OBJECTIVES: The prevalence and necessity for early detection of vision problems illustrate the need for improved methods of vision screening in preschool children. This study assessed the validity and reliability of a new device, the MTI photoscreener in a cross-sectional field study. METHODS: An appropriate sample size (> 140) was calculated and recruited for the study. All children (N = 161) in a migrant workers summer education program were screened with the MTI Photoscreener. Simultaneously and in a masked design, disease status was determined by the Modified Clinic Technique, a well established method for diagnosing the conditions which the MTI screener was designed to detect. RESULTS: Validity measures revealed a sensitivity of 54%, specificity of 87%, phi coefficient of 0.40, and positive predictive value of 52%. Repeatability was assessed by the kappa coefficient, by a test for effect modification by examiner, and by comparison of sensitivity and specificity across 12 masked examiners. The kappa coefficient was 0.38. A test for effect modification suggested that differences existed among the examiners. Variability of sensitivity was high, but variability of specificity was low. CONCLUSIONS: Methods for vision screening in preschool children are limited. The MTI Photoscreener is an easy and efficient method, but the validity and reliability is a concern. Comparison of our results with other studies suggests future potential for this instrument provided protocols are refined and further field studies reveal efficacy.
OBJECTIVES: The prevalence and necessity for early detection of vision problems illustrate the need for improved methods of vision screening in preschool children. This study assessed the validity and reliability of a new device, the MTI photoscreener in a cross-sectional field study. METHODS: An appropriate sample size (> 140) was calculated and recruited for the study. All children (N = 161) in a migrant workers summer education program were screened with the MTI Photoscreener. Simultaneously and in a masked design, disease status was determined by the Modified Clinic Technique, a well established method for diagnosing the conditions which the MTI screener was designed to detect. RESULTS: Validity measures revealed a sensitivity of 54%, specificity of 87%, phi coefficient of 0.40, and positive predictive value of 52%. Repeatability was assessed by the kappa coefficient, by a test for effect modification by examiner, and by comparison of sensitivity and specificity across 12 masked examiners. The kappa coefficient was 0.38. A test for effect modification suggested that differences existed among the examiners. Variability of sensitivity was high, but variability of specificity was low. CONCLUSIONS: Methods for vision screening in preschool children are limited. The MTI Photoscreener is an easy and efficient method, but the validity and reliability is a concern. Comparison of our results with other studies suggests future potential for this instrument provided protocols are refined and further field studies reveal efficacy.