Clinical Science
Effectiveness of myopia prediction model in screening children and teenager myopia
Shiming Li, Mingyang Ren, Sanguo Zhang, He Li, Luoru Liu, Ningli Wang
Published 2019-04-10
Cite as Chin J Exp Ophthalmol, 2019, 37(4): 269-273. DOI: 10.3760/cma.j.issn.2095-0160.2019.04.006
Abstract
ObjectiveTo compare the effects of different combinations of common ocular parameters on the prediction of myopia in children and teenager, with the results of cycloplegia autorefraction as the gold standard for the diagnosis of myopia.
MethodsA diagnostic trial was adopted.Total of 2 739 first grade students in primary school and 1 797 grade 7 students in junior high school from the "Anyang Childhood Eye Study" were included from 2011 to 2012.Cycloplegic autorefraction using 1% cyclopentolate was used to diagnose the myopia (spherical equivalent [SE] ≥-0.5 D). Uncorrected distant visual acuity (UCDVA), non-cycloplegic autorefraction, axial length (AL) and corneal power were combined randomly, and Logistic regression method was used to establish the prediction models on myopia and evaluate their sensitivity, specificity, positive predictive value and negative predictive value.This study followed the Declaration of Helsinki and this study protocol was approved by the Ethics Committee of Beijing Tongren Hospital.
ResultsWhen using a single index, the effect of UCDVA was the best, but the sensitivity was only 39.18% in the first grade of primary school students, and the positive predictive value was 78.35%.Some improvement could be obtained by using two indexes, the effect of UCDVA+ AL was the best, the sensitivity was 48.46%, and the positive predictive value was 78.79%, followed by UCDVA+ non-cycloplegic autorefraction, the sensitivity was 45.08% and the positive predictive value was 79.15%.Three indices could further improve the effect of prediction model.The effect of UCDVA+ AL+ corneal power was the best, the sensitivity was 54.45% and the positive predictive value was 85.45%, followed by UCDVA+ AL+ non-cycloplegic autorefraction, the sensitivity was 52.12% and the positive predictive value was 81.21%.The best results could be obtained by using all four indices, but the improvement was limited, the sensitivity was 55.26%, and the positive predictive value was 84.57%.In the grade 7 students, sensitivities of the various indices were generally higher (86.63% to 97.73%), but the specificity was decreased (3.62% to 90.52%).
ConclusionsAL can significantly improve the sensitivity of predicting myopia in children, it is non-invasive, simple and easy to perform, and it is an important indicator of eye development and myopia development in children.Therefore, we suggest that AL should be list as a routine indicator of myopia screening, as well as clinical diagnosis and treatment in children.
Key words:
Myopia/diagnosis; Prediction model; Axial length; Sensitivity; Screening (General survey); Children; Adolescents
Contributor Information
Shiming Li
Beijing Tongren Eye Center, Beijing Tongren Hospital, Beijing Ophthalmology &
Visual Science Key Laboratory, Beijing Institute of Ophthalmology, Capital Medical University, Beijing 100730, China
Mingyang Ren
College of Mathematics, Chinese Academy of Sciences University, Beijing 100083, China
Sanguo Zhang
College of Mathematics, Chinese Academy of Sciences University, Beijing 100083, China
He Li
Anyang Eye Hospital, Anyang 455000, China
Luoru Liu
Anyang Eye Hospital, Anyang 455000, China
Ningli Wang
Beijing Tongren Eye Center, Beijing Tongren Hospital, Beijing Ophthalmology &
Visual Science Key Laboratory, Beijing Institute of Ophthalmology, Capital Medical University, Beijing 100730, China