Recent peer-reviewed research published in the British Medical Journal—and announced by global nonprofit Orbis International—examined the use of an online training course to educate non-ophthalmic healthcare workers in detecting glaucoma.
Let’s start with some background.
As the leading cause of irreversible blindness across the globe, glaucoma has led to a 62% increase in blind or visually impaired patients between 1990 and 2010. Per research, the majority of these cases have occurred in low- to middle-income countries (LMIC).
As such, previous research has found the rate of undiagnosed glaucoma in 12 different countries observed on several continents was an estimated 81.87% (Africa, 92.4%; Asia, 76.75; Australia, 81%; Europe, 71%))—making screening, a form of secondary prevention of any disease, all the more critical for these patients.
I’m sensing a “but”…
Yup. Unlike diabetic retinopathy (DR) screening, which has become increasingly available in LMIC, glaucoma screenings have yet to be deemed cost effective and widely accessible in these countries.
So what might the solution be?
Investigators hypothesized that developing a screening protocol for both DR and glaucoma—two of the major blinding ophthalmic diseases—leveraging the same equipment and personnel could potentially result in a more cost effective approach.
And their goal?
To develop and test an online course to train nurses, technicians and non-ophthalmic physicians to screen for glaucoma..
Now talk about the program.
Funded by the eXcellence in Ophthalmology Vision Awards (XOVA), a Novartis independent grant program—the uncontrolled, experimental, before-and-after study enrolled 43 non-ophthalmic DR graders who previously trained for DR screenings through a project supported by Orbis and the Fred Hollow Foundation (FHF) in Vietnam..
Each grader underwent an online training course designed specifically to focus on a standardized approach to recognize glaucomatous damage to the optic nerve.
Break it down for me.
- DR graders underwent testing on a standard image set at baseline.
- Participants completed a self-paced, online training course with practice questions at the end of each course module (those who failed were asked to review the course again).
- Participants then repeated the test on the standard image set with identical images to those at baseline (but presented in a random order).
- Test scores were compared with baseline and to scores of 29 Vietnamese ophthalmologists also partaking in the same program (but who did not take the optic nerve grading course).
- Each participant received 20- to 30-minute, one-on-one training conducted by a glaucoma specialist.
- Following this, they were tested (for a third time) on the standard image set.
What kind of images were these?
All 60 digital photographic images of optic nerves—with 27 from healthy eyes and 33 from glaucoma eyes—were fully anonymized, de-identified, and of gradable quality based on glaucoma specialists’ feedback. The images were taken from three sources:
- The Orbis Vietnam program (19)
- Northern Ireland Cohort for the Longitudinal Study of Ageing (NICOLA) study (11)
- Glaucomatous Optic Neuropathy Evaluation (GONE) (23)
And the test itself?
The three-module, 5-hour course was conducted via Orbis Cybersight, a telemedicine and learning platform, with 217 Vietnamese-language slides (with voiceover narration in Vietnamese) that described a standardized, step-by-step process for examining healthy and glaucomatous optic nerve heads.
- Module 1→ Intro to glaucoma
- Module 2 → Intro to normal optic nerve head and systemic descriptions of features
- Module 3 → Systemic method for assessing optic nerve head photos to screen for glaucomatous features
Findings?
All study participants completed both pre- and post-training tests, as well as the one-on-one training.
Compared to the 29 local ophthalmologists who also took the training test, the investigators noted no statistically significant difference between any of the following factors:
- Age
- Sex distribution
- Work experience
How did the DR graders perform?
Per the data, the participants’ mean scores significantly improved for both pre-training (33.3%±14.3%) and post-training (55.8%±12.6%), p<0.001.
Compared to the ophthalmologists’ scores, there was no significant difference (58.7%±15.4%, p=0.384].
According to the investigators, “these findings… were consistent across optic nerve images from all sources.”
And overall?
Per the study authors, the data suggests, “that it is possible to train non-ophthalmic DR graders to screen accurately for glaucoma using optic nerve photos using a short, self-paced online course.”
They further stated that the addition of expert-led one-on-one training didn't result in any further significant improvement of grading performance—”suggesting encouragingly that results obtained with a convenient and scalable tool like our course approach is what may be readily achievable with such cadres.”
Conclusion.
The authors concluded the ability to train non-ophthalmic DR grades to screen for glaucoma (using optic nerve images) can be scaled quickly and at a low cost—potentially leading to an improvement in “glaucoma detection and ultimately reduce the prevalence of blindness from glaucoma in low-resource settings,” they wrote.