Chart-Based Visual Function Tests Limited in Detecting Intermediate AMD

Visual function (VF) tests for best-corrected visual acuity (BCVA), low luminance visual acuity (LLVA), and Moorfields Acuity Test (MAT), among others, were not able to distinguish between no age-related macular degeneration (AMD) and intermediate AMD (iAMD), according to a study published in JAMA Ophthalmology.

The researchers recruited from 20 clinical centers, and there were participants from 18 European clinical sites for the cross-sectional analysis. Patients with no AMD, early AMD, iAMD, and late AMD were recruited from 5 sites, and the remaining 13 sites focused on participants with iAMD only.

Disease classification was determined by Beckman classification based on multimodal imaging. Participants had to be aged 55 to 85 years; they were excluded if they had ocular conditions that would require surgical intervention, had known systemic illnesses, or had cognitive impairment, illiteracy, or did not speak the national language.

VF tests included in this study were the BCVA, the LLVA, the MAT, the Pelli-Robson Contract Sensitivity assessment (CS), and the International Reading Speed Test (IReST; small-print standardized [SPS] and large-print standardized [LPS]).

There were 301 participants included in this study, with a mean (SD) age of 71 (7) years; 62.1% were women. There were 34 (11.3%) participants with early AMD, 168 (55.8%) with iAMD, 43 (14.3%) with late AMD, and 56 (18.6%) with no AMD. Participants with no AMD were younger than those with iAMD (mean age, 68 [6] vs 71 [8] years), and participants with iAMD were younger those with late AMD (mean age, 71 [8] vs 75 [6] years).

There were 290 participants who attended both study visits. Their median (IQR) time between visits was 14 (12-18) days, and no disease progression events were observed in that period.

Intraclass correlation coefficients (ICCs) values for chart-based VF measures ranged between 0.88 for CS and 0.96 for BCVA, which indicated good to excellent repeatability. When calculated by disease group the results were slightly lower, with iAMD (0.73 for CS to 0.89 for LPS and SPS) and late AMD (0.79 for LLVA to 0.95 for LPS) having good reliability.

The limits of agreement (LoA) were tighter within the no AMD, early AMD, and iAMD groups vs the late AMD cohort. The late AMD cohort had the tightest LoA of the disease groups for the IReST measurements, although a quarter of the participants with late AMD could not read any words during either test.

There were 8 of 10 site groupings for BCVA, LLVA, and MAT; 7 of 10 for CS; 5 of 8 for SPS; and 7 of 8 for LPS that had good to excellent ICC values for BCVA, LLVA, and MAT. Corresponding LoAs were in line with the iAMD cohort.

All measures were great at distinguishing iAMD from late AMD (AUC, 0.92-0.99). Early AMD was not distinguishable from iAMD on all measures (AUC, 0.54-0.64). CS was best at distinguishing between no AMD and iAMD (AUC, 0.77). BCVA, LLVA, and MAT were adequate at discriminating in the same conditions (AUC, 0.69-0.71) where low-luminance deficit was not able to discriminate at all (AUC, 0.59).

IReST offered poor distinguishing ability between no AMD and iAMD (AUC, 0.57-0.61). The ability to discriminate between no AMD and iAMD was highest in the combination of CS and LLVA tests (AUC, 0.79; 95% CI, 0.73-0.86) but was only slightly better than CS alone in discriminating between the 2 groups.

There were some limitations to this study. Tests were given in a certain order so that consecutive tests did not have the same letter series, and technicians were not masked to the severity of a participant's disease. In addition, test selection to determine VF was limited to literature from no later than 2016.

The researchers concluded that the study presented the first evidence that BCVA, LLVA, MAT, CS, and IReST were repeatable in a multicenter, multiexaminer setting.

Reference

Dunbar HM, Behning C, Abdirahman A, et al. Repeatability and discriminatory power of chart-based visual function tests in individuals with age-related macular degeneration: A MACUSTAR study report. JAMA Ophthalmol. Published online June 23, 2022. doi:10.1001/jamaophthalmol.2022.2113