Modelling the results of the orthoptic treatment of divergence excess

Using data compiled from an earlier study of the results of the orthoptic treatment of divergence excess [Daum, K. M. Ophthal. Physiol. Opt. 4, 15-24 (1984)], a model has been presented regarding the results of the treatment. Provided the calibration sample is representative, the model may be used after an initial examination to predict the class of success a given subject will eventually be placed in as a result of orthoptic treatment. After treatment with a standard orthoptic regimen, the subjects were classified as being either totally successful (no objective or subjective visual difficulties) or partially/not successful. Stepwise discriminate analysis of 15 independent variables for the dependent variable success was used in order to choose the model with the highest coefficient of determination (r2) and the best statistical significance. The three variables chosen for the model were the angle of deviation at near and the blur and break values of the negative vergences at near. Using these variables, discriminant analysis was performed on the calibration sample of nine subjects. Each was classified correctly. The use and limitations of the model are briefly discussed.