Evaluating retinal thickness classification in children: A comparison between pediatric and adult optical coherence
Tian Siew Pua1, Mohd Izzuddin Hairol1
1Centre for Community Health Studies (ReaCH), Faculty of Health Sciences, Universiti Kebangsaan Malaysia, Kuala Lumpur, Malaysia.
Insights
Pediatric retinal thickness measurements show moderate agreement with adult data, with no significant age correlation but a positive association with spherical equivalent in children.
Area of Science:
- Ophthalmology
- Medical Imaging
- Pediatric Eye Care
Background:
- Accurate retinal thickness measurements are crucial for diagnosing pediatric eye conditions.
- Existing optical coherence tomography (OCT) devices often use adult reference databases, potentially impacting pediatric diagnostic accuracy.
Purpose of the Study:
- To evaluate the agreement between pediatric retinal thickness classifications using adult reference data versus a new pediatric database.
- To assess the correlation of retinal thickness with age and spherical equivalent (SE) in children.
Main Methods:
- A cross-sectional study included 160 healthy children (aged 6-18 years).
- Peripapillary retinal nerve fibre layer (pRNFL) and macular thickness were measured using Topcon 3D OCT-1.
- Agreement was assessed using Cohen's κ, comparing pediatric data against adult reference data.
Main Results:
- Overall agreement for pRNFL was ≥90% and for macula was >72% when compared to adult data.
- Significant agreement (κ > 0.6) was found for temporal RNFL, macular perifoveal regions, and inferior ganglion cell + inner plexiform layer (GCIPL).
- Retinal thickness showed no significant correlation with age but a positive association with spherical equivalent (SE).
Conclusions:
- Adult reference data provides 72-90% agreement for pediatric pRNFL and macular thickness measurements.
- Age is not a significant factor, but spherical equivalent is positively associated with retinal thickness in children.
Purpose:
This study investigates the agreement of children's retinal thickness classification by color category between Topcon 3D OCT-1's built-in adult reference data and our new pediatric database and assesses the correlation of retinal thickness with age and spherical equivalent (SE).
Methods:
160 eyes of 160 healthy children (74 boys, 86 girls) aged 6-18 years (mean: 11.60 ± 3.28 years) were evaluated in this cross-sectional study. The peripapillary retinal nerve fibre layer (pRNFL) and macular thickness were determined for the 1st, 5th, 95th, and 99th percentile points. Cohen's κ value and specific agreement between pediatric data and adult reference database were estimated. The correlation between retinal thickness with age and SE was also determined.
Results:
The mean thickness for the total RNFL, average macular, and central macula were 112.05±8.65 μm, 280.24±12.46 μm, and 220.55±17.53 μm, respectively. The overall agreement between the classification of the adult database and pediatric data for pRNFL was ≥90%, with discrepancies in 46 out of 150 eyes (30.67%); for macula, it was above 72%, with discrepancies in 93 out of 153 eyes (60.78%); and for ganglion cell complex and ganglion cell + inner plexiform layer (GCIPL) the agreement was above 84% and 85%, respectively. A significant level of agreement between pediatric data and adult reference data was achieved for temporal RNFL (κ = 0.65), macular perifoveal superior (κ = 0.67), and inferior (κ = 0.63) and inferior GCIPL (κ = 0.67). The correlations between age and retinal thickness were not significant (all p>0.05). Most retinal thickness parameters were positively associated with SE (Pearson's coefficient, r = 0.26 to 0.49, all p<0.05).
Conclusions:
The overall agreement for pRNFL and macular thickness measurements in children with the adult reference database was between 72% and 90%. Children's retinal thickness was not significantly correlated with age but was positively associated with spherical equivalent.
