Non Mydriatic Fundus Camera Shows Good Repeatability, Reproducibility

A non mydriatic digital fundus camera demonstrated repeatable and reproducible meibography results, showing its reliability for clinical use, according to a study published in Ophthalmic & Physiological Optics.

Two researchers assessed the interexaminer reproducibility of the Cobra® (Hanson Instruments) HD fundus camera by performing meibomian gland (MG) imaging on 74 patients (mean age, 23 years) on the same day. Among these participants, 72 individuals underwent retesting on subsequent days to assess intersession repeatability. Study participants completed the Ocular Surface Disease Index (OSDI) to determine the presence of dry eye symptoms (interexaminer assessment cohort, n=37; intersession repeatability cohort, n=36) and underwent MG imaging of both the top and bottom lids. The investigators measured the percentage and grade of MG loss according to the meiboscale grading system. 

The mean MG loss of the upper (session 1: 13.5±9.5%; session 2: 12.8±8.5%) and lower eyelids (session 1: 7.5±6.9%; session 2: 7.3±6.3%) was not significantly different between sessions, regardless of dry eye symptom status, the report shows. 

Interexaminer reproducibility was also good with mean MG loss of the upper (examiner 1: 12.7±8.2%, Examiner 2: 13.1±8.0%) and lower eyelids (examiner 1: 7.0±6.2%; examiner 2: 7.4±6.2%) not yielding significant differences between examiners. 

Within-participant variability was low with intraclass correlation coefficient (ICC) values greater than 0.89 observed for both eyelids, examiners, and sessions, according to the report.

“[T]he Cobra HD fundus camera meibographer demonstrates good repeatability and reproducibility, and clinically similar findings should be obtained when used by different examiners on different occasions,” according to the researchers. “Thus, it is suitable for the meibographic assessment and follow-up of disease progression or treatment outcomes.”

Study limitations include the potential for volunteer bias.