advertisement

---

Editors Selection IGR 7-3

Epidemiology

Comment by Roy Wilson on:

13017 Prevalence of open-angle glaucoma in a rural south Indian population, Vijaya L; George R; Paul PG et al., Investigative Ophthalmology and Visual Science, 2005; 46: 4461-4467

Find related abstracts

---

Several population-based prevalence studies of primary open-angle glaucoma (POAG) have been performed in India over the past decade. This most recent study by Vijaya et al. (744) in a rural population in southern India found a prevalence of POAG (1.6%) that was intermediate between that found in another rural population in Southern India in the Aravind Comprehensive Eye Survey (1.2%) and in an urban population in the Andhra Pradesh Eye Diseases Study (2.6%). Vijaya et al. suggest that the differences in prevalence estimates between these three studies may be due to the different definitions used to diagnose glaucoma. Although that may certainly be one explanation, there may be other issues to consider.

In a rural Indian population, almost everyone (98.4%) was previously undiagnosed. This finding suggests that efforts to screen for advanced glaucoma in this population may be particularly rewarding

The authors very nicely discuss several issues that may have lead to either an under or over-diagnosis of POAG in their study. In addition to these, there are at least two methodological issues that should be considered as potential sources of a biased estimate. First, this study defined POAG according to a classification suggested by the International Society of Geographical and Epidemiological Ophthalmology which is based on both structural and functional criteria. Three categories for glaucoma diagnosis were used: category 1 required both structural and functional evidence, category 2 required only structural damage, and category 3 was based on IOP only. Each category was stringently defined. However, a potential for under-diagnosis exists with subjects who may have had evidence of functional damage, but whose optic disc appearance did not satisfy the definition of structural damage as stipulated for a category 1 diagnosis (cup to disc ratio ≥97.5^th percentile of the normal population). This reliance on cup to disc ratio for defining glaucoma is problematic since cup to disc ratios are influenced by disc size which is not considered.

The distributions for cup to disc ratio and for IOP in the normal population were derived from a subset of 1818 subjects (study population was 3924 subjects) in the study with normal and reliable suprathreshold visual field tests by frequency doubling perimetry. This methodology is unlikely to have yielded a truly normal population distribution. This is problematic since the cup to disc ratios and IOPs used for the three diagnostic categories were based on percentile deviations from normality as determined by this unorthodox method.

The first issue, i.e., subjects with functional loss but with cup to disc ratio less than 97.5th percentile of normal population, wouldhave resulted in an under-estimate of true glaucoma prevalence. The second issue, i.e., normal population distribution defined by a subset of super-normal subjects within the same study, could have resulted in an overly conservative distribution of parameters used in diagnosing glaucoma and an over-estimate of true glaucoma prevalence.

Notwithstanding the potential biases discussed above, this is a very good population-based study and important information is reported. Of interest is, that many population-based studies in other populations have reported that up to half of the study subjects diagnosed with glaucoma did not know that they had the disease. In this rural Indian population, almost everyone (98.4%) was previously undiagnosed. This finding suggests that efforts to screen for advanced glaucoma in this population may be particularly rewarding. Additionally, this study is part of a larger study that includes an urban population. It will be interesting to compare the prevalence in this rural population with the glaucoma prevalence in the urban population as derived within the same study using an identical methodology. Finally, since gonioscopy was performed, it is hoped that the proportion of glaucoma that was of the angle-closure variety will be reported in subsequent publications.

---

• ← Previous
• Next →

---