Reliability of fixation preference testing in diagnosing amblyopia.

We evaluated the sensitivity and specificity of standard fixation preference testing and the 10-prism diopter (PD) fixation test in 427 patients. Standard fixation preference testing showed good sensitivity and specificity for than 10 PD. Patients with small angle tropias, however, had a high false-positive rate, with 40% of nonamblyopic patients inappropriately responding with strong fixation preference. The 10-PD fixation test reduced this high false-positive rate and showed only 1.5% false-positive results. Unfortunately, this shift to equal fixation preference was also seen in some patients with amblyopia, as five of 19 patients with three or more lines difference alternated fixation with the 10-PD fixation test. Retesting disclosed that, although these children would briefly alternate to the amblyopic eye while viewing a stationary target, they would not hold fixation with the amblyopic eye through smooth pursuit. Our conclusion is that standard fixation preference testing can be used to diagnose moderate to severe amblyopia in patients with tropias greater than 10 PD. Patients with small angle deviations or straight eyes should be examined with the 10-PD fixation test, with the criterion for equal vision being the ability to hold fixation well with either eye through smooth pursuit.