Agreement between droopy eyelid diagnoses made by in-person examination compared to analysis of digital images

Abstract

Purpose : Digital external photography shows promisng role in oculo-telemedicine consultation. We performed a retrospective, observational study to assess the agreement between external images and conventional face-to-face consultation in arriving at clinical diagnoses and management of droopy eyelids.

Methods : 18 patients at a single institution were randomly selected from droopy eyelid patient database. All were verified to have a complete sequence of external photographic images (Figure 1). Images were taken with Canon 6D camera with 100mm r/2.8 macro lens by professional photographers at the time of consultation. Six graders (oculoplastic surgeons with varying practice experiences) were asked to identify the cause(s) of droopy eyelids (dermatochalasis, blepharoptosis, and/or brow ptosis) and recommend surgical managements from the images. Specificity, sensitivity, Kappa coefficients and 95% confidence intervals were calculated to assess agreement between each grader and the in-person examination results.

Results : Graders had low levels of agreement with the in-person examination when diagnosing dermatochalasis or blepharoptosis and deciding whether to recommend upper lid blepharoplasty and ptosis repair, whereas they did better when diagnosis brow ptosis and deciding whether to recommend a brow lift (Figure 2a). Several graders had very high sensitivity coupled with a very low specificity when diagnosing upper lid dermatochalasis and blepharoptosis and recommending blepharoplasty and/or ptosis repairs. On the other hand, diagnosing brow ptosis and recommending brow lifts showed the opposite coupling of low sensitivity and high specificity (Figure 2b). The graders required only 3 images (full face and close-up view of both eyes open in primary gaze and closed) to provide the diagnoses and felt that additional photos in the sequence did not change their diagnoses in most cases.

Conclusions : External ophthalmic photography in oculoplastics research and telemedicine consultation should be used with caution, given lack of agreement with the in-person examination in arriving at diagnoses and surgical management agreement shown in our study.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.

 

Figure 1. External photography sequence for droopy eyelid evaluation

 

Figure 2. a) Kappa coefficient plot; A kappa coefficient of 1 indicates perfect agreement, while 0 indicates a level of agreement expected by chance. b) Specificity and Sensitivity plot