GLAUCOMA SCREENING USING AN IPAD-BASED VISUAL FIELD TEST IN A WEST AFRICAN POPULATION.

OBJECTIVE To investigate the performance of the Melbourne Rapid Field (MRF) for use in clinic-based visual field testing in a low-resource setting. DESIGN Prospective cross-sectional study. PARTICIPANTS One-hundred and three subjects (66 glaucoma, 37 controls) attending their clinical appointment at the Tema Eye Center, Tema, Ghana. METHODS Glaucoma patients and normal controls were subjected to MRF and Humphrey Field Analyzer (HFA). MAIN OUTCOME MEASURES Mean deviation (MD), pattern standard deviation (PSD), reliability parameters, sensitivity, specificity, and area under the receiver operating characteristic curves (ROC). RESULTS Mean MD was less negative and mean PSD more positive on MRF than HFA in both groups (all p< 0.001). False positive and false negative rates were comparable between methods (p=0.09 and 0.35, respectively). In glaucoma patients, MD and PSD from the two devices were strongly (r=0.84, p<0.001) and moderately (r=0.61, p<0.001) correlated, respectively. Agreement analysis revealed that MRF tended to generate significantly higher MD (bias: 3.3±4.1 dB, p=0.03) and PSD (bias: 1.9 dB±2.8 dB; p=0.03) with wide limits of agreements. For detecting moderate to advanced glaucoma the sensitivity was 60.9% for MRF and 78.3% for HVF (p = 0.10); respective specificities were 86.5% and 83.8% (p = 0.76). CONCLUSIONS MRF underestimated MD and overestimated PSD values compared to HFA. Agreement biases were significant, suggesting a weak agreement between the two devices. However, MRF shows potential for screening in a low-resource setting, particularly for detecting moderate to advanced glaucoma.