Statistical validation of a strabismus index calculated from objective ocular alignment data

Statistical analysis of a strabismus index used for objective interpretation of eye alignment data and improvement of this strabismus index.

Primary position eye alignment data of 330 subjects aged from a few months to young adulthood were collected with Purkinje Reflection Pattern Evaluation. Among these there were 95 strabismic subjects with inapparent, very small ocular misalignments from o° to about ±5°. The subjects were split into four age groups: up to 12 months of age, between 12 and 24 months, between 24 and 48 months and above 48 months of age. The measured and calculated angular variables and the strabismus index were tested statistically for age dependence and differences between the three orthoptic groups 'strabismus', 'referral' and 'non-referral'. The level of significance chosen was $aL=0.05.

There was a significant dependence of most of the angular variables on the orthoptic group. A dependence of the variables on age was not found except for a few angular variables which seemed to be related to the somewhat larger fluctuations of fixation in the two lower age groups. There was no age dependence of the strabismus index. With a strabismus index threshold of 0.7, the sensitivity was 85% for the detection of cases who needed referral according to the orthoptic examination.

This study provides the statistical basis for the strabismus index procedure which was initially introduced as an empirical procedure. It indicates that the strabismus index does not depend on age, so that limited cooperation in toddlers and infants has no major influence on the result of the data classification with the strabismus index. The strabismus index is a suitable means for objective, examiner-independent decision making in screening for very small ocular misalignment.