Purchase this article with an account.
Sergul Ayse Erzurum, Rui Wu, Nicholas A. Sala, Robert W. Arnold, David I. Silbert, John W Erickson, Susan A Cotter, Raymond T Kraker, Jonathan M Holmes; Parent Provided Photographs as an Outcome Measure for Childhood Chalazia. Invest. Ophthalmol. Vis. Sci. 2020;61(7):2148.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To determine whether smartphone photographs of children’s eyelids, taken by parents, could be used to determine whether chalazia were present or absent.
Sixty children age 6 months to <17 years with at least 1 new or recurrent chalazion measuring ≥ 2mm in diameter diagnosed on clinical exam were enrolled. The clinical investigator examined and documented presence or absence of chalazion in each of 4 eyelids of every child on clinical exam, which was considered the Gold standard. Parents were instructed to take 4 digital photos of their child’s face using a smartphone (1 with eyes closed, 1 with eyes looking straight ahead, 1 with eyes looking up, and 1 with both lower eyelids pulled down, if the child would allow). Additional photos could be taken at parental discretion. Parents uploaded the photos to the study website. Based on the photographs, a masked reader assessed whether or not each eyelid could be graded; and judged whether a chalazion was present on each eyelid. Measures of agreement between the clinical examiner and the masked reader were calculated as sensitivity and specificity with 95% confidence intervals, adjusting for the correlation between the eyelids from the same participant.
58 of 60 participants had all 4 required photos taken and uploaded (photos with both lower eyelids pulled down were unsuccessful for 2); 24 participants had additional optional photographs taken. Photos were judged by the masked reader to be of sufficient quality for grading of all 240 eyelids. Using only the 4 required photos for each child, of 85 eyelids for which the clinical investigator documented a chalazion, the masked reader graded a chalazion present in 66 (sensitivity = 72%, 95% confidence interval (CI): 61% to 81%). Of 155 eyelids for which the clinical investigator documented a chalazion was absent, the masked reader graded a chalazion absent in 149 (specificity = 96%, 95% CI: 92% to 98%). Using additional optional photographs, sensitivity increased to 82% (95% CI: 72% to 89%) while specificity was 95% (95% CI: 91% to 98%).
Remote masked assessment of eyelid photographs taken by parents may be useful as an outcome measure for future clinical trials evaluating treatments of childhood chalazia.
This is a 2020 ARVO Annual Meeting abstract.
This PDF is available to Subscribers Only