Abstract
Purpose:
To evaluate the quality of photographs taken in a consult setting and to analyze whether remote readers accurately assessed the need for an ophthalmology referral and the ophthalmic diagnosis compared to a gold-standard clinical examination based on 1) photographs obtained with a hand-held portable camera alone and 2) photographs plus the patient’s chief complaint (CC).
Methods:
We performed a prospective observational study of all patients that had ophthalmology consult requests over a two-week period in a tertiary care center. Anterior segment (AS) and posterior segment (PS) photographs from a portable camera (PictorPlus, Volk Optical, Cleveland, OH) were obtained. The CC and gold-standard clinical ophthalmic examination were recorded and compared to the assessment by two remote ophthalmologist-readers. We analyzed the quality of the photographs. We compared each readers’ assessment (diagnosis and decision to refer for ophthalmic consultation) to each other and to the gold standard clinical exam. Kappa values were calculated for inter-rater agreement.
Results:
Thirty-four patients were included in the study (24 eyes imaged for AS, 39 for PS). Only 6% of AS and 22% of PS images were not gradable. Readers agreed on image quality (p<0.05). For gradable AS images, readers agreed with the gold standard diagnosis 81-86% of the time without and 80-85% with the CC. For gradable PS images, readers agreed with the gold standard diagnosis 72-81% of the time without and 73-91% with the CC. The two readers agreed on the presence of pathology and need for referral for 86% of AS patients (κ=0.72, p=0.0003) and 82% of PS patients (κ=0.65, p<0.0001) without the CC. With the addition of the CC, reader agreed on the need for referral for 73% of AS patients (p=0.04) but was not significant for PS patients (46%, p=0.5).
Conclusions:
Portable imaging technology can produce high quality AS and PS photos, though more PS images were not gradable. The telemedicine reader correctly assessed the clinical picture based on limited information the majority of the time. The reader’s decision-making may be more reliable for AS versus PS pathology; this will need to be explored further with a larger sample size. In the future, additional information such as visual acuity, subjective ophthalmic complaints, and the general condition of the patient may improve the accuracy of telemedicine readers’ assessment and decision to refer for further in-person evaluation.