Web-based vs. conventional: a comprehensive analysis of visual acuity assessment using the PocDoc tool in a tertiary eye care centre

Abstract

Background: Visual acuity (VA) represents a fundamental measure of visual function. The significant prevalence of underdiagnosed ocular disorders underscores the importance of effective VA assessment. This study evaluates the efficacy of a web-based VA assessment tool ("PocDoc") versus conventional VA testing.

Methods: Prospective observational study including 353 participants recruited from various eye clinics in a tertiary referral centre. Age, diagnosis, and VA related information (i.e. VA measurements from PocDoc and conventional VA test [Snellen chart], test type, etc) were collected. Spearman's rank correlation, Intraclass Correlation, and Bland-Altman plot compared outcomes of both tests. One-way ANOVA and paired-T test were used to compare means.

Results: Most patients were males (59.2%) with a mean age of 52.2 ± 20.6 years. PocDoc had moderate positive correlation to conventional testing (rho = 0.50, p < 0.001). PocDoc led to higher logMAR scores compared to conventional testing (mean logMAR 0.19 and 0.13 respectively, p < 0.01). Moreover, PocDoc demonstrated a sensitivity of 82.8% and specificity of 79% for detecting visual impairment. The discrepancy between PocDoc and conventional VA testing increased with higher logMAR values, indicating greater inconsistency between the tests for patients with poorer VA. Age, test type, and disease type contributed to this variability.

Conclusions: The concordance between PocDoc and conventional testing for VA measurement across various ages and conditions makes it a suitable screening tool. Future technological inventions should consider age, test type, and disease type as critical factors related to the level of agreement and correlation between digital and conventional VA testing methods.

Fig. 1. PocDoc digital tool.

Instruction manual on how to carry out PocDoc VA test. Specific instructions for Snellen or LogMAR charts are under step 7; steps for Landolt C or Tumbling E are under step 8, steps for Tail the Dog are under step 9 (A). VA test options are shown below. Image showing calibration using a credit card on PocDoc (B). Users would overlay their credit card on the screen and move the blue toggle to adjust the size of the orange box. Example of a test performed in a patient evidencing how the results are generated and recorded (C). Icons credits to Flaticon.

Fig. 2. Sensitivity analysis of factors influencing LogMAR difference between PocDoc and Conventional VA tests.

This figure illustrates the variability in LogMAR differences discriminated by disease type (A1), test type (A2), and patient age group (A3). The multivariate plot, segmented by age group, reveals distinct preferences in test usage: the Tail of the Dog test predominates among patients under 19 years (B1), whereas the Landolt C test is primarily utilized in individuals aged 20–59 and those over 60 (B2), with the extent of LogMAR difference varying by disease. Additionally, the Snellen test (B3), employed across similar age demographics as the Landolt C, exhibits increased variability in LogMAR differences among cataract patients, highlighting the effect of test selection on visual acuity measurement across different age groups and conditions.

Fig. 3. A Bland-Altman plots and scatter plots for agreement analysis and factors influencing heteroscedasticity.

General Bland-Altman plots of logMAR difference (y-axis) against logMAR average (x-axis) (A1). Mean logMAR difference shown by the blue bar in the middle, which differs from 0.00 logMAR, the straight black line. 95% limits of agreement (LOA) are shown by the green and red bars. Bland-Altman plots of logMAR difference against logMAR average of Tail the Dog charts (A2), Landolt C charts (A3), and Snellen charts (A4). Bland-Altman plot showing the differences in LogMAR between the two measurement methods against their averages (B1, B2, and B3). In each image the data points are coloured based on different categories. B1—Age categories: 5–19 years (grey), 20–59 years (blue), and 60 onwards (yellow); (B2)—Test types: C = Landolt C, D = Tail the Dog, S: Snellen test; (B3)—Condition or disease (Nil, Myopia, AMD, Cataract, Glaucoma, Others, Uveitis). Scatter plot showing the relationship between the two measurement methods for LogMAR (B4, B5, and B6). In each plot the data points are differentiated by categories with distinct colours and regression lines. B4—Age categories; B5—Test Types; B6—Condition or disease.