Abstract
Purpose :
Telemedicine-enabled primary eye-care centres (Vision Centre) may be an important component of achieving universal eye care. Thus, we sought to evaluate the use of telemedicine-enabled Vision Centres (VCs) compared with the gold-standard, in-person examinations (IPEs) by an ophthalmologist.
Methods :
This prospective, diagnostic accuracy validation study was conducted at a VC belonging to Aravind Eye Hospital (AEH) in south India. We included consecutive new patients, aged ≥16 years, visiting the VC. All participants underwent the tele-ophthalmic examination (TOE), followed by IPE on the same visit. The in-person ophthalmologist was masked to the TOE diagnosis and treatment advice. The primary outcomes were percent agreement, kappa (k), sensitivity, and specificity of TOE versus gold standard, IPE.
Results :
The TOE had high percent agreement for all eye conditions ranging from 91.3% (allergic conjunctivitis) to 98.8% (injury, and pterygium). There was substantial agreement between the TOE and IPE for cataracts (k = 0.72; 95% CI: 0.63-0.82) and moderate to substantial agreement for injury (k = 0.55; 95% CI: 0.27 to 0.83), allergic conjunctivitis (k = 0.51; 95% CI: 0.41 to 0.62), infective conjunctivitis (k = 0.61; 95% CI: 0.43 to 0.78), and pterygium (k = 0.55; 95% CI: 0.27 to 0.83) and fair agreement for diabetic retinopathy (k = 0.34; 95% CI: 0.09 to 0.59), and slight agreement for glaucoma (k = 0.16; 95% CI: -0.13 to 0.45). Overall, the sensitivity of TOE ranged from 12.5% (glaucoma) to 91.7% (cataracts) and the specificity ranged from 93.5% (normal eyes) to 99.8% (injury). With respect to treatment advice (cataract surgery, spectacles and medications), there was substantial agreement (k≥0.67) with a sensitivity ranging from 58.1% to 77.2% and specificity ranging from 96.9% to 100%.
Conclusions :
The TOE in rural VCs has comparable accuracy to an IPE by an ophthalmologist for correctly identifying and treating major eye ailments. Through providing universal eye care to rural populations, this model may contribute to work toward achieving the U.N. Sustainable Development Goals.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.