To assess the performance of the Odysight® self-monitoring tool for primary and secondary prevention of exudative maculopathies.

A one-year retrospective evaluation was conducted in a private practice. An alert was considered significant (SA) when patient was symptomatic, and non significant (NSA) when the patient was asymptomatic and the VA drop was refuted by an additional test carried out the next day.

Of the 92 patients who received a prescription, 44 (48%) downloaded and used the application and were analyzed. At the end of the year, there were 24 (55.8%) active patients. 29 alerts were generated, 13 were considered as NSA, and no anticipated appointment was given. The remaining 16 alerts were SA, 6 (21%) of them were true positive, i.e, revelealed a disease activity on OCT, leading to anticipated injection (representing 9.3% of the patients). 83% of the true positive alerts were generated by eyes with retinal vein occlusion (RVO). Eleven false negative events occurred (6 (55%) for AMD, 4 (36%) for RVO, 1 (9%) for myopic MNV). No primary exudation occured, although 7 false positive alerts were generated. The overall sensitivity of alerts for detecting OCT recurrence was 35.3% (95% CI [12.6; 58]), specificity was 93% (95% CI [90.3; 95.8]), positive predictive value was 20.7% (95% CI [6; 35.7]), and negative predictive value was 96.5% (95% CI [94.5; 98.6]). Sensitivity was higher for RVO (55.6% (95% CI [26.7; 81.1]) than for AMD (14.3% (95% CI [2.6; 51.3]).

In this real-life experimentation of Odysight, half of the patients used the application after initial prescription. Participation was low among diabetics. Odysight® allowed about 10% of the cohort to receive earlier intravitreal injection, albeit with a high number of false positives. Sensitivity was low in secondary prevention for AMD, and better for patients with RVO. Additional use in primary prevention may be questionable.