Marked variability amongst pediatric ophthalmologists in designating the diagnostic gaze position angles in ocular motility evaluations.

BACKGROUND

Management of strabismus relies on accurate evaluation of binocular alignment in standard gaze positions. In 1962, Stuart & Burian noted that, without adopting a standard routine, "measurements of various patients could not be compared, and there will be considerable difference in measurement from one examination to another and by different examiners" (1). Diagnostic position gaze angles are not routinely measured. Is this important?

SUBJECTS AND METHODS

Subjects were 82 volunteer experts recruited from attendees at the 1998 American Association for Pediatric Ophthalmology and Strabismus (AAPOS) scientific meeting. One author served as examinee for all testing. The actual head posture was measured for the expert designated primary position, right and left gazes, head tilts and up- and downgazes using the CROM device. The examinee fixated at a six meter distance accommodative target. Examiners were asked to mimic their office routine. Eighty-two subjects ranging from 29 to 69 years of age, consisting of 24 females and 58 males were recruited. Sixty-nine were pediatric ophthalmologists, 7 orthoptists, 4 international members and 2 members in training. Eight subjects also underwent re-testing. Years in practice averaged 11.5.

RESULTS

Range of head posture measurements: For "Horizontal Gaze": 10 to 50 degrees; For "Vertical Gaze": 4 to 58 degrees; For "Head Tilts": 20 to 50 degrees. There was no substantial difference between initial and repeat measurements.

CONCLUSION

There is a surprisingly high degree of variability amongst expert observers in defining standard gaze positions. These results may explain some of the inconsistent outcomes noted in the strabismus literature. The implication for transferring data from publication to practice and in designing multicentered protocols is concerning. Without defining and maintaining a standard for binocular alignment measurements, comparison between studies and examiners is not possible.