Abstract

A short-cut systematic review was conducted using a described protocol. The three-part question addressed was: In adult patients presenting to the ED with diabetes-related visual symptoms, how effective is using a portable handheld fundus camera in diagnosing diabetic retinopathy?

MEDLINE, Embase and Cochrane databases were searched for relevant evidence. Altogether, 237 papers were found using the search strategy developed. 12 provided the best evidence to answer the three-part question. The data on first author name, publication year, country of origin, study type, study sample size, participant’s gender, reported effect sizes, main findings and limitations were extracted from the relevant studies and listed in a table.

Following a thorough examination and review of the literature, our analysis identified 12 articles for detailed evaluation. Of these, three provided the most compelling evidence concerning the use of portable handheld fundus cameras for the diagnosis of diabetic retinopathy in emergency settings. Ruan et al (2022) reported superior image quality and a sensitivity of 82.1% (95% CI: 72.1% to 92.2%) with a specificity of 97.4% (95% CI: 95.4% to 99.5%) for a handheld camera combined with artificial intelligence interpretation. Jin et al (2017) demonstrated high-quality images with 63% rated as excellent, showing a comparable efficacy to a traditional tabletop camera. Das et al (2022) found that Remidio and Pictor handheld cameras had high success rates and image quality, with sensitivities of 77.5% (95% CI: 65.9% to 89.0%) and 78.1% (95% CI: 66.6% to 89.5%), respectively, comparable to the Zeiss tabletop camera’s sensitivity of 84.9% (95% CI: 78.2% to 91.5%). The clinical bottom line is that the best available evidence supports the effectiveness of portable handheld fundus cameras for diagnosing diabetic retinopathy in emergency settings.