal thickness	13	16	1	2
FFA	21	14	1	0
Corneal confocal	3	7	0	0
Retinopathy detection software	60	22	1	0
Digital fundus photography	67	31	1	0
Retinal image processing (including SVP)	69	31	0	0

While retinopathy detection software, digital fundus photography, and retinal image processing all had strong levels of indicative, if not compelling evidence for diabetic changes within the parameters of individual studies, there is little agreement on software or algorithm usage. Furthermore, in many cases, evidence was of early-stage clinical diabetic retinopathy, not preclinical retinal neuropathy. While OCT is an excellent imaging modality for many purposes in evaluating retinal health, relatively it was not a reliable indicator of diabetic retinal neuropathic changes in the literature.