Validity of self-report in type 1 diabetic subjects for laser treatment of retinopathy.

Kleins Lab // Publications // Ronald Danis // Dec 01 2013

PubMed ID: 23890420

Author(s): Grassi MA, Sun W, Gangaputra S, Cleary PA, Hubbard L, Lachin JM, Gao X, Kiss S, Barkmeier AJ, Almony A, Davis M, Klein R, Danis RP; DCCT/EDIC Research Group. Validity of self-report in type 1 diabetic subjects for laser treatment of retinopathy. Ophthalmology. 2013 Dec;120(12):2580-2586. doi: 10.1016/j.ophtha.2013.06.002. Epub 2013 Jul 24. PMID 23890420

Journal: Ophthalmology, Volume 120, Issue 12, Dec 2013

PURPOSE This study sought to determine the validity of self-report of prior panretinal photocoagulation (PRP) and focal photocoagulation (FP) compared with fundus photography.

DESIGN Prospective cohort study.

PARTICIPANTS One thousand three hundred sixty-three type 1 diabetic subjects from the Epidemiology of Diabetes Interventions and Complications (EDIC) study, a subset of the 1441 subjects originally enrolled in the multicenter Diabetes Control and Complications Trial.

METHODS At each annual visit, subjects were asked by EDIC staff whether they had undergone PRP, FP, or both since the last completed annual clinic visit. Fundus photographs were collected from one quarter of the cohort each year and from the entire cohort at EDIC years 4 and 10. Photographs were graded for the presence and extent of PRP and FP. Seventeen years of subject reporting and photograph grading of PRP and FP were compared in EDIC subjects.

MAIN OUTCOME MEASURES The κ, sensitivity, specificity, and positive and negative predictive values were calculated for subject-reported PRP and FP. Factors influencing subject misreporting were investigated.

RESULTS For subject reporting, 1244 (96%) of 1296 subjects with gradable photographs accurately reported whether they had a history of PRP in one or both eyes, and 1259 (97.5%) of 1291 with valid photographs correctly reported their history of FP. For PRP and FP, sensitivities were 90.4% and 74.0%, respectively; specificities were 96.0% and 98.8%, respectively; positive predictive values were 75.9% and 80.3%, respectively; negative predictive values were 98.9% and 98.4%, respectively; and κ values were 0.80 and 0.76, respectively. Risk factors associated with misreporting included prior laser for diabetic retinopathy and prior ocular surgery (each P<0.04).

CONCLUSIONS For subjects with type 1 diabetes, in the absence of a clinical examination or fundus photographs, subject self-report could be a reliable tool in a well-monitored study for assessing laser treatment type in diabetic retinopathy.

Copyright © 2013 American Academy of Ophthalmology. Published by Elsevier Inc. All rights reserved.