RESUMO
PURPOSE: To assess the sensitivity and specificity of microperimetry as a screening test to detecting hydroxychloroquine retinopathy. METHODS: Retrospective cohort study. Patients with history of hydroxychloroquine use for more than 5 years and with concomitant microperimetry and multifocal electroretinogram testing were retrospectively reviewed. Microperimetry was considered positive if there were three or more contiguous scotoma points in the parafoveal region. Multifocal electroretinogram was used as gold standard and was considered positive if there was an increased R1/R2 ring ratio (>2.5) or reduced R1 absolute amplitude (<9.0). Sensitivity, specificity, positive predictive value, and negative predictive value of microperimetry were calculated. RESULTS: A total of 197 patients were reviewed. Hydroxychloroquine retinopathy was present in 22 (11%) patients. Their mean (SD) age was 54 (14) years, and 96% were women. Their mean (SD) daily dose was 5.7 (1.3) mg/kg, cumulative dose was 2041 (1,548) g, and duration of use was 15 (10) years. Sensitivity, specificity, positive predictive value, and negative predictive value of microperimetry were 73%, 93%, 53%, and 96%, respectively. CONCLUSION: Microperimetry has inferior sensitivity but good specificity in detecting hydroxychloroquine retinopathy (compared with multifocal electroretinogram). As such, it may be a useful ancillary test to exclude retinopathy, especially in high-risk patients or those with conflicting results on different modalities.