ABSTRACT
Purpose : Telemedicine allows physicians to provide care virtually, limiting patient exposure during the COVID-19 pandemic. As a follow up to a previous analysis (Gore R, et al. IOVS 2021;68:ARVO E- 146), we performed a retrospective chart review on a separate patient cohort to test whether virtual visual acuity (VA) tests are a reliable means of screening vision in a pediatric population. Methods : Families screened were sent instructions and an Early Treatment Diabetic Retinopathy Study (ETDRS) eye chart calibrated for 5 feet. Volunteer screeners instructed children over the phone, and those who failed the screen were referred to Wills Eye Hospital (WEH) for further evaluation. The institutional review board at WEH approved the current study. Demographic characteristics and VA screenings (virtual and in-person) were analyzed and compared using Pearson correlation coefficients. Bland-Altman plots were constructed to visualize this comparison. Results : A total of 242 eyes of 121 patients were included. The mean ± SD age was 11.7±3.1 years. Using Pearson correlations, the in-person and virtual VA measurements were moderately correlated in both the right eye (OD: r=0.506, p<0.001) and left eye (OS: r=0.610, p<0.001). Comparing in-person to virtual settings, the mean ± SD LogMAR of VA of the right eye (OD) was 0.37±0.36 (Snellen 20/46) vs. 0.55±0.30 (Snellen 20/70), respectively (p<0.001). The mean ± SD LogMAR of VA of the left eye (OS) was 0.38±0.43 (Snellen 20/47) in-person vs. 0.52±0.30 (Snellen 20/66) in the virtual setting (p<0.001). However, when the VA range was classified based on the median of in-person VA (LogMAR 0.3, Snellen 20/40), there was no significant difference between in-person and virtual VA measurements in VA >20/40 OD (0.70±0.30, 20/100 vs. 0.67±0.29, 20/93, p=0.538) and OS (0.76±0.45, 20/115 vs. 0.70±0.29, 20/100, p=0.274) (Table 1). Compared to in-person VA measurement (gold standard), the Bland-Altman plot shows a mean difference of 0.17 LogMAR with virtual VA measurements OD (Fig 1A) and a mean difference of 0.14 LogMAR with virtual VA measurements OS (Fig 1B). Conclusions : An ETDRS VA test delivered at home demonstrated good correlation with a standard ophthalmologist-administered VA test in a pediatric population. Virtual screening may underestimate vision in patients with good vision at the in-person visit, but the virtual screening does not fail to detect poor vision. (Figure Presented).