Variability of panretinal photocoagulation lesions across physicians and patients. Quantification of diameter and intensity variation.

Background: Photocoagulation lesion intensity relies on the judgement of retinal blanching. Lesions turn out variable due to observer-dependent judgement and time dependency of blanching. We investigated lesion variability per patient and per physician in clinical routine treatments.
Methods: In this observational clinical trial, different physicians performed panretinal photocoagulation for diabetic retinopathy. Study eyes received 20-30 study lesions at 20 ms (three physicians, nine eyes) and 200 ms (four physicians, 12 eyes) irradiation time (532 nm continuous wave photocoagulator, 300 μm spot size). Lesions were imaged after 1 hour with photography and optical coherence tomography (OCT). We measured lesion diameters in fundus and OCT images, and graded intensities according to a previously published six-step classifier.
Results: 200-ms lesions were larger and more severe (568, 474-625 μm [median, IQR], predominantly class 6) than 20-ms lesions (397, 347-459 μm, predominantly classes 3-4). The impact of laser power was small compared to other factors. Lesion intensities and diameters in fundus and OCT images varied significantly between patients and between physicians. Median photographic lesion diameters varied by up to a factor of 1.61 (20 ms) or 1.5 (200 ms) respectively.
Conclusions: In this study, the treated area of retina varied by up to a factor of 1.61 ²  = 2.59 for a given spot number. As clinical efficacy depends on the treated area, which is a function of lesion number by area per lesion, our results implicate poor control of the overall treatment effect if treatments are administered according to lesion number or spacing alone. Better ways of laser effect control should be sought.