Abstract
Purpose: :
Patients with socially significant strabismus may suffer both visual and psychosocial consequences. The magnitude at which strabismus becomes socially significant is unclear. Previous publications have suggested exotropia (XT) is easier to detect than esotropia (ET), with critical magnitude of detection (>70% detection) described at approximately 8Δ for XT and 14.5Δ for ET. However, both these studies used only one, female, Caucasian model. This study will further examine the critical magnitude of detectable strabismus by lay observers in a more culturally diverse model set.
Methods: :
Twelve models of variable ethnicity (4 African American (AA), 4 Asian (AS) and 4 Caucasian (CA)) with equal number of female and male models within each race were digitally photographed in a range of horizontal gaze positions. Strabismus ranging from ET 21Δ to XT 21Δ was simulated using photo manipulation. Images were presented in a randomized order to 102 non-health care professional observers who were asked to decide whether strabismus was present or not. The magnitude of strabismus with a 70% positive response rate was defined as the critical magnitude.
Results: :
Overall, an increase in likelihood of strabismus detection was noted as the angle of strabismus increased. For all models combined, critical magnitude for detecting ET was 18.82Δ (95% C.I. (18.34Δ, 20.38Δ)) and XT was 16.63Δ (95% C.I. (15.46Δ, 17.26Δ)), indicating XT was slightly easier to detect than ET. Differences were noted among race. The critical magnitude for detecting ET for AS models was 14.25Δ (12.91, 15.60) compared to 20.88Δ (18.54, 25.32) for XT. AA and CA data showed an opposite trend with critical magnitude being 20.82Δ (19.26, 22.37) for ET and 16.29Δ (15.52, 17.25) for XT in AA models and 23.17Δ (20.43, 27.43) for ET and 13.53Δ (12.70, 14.64) for XT in CA models. The positive response rate at ortho position (false positive rate) was lowest for AA followed by CA with a higher response rate for AS models. Large variability in critical magnitude of detection was found among the 12 models implying no uniform detection angle may exist across individuals.
Conclusions: :
XT was slightly easier to detect than ET for all models combined. Race differences were observed with ET being easier to detect than XT in AS models and the opposite trend in AA and CA models. Large inter-model variability may suggest difficulty generalizing results of this and similar studies and raises the question whether guidelines for socially significant strabismus can be applied to an entire population.
Keywords: strabismus: diagnosis and detection