Vison

Week in Review: Hospitalizations for Syphilitic Uveitis, Online Training to Detect Glaucoma, AI and Japanese Ophthalmology Exams





Comprehensive Ophthalmology, Glaucoma, Uveitis

A weekly roundup of ophthalmic news from around the web.

In the United States, the rate of hospitalizations for syphilitic uveitis is rising. Syphilis cases have been increasing every year since 2001. Since 2010, the same trends have been noted for syphilitic uveitis, based on the number of hospitalizations for the disease, according to researchers who conducted a cross-sectional study of data from a sample of 444,674 patients included in the Nationwide Inpatient Sample. From 2010 to 2019, there were 5581 hospital admissions related to syphilitic uveitis; adjusting for inflation, this represented $385 million in hospitalization charges. The national incidence rate increased from a low of 0.08 per 100,000 population in 2011 to 0.23 per 100,000 population in 2019. Black patients represented 32% of syphilitic uveitis–related hospital admissions, a disproportionately high number. The authors note that the increases seen emphasize “the need for ophthalmologists to maintain a high index of suspicion for syphilitic uveitis when evaluating patients with intraocular inflammation.” JAMA Ophthalmology

Online training could help non-ophthalmologists spot glaucoma in underserved populations. In this study conducted in Vietnam, an online course was used to train 43 non-ophthalmologists (e.g., nurses) already trained as diabetic retinopathy (DR) graders on how to spot glaucomatous optic nerves using ophthalmic images. The course consisted of 3 modules that provided step-by-step instructions on identifying both healthy and glaucomatous optic nerves, using 60 separate images. Performance scores significantly improved following completion of the course, and no significant differences were found between scores from local ophthalmologists (who had not taken the course) and post-training scores from the non-ophthalmologists. In areas where there are few ophthalmologists, this online course could help local health care workers improve identification of glaucoma in the patients they serve. BMJ Open

ChatGPT is found to correctly answer nearly half of Japanese-language ophthalmology exam questions. That was the result of an investigation comparing the performance of 2 versions of ChatGPT against examinees’ answers to board examination questions from the Japanese Society of Ophthalmology. On average, examinees answered 66% of questions correctly; with zero-shot prompting, ChatGPT-3.5 and -4 answered 22% and 46% of questions correctly, respectively. While ChatGPT-4 had a much lower correct answer rate than human test-takers, its answer rate was close to that seen in earlier studies where it was applied to English-language ophthalmology exam questions. Cureus



Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button