In this population-based study of Chinese residents of Shanghai, 291 subjects were observed to have a peripheral ACD less than 0.5 CT in their right eyes. Among these subjects, 91 (31.27%, 91/291) subjects were diagnosed as having PACD. In this study, digital remote screening and clinical re-examination for a diagnosis were used at a tertiary eye hospital. Information on all subjects was first collected and analyzed remotely, including visual acuity examination, refraction, IOP measurement, digital anterior slit lamp eye structure photography and digital fundus photography. After this information transmitted to the Shanghai Eye Disease Prevention and Treatment Center, glaucoma suspects were asked permission to proceed with additional and confirmatory reexamination.[20]
In this study, the peripheral ACD in 2463 subjects were 0.94 ± 0.44 CT, and 11.8% subjects had reported a peripheral ACD < 0.5 CT. The population distribution of peripheral ACD has been studied in Europe, North America, and Asia.[22–24]
Based on peripheral anterior chamber depth, Van Herrick classification has been used to detect angle closure glaucoma (ACG) eyes.[25] This method has proved to be effective and suitable for glaucoma screening. However, Van Herrick’s classification only had five categories (0, 1/4, 1/2, 3/4, 1 CT). The Van Herrick’s classification has been suggested to have low sensitivity and specificity to detect patients with ACG.[11] Studies in southern India, and Greenland reported suboptimal performance of this test in screening for PAC.[26, 27] Therefore, in this study, a modified Van Herrick grading scheme was performed with eight categories (0, 0.05, 0.15, 0.25, 0.35, 0.45, 0.75, and ≥ 1.0 CT) instead of the usual five categories. These values were chosen to give class limits of 0, < 0.1, < 0.2, < 0.3, < 0.4, < 0.5, < 1.0, and ≥ 1.0 CT. This technique has been demonstrated to be valid in other studies.[28–30]
Our research shows that the peripheral ACD declines by 0.31 CT per diopter of SE and was 0.19 CT shallower in women than in men. Our findings agree with previous studies that women are more likely to have a shallower peripheral ACD.[31] Similar to other reports, our results are consistent with hyperopic patients likely to have shallower ACDs [32].
The present study used three different indexes (SE, peripheral ACD, SE combined with peripheral ACD) to investigate the most efficient way for PACD screening. The independent SE as the screening index for PACD was not statistically observed and did not significantly improve the screen capability of peripheral ACD for PACD, too. On the basis of these findings, only peripheral ACD measurement should be used to PACD screening in community elderly Chinese.
In our research, when peripheral ACD was < 0.3 CT, Youden index performed best in all groups and the positive predictive value was 0.703, however, a higher sensitivity is needed for effective community screening. The use of a CT cut off value of 0.4 would achieve much higher sensitivity (0.846). Our findings disagree with those from North America, Greenland, and Australia, which reported a cut-off value of 0.25 CT.[26, 28, 33] The reasons may be that the pathogenisis of angle closure is different in different population and the assessment method of peripheral ACD in this study is different from that in other studies. Similarly, Foster et al also reported the augmented Van Herrick scheme offers enhanced performance in detection of established PACG. [29]
A limitation of this study is that only subjects with peripheral ACD < 0.5 CT were asked to undergo gonioscopy, which may give rise to a potential cause of bias. Three major rationales were used for our screening method design. First, according to Van Herrick, a subject with a peripheral ACD ≥ 0.5 CT would have a very low probability of having angle closure.[25] Second, there was a study in Australia, in which goniscopy was carried out only with peripheral ACD ≤ 0.3 CT.[28] Third, Foster et al performed goniscopy on subjects and indicated that in subjects witth peripheral ACD ≥ 0.5 CT, no patients had closed angles. Therefore, a peripheral ACD < 0.5 CT is likely to be a relatively cost-effective index for glaucoma screening, although we do not recommend that only the estimation of peripheral ACD be used instead of gonioscopic examination in patients suspected of glaucoma.