Diagnostic accuracy of online visual acuity testing of paediatric patients

Abstract Background/Objectives: Remote assessment of children’s visual acuity became necessary during the COVID19 pandemic. We aimed to assess the extent of agreement between hospital-based clinical testing and clinician-led home-based testing.Subjects/Methods: 50 children aged 2–16 years attending hospital eye services at two UK hospitals had routine hospital-based acuities compared with subsequent online, orthoptist-supervised home visual acuities. Agreement was assessed using Bland-Altman plots, as was test-retest (TRT) agreement of two, repeated home acuity tests.Results: Monocular acuities tested at hospital and at home were obtained from all 50 children; 33 also had binocular acuities in both settings and 35 had acuities retested immediately at home. Most children were tested at home using a computer or tablet: two were tested with a smartphone. No mean test differences were found for hospital vs home testing, and limits of agreement (LOAs) were ± 0.32 and ± 0.35 logMAR for binocular and monocular testing respectively. LOAs for inter-ocular acuity differences were -0.15–0.25 logMAR. TRT agreement was excellent, with LOA of ± 0.14 logMAR.Conclusions: We found excellent agreement between hospital and home acuity testing. LOAs were in keeping with multiple changes between measures (test; setting; time; tester) and a cohort including patients as young as two years old. Even smartphone testing proved feasible. Inability of the supervising orthoptist to check test distance or device calibration/orientation were limitations likely contributing to breadth of LOAs. Home vision testing is feasible and accurate, but its precision, acceptability, health economic impact and carbon impact require more attention.