23. Diabetic retinopathy features at baseline ‐ Studies undertaking multivariable regression analyses to determine the effect of diabetic retinopathy features at baseline on progression to PDR.

Study	Study type	Time years	N at baseline	Adjustment factors	Effect estimate			P value	Comments
					Type	Value	95% CI
Type 1 and 2 diabetes
Lee 2017	Electronic database	5	2823	Age, sex, race, initial visual acuity	HR	1.77a 1.47b	1.25 to 2.49 0.94 to 2.31	0.001     0.88
WESDR Klein 1995	Prospective cohort	4	236	HbA1c, duration of DM, age, sex, age at DM diagnosis, SBP, DBP, BMI, proteinuria and type of DM	OR	1.04c 1.05d 5.77e	1.02 to 1.07 1.01 to 1.09 2.24 to 14.89	< 0.001   0.006     < 0.001

BMI: body mass index; CI: confidence interval; DBP: diastolic blood pressure; DM: diabetes mellitus; DR: diabetic retinopathy; HbA1c: glycated haemoglobin/haemoglobin A1c; HR: hazard ratio; NPDR: non‐proliferative diabetic retinopathy; OR: odds ratio; PDR: proliferative diabetic retinopathy; RR: risk ratio; SBP: systolic blood pressure; vs: versus

^aIRMA vs venous beading in four quadrants
^bDot/blot haemorrhages vs venous beading in four quadrants
^cDifference in number of microaneurysms at baseline and follow‐up
^dRatio between number of microaneurysms at baseline and follow‐up
^eDifference of ≥ 16 microaneurysms at baseline and follow‐up