Evaluating photoscreeners II: MTI and Fortune video refractor

Background: Amblyopia is a public health problem, usually amenable to treat ment if detected early. Photoscreeners are camera-based instruments which c an detect risk factors for amblyopia such as squint, refractive errors and media opacities. Objective: To evaluate two commercially available photoscreeners, namely th e MTI photoscreener (Medical Technology, Iowa City, IA, USA) and the Fortun e Optical VRB-100 (Fortune Optical, Padova, Italy) videophotorefractor; in a selected childhood population, having a high prevalence of amblyopia, bef ore undertaking a full-scale trial in the general population. Subjects and Methods: The study design was a double-masked study, One-hundr ed and five children aged between 12 and 44 months with either normal visio n or known visual disorders were photoscreened without cycloplegia using th e Fortune and the MTI photoscreeners. Each child had a full ophthalmic exam ination either on the day of screening or in the preceding 6 months. Risk f actors were: any manifest strabismus (squint), hypermetropia >3.5 D, anisom etropia greater than or equal to 1 D sphere, myopia greater than or equal t o 2 D sphere, astigmatism greater than or equal to 2 D, media opacity or fu ndus abnormality affecting vision, The prevalence of risk factors for ambly opia was 60%. Photoscreen images were reviewed by two independent masked ob servers for indicators of amblyopiogenic risk factors and compared to the f ull ophthalmological examination to determine sensitivity and specificity f or each instrument. Results: Sensitivities and specificities for the detection of risk factors for amblyopia were as follows. Fortune photoscreener: reader 1,sensitivity 60%, specificity 75%; reader 2, sensitivity 68% specificity 86%. MTI photos creener: reader 1, sensitivity 56%, specificity 79%; reader 2, sensitivity 61%, sensitivity 86%. The results for reader 1 and 2 showed no significant differences for either instrument. Conclusion: Both instruments performed unsatisfactorily in a study populati on aged 1-4 years with a high prevalence of amblyopiogenic risk factors. Ac cordingly, we do not believe that either instrument can be recommended to s creen for eye disorders in children between the ages of 1 and 4 years becau se the low prevalence of amblyopia in this population demands very high sen sitivity and specificity in order to avoid inappropriate over- or under-ref erral.