In this prospective observational study, West et al compared the performance of the 10-2 and 24-2 visual field patterns to detect macular damage in a population with early glaucomatous visual field loss. They included 97 glaucoma patients with average 24-2 mean deviation (MD) of -2.3 dB and 65 healthy participants. When looking at different pointwise criteria to define abnormalities within the central 10° of the visual field, they found no significant difference between the central 10° of the 24-2 and the 10-2 regarding their ability to detect central field abnormalities. Although there was a non-significant tendency for better performance of the 10-2 compared to the 24-2 (as measured with area under the curve of the receiver operating characteristics curves (AUC-ROC)), at matched 5% specificity there was a non-significant trend for better performance of the central 10° of the 24-2 pattern. The authors concluded that, based on their data, there is little evidence for adding the 10-2 visual field test when searching for central visual field loss in early glaucoma and it may be more prudent to use the 10-2 to follow patients with higher risk of central visual field progression.
While there is no arguing with their data per se, there are a number of reasons to question their conclusions. First, glaucomatous functional damage was defined based upon the presence of 24-2 visual field loss in glaucomatous eyes and its absence in healthy controls. In any study that aims to compare two different diagnostic techniques (say A vs. B), one should not employ one of them (say, A) as the reference standard. By doing so, one will inevitably favor the performance of that technique over that to which it is being compared. By requiring the 24-2 to be abnormal in glaucomatous eyes and normal in the healthy group, the study inserted bias favoring the performance of the 24-2. Second, given the much smaller number of points within the central 10° of the 24-2 (12 points) compared to the 10-2 (68 points), the comparison of sensitivity at matched specificity between the two tests is flawed from a clinical standpoint. In Figure 6, for instance, one notes that a single abnormal point on the 24-2 yielded similar to higher sensitivity with 24-2 than the 10-2, which required 6 abnormal points. That one point on the 24-2 has little to no value clinically unless it belongs to a cluster coming from outside the central 10°. In practice, most clinicians would probably consider a single abnormal point in the central 10° of the 24-2 (in the absence of a cluster coming outside the 10°) as the result of test variability and would likely ask the patient to repeat the test next visit to confirm the abnormality. Yet, 6 or so points on the 10-2 may constitute a cluster more likely to be real and to agree with OCT results, increasing the level of certainty and decreasing the need and burden of repeat testing. Third, there was no reference to any structural tests (namely OCT of the macula) to confirm whether the abnormalities seen on the 24-2 or 10-2 were real and not just false-positive results. It is therefore very likely that in many of the eyes in which the 24-2 showed abnormalities not shown on the 10-2 may in reality constitute false positive results of macular damage, especially if a single point was present. Finally, and most importantly, the study's main conclusion contradicts some of the data presented. Despite the large overlap between the two modalities regarding the presence of central loss, in reality the study results favor following all eyes with central loss with 10-2 tests. In other words, it is not only eyes "with higher risk of central visual field progression" who will benefit from the 10-2 over time but every patient with central loss - whether or not it was also seen on the 24-2. This is obvious given the better sampling of the central field with the 10-2 pattern, which allows for further and better monitoring of progressive changes as opposed to only a few points in the central 10° of the 24-2.
For those patients with 24-2 detected loss not seen with 10-2, one should look closely for false-positives, preferably with the aid of OCT
In sum, for those patients with 24-2 detected loss not seen with 10-2, one should look closely for false-positives, preferably with the aid of OCT.1 Also, one should keep in mind that many of these cases may be due to clusters coming from the more peripheral region (not sampled by the 10-2, by definition) towards the central 10 (see their Supplemental Fig 11).
For those patients whose 10-2 detected loss not seen with 24-2, the benefit of 10-2 is clear and has implications not only for the detection of glaucomatous damage, but also staging its severity.2 Finally, for the majority of patients identified by West et al in whom the presence of damage overlapped between the 24-2 and 10-2, the value of the 10-2 to monitor progression is evident not only among those at high risk of progression. but for all those in whom central field loss has been detected or suspected.