Photographic and clinical techniques for outcome assessment in sixth nerve palsy

Purpose: There are few data on the reproducibility of clinical assessment of ductions and alignment. We evaluated photographic methods that may be used for masked outcome determination in a clinical trial and compared them with clinical measures. Design: Interexaminer reliability study. Participants: Twenty-three patients with unilateral sixth nerve palsy and three control participants were clinically evaluated by two masked examiners. Main Outcome Measures: Abduction deficit was graded as 0 to -5. Simultaneous prism and cover tests (SPCT) and alternate cover tests (ACT) were performed at distance and near fixation. Photographs were taken of abduction and distance alignment by each examiner. The photographs were evaluated by a third masked reader, who assigned abduction grade, measured absolute abduction (mm) and relative abduction (%), and calculated alignment in prism diopters (pd). Agreement was evaluated by calculating intraclass correlation coefficients (r_i), weighted kappa statistics (κ), and Spearman rank correlation coefficients (r_s). Results: There was excellent agreement between the two clinicians in clinical abduction deficit (κ = 0.86) SPCT and ACT at distance and near (r_i 0.94-0.96), between the clinical grade and masked photographic grade (κ = 0.83), and between the two sets of photographs for absolute abduction and relative abduction (r_i = 0.98 and 0.97). Both photographic measures of abduction correlated well with the clinical grade (r_s = -0.96 for each). Measurements of alignment from photographs correlated with clinical SPCT measurement (r_i = 0.88), but had a lower level of absolute agreement (38% within 5 pd) than between two independent SPCT measurements (96% within 5 pd). Conclusions: The excellent interexaminer agreement of our new photographic abduction assessment and of masked clinical measures suggest that these methods would be useful in clinical trials. In contrast, our simple method of photographic assessment of alignment lacks excellent agreement with the clinical assessment. These data are important in planning clinical trials in strabismus.