advertisement

Editors Selection IGR 15-2

Comments

Comment by Tony Realini on:

53737 Do findings on routine examination identify patients at risk for primary open-angle glaucoma? The rational clinical examination systematic review, Hollands H; Johnson D; Hollands S et al., JAMA (Journal of the American Medical Association), 2013; 309: 2035-2042

Find related abstracts

Hollands and colleagues have conducted an interesting meta-analysis to identify demographic and physical factors that increase the risk of being diagnosed with primary open-angle glaucoma (POAG). Their research question was whether primary care physicians can identify the people most at risk for glaucoma based on historical and physical findings accessible within their scope of practice (i.e., direct ophthalmoscopy). Unfortunately, in the absence of any studies that directly assess the diagnostic yield of patients referred for glaucoma evaluation from primary-care providers, this question was unanswered. Instead, the authors pooled the results of many high-quality population-based studies from which they retrieved data on both predictive factors (both from the history and the examination) and the outcomes of formal glaucoma evaluations. Their findings are familiar. Predictive demographic risk factors included high myopia, family history, black race, and increasing age. Predictive examination findings included increased cup-disc ratio (CDR), increased CRD asymmetry, disc hemorrhage, and elevated intraocular pressure (IOP).

in the hands of primary care physicians, it is likely that detection rates for early glaucoma will be even lower as subtle findings are likely to be overlooked

Not surprisingly, none of these factors cleanly discriminated between glaucomatous and healthy states at any cut-off values. Also not surprisingly, the diagnostic yield is disease stage-specific, with lower detection rates for early glaucoma and higher detection rates for advanced glaucoma. When we think about how these observations might play out in the hands of primary care physicians, it is likely that detection rates for early glaucoma will be even lower as subtle findings are likely to be overlooked. Similarly, there will likely be more advanced disease detected given that the most obvious outlying findings (very large CDR or very high IOP) are most likely to be recognized and referred for evaluation. Phrased in the parlance of epidemiology, screening for glaucoma by primary care providers is likely to be characterized by low sensitivity but high specificity. This is okay. I can live with this. We have to stop discarding less-than-perfect strategies for identifying undiagnosed glaucoma cases. Until (if ever) we have a perfect screening test for glaucoma, good enough has to be good enough.

Consider my perspective, which is informed by our research into effective glaucoma management within the developing world. In a setting where the majority of people with glaucoma are untreated or undertreated, a moderately effective intervention is better than no intervention at all. First-world standards have to be checked at the border. Lowering IOP adequately to halt progression is a great first-world treatment goal, but an intervention that is safe, effective and cost effective but only delays blindness by a year or two is still worthwhile in a region where almost everyone is undertreated and with little if any social services support for the blind.

Lowering IOP adequately to halt progression is a great first-world treatment goal, but an intervention that is safe, effective and cost effective but only delays blindness by a year or two is still worthwhile in a region where almost everyone is undertreated and with little if any social services support for the blind

This philosophy applies to screening in the developed world. I'm unconvinced that we should be concerned about missing early glaucoma cases in screening. I'd rather sacrifice sensitivity for specificity. I want a high yield on the cases referred in for confirmatory testing ‐ because that is what costs the most money.

So I propose we focus on finding the moderate and advanced cases of glaucoma. Any further research into glaucoma screening should not focus on better screening tests, but rather on two other key components of the process: firstly, improving the rate of follow-up for those with positive screening tests; and secondly, ensuring that people with negative screening tests are periodically rescreened throughout their lives. The importance of the first is obvious: screening is pointless if it doesn't lead to treatment. The importance of the second is fundamental: if we are going to sacrifice the early cases of glaucoma, we need to be sure they stay in the system and get rescreened and identified once they progress to moderate disease.

• ← Previous
• Next →