Accuracy of Amsler Grid in Screening for Chloroquine Retinopathy
Apichat Suansilpong MD*, Somchai Uaratanawong MD**
Affiliation : * Department of Ophthalmology, BMA Medical College and Vajira Hospital, Bangkok, Thailand ** Department of Medicine, BMA Medical College and Vajira Hospital, Bangkok, Thailand
Objective : Determine the accuracy, sensitivity, and specificity of an Amsler grid testing for chloroquine retinopathy screening.
Material and Method: One hundred and forty patients who received chloroquine phosphate or hydroxychloroquine sulfate
and attended the rheumatology clinic of BMA Medical College and Vajira Hospital between March 2008 and May 2009 were
included. The patients underwent Amsler grid testing, which would be interpreted by a rheumatologist, for any evidence of
chloroquine retinopathy. The results from Amsler grid testing were then compared to the results from a Humphrey 10-2 fields
testing, which was subsequently performed by an experienced ophthalmologist and was used as a gold standard.
Results : Out of 140 patients, chloroquine retinopathy was evidenced in 11 patients (7.9%). Kappa value of the Amsler grid
testing interpreted by rheumatologist and the Humphrey 10-2 fields testing interpreted by ophthalmologist was 0.89. The
accuracy for screening chloroquine retinopathy by the Amsler grid testing was 98.6% (95% confidence interval [CI],
98.1-100.0%) with the sensitivity of 81.8% (95% CI,75.4-88.2%). The specificity and positive predictive value were 100.0%
while the negative predictive value was 98.4% (95% CI, 96.4-100.0%).
Conclusion : Amsler grid testing is an accurate screening test for chloroquine retinopathy with very high specificity. The test
could be achieved by a rheumatologist who could practically serve the patients in one visit at the rheumatology clinic.
Keywords : Chloroquine retinopathy, Screening, Amsler grid testing
