Screening for glaucoma using intraocular pressure alone

BMJ 2017; 358 doi: https://doi.org/10.1136/bmj.j4160 (Published 13 September 2017) Cite this as: BMJ 2017;358:j4160
  1. Henry D Jampel, Odd Fellows professor of ophthalmology
  1. Johns Hopkins Wilmer Eye Institute, 600 N Wolfe St, Baltimore, MD 21287-9205, USA
  1. hjampel{at}jhmi.edu

An outdated concept that should be abandoned

In a linked paper, Chan and colleagues (doi:10.1136/bmj.j3889) report the distribution of intraocular pressure (IOP) and the frequency of glaucoma in the EPIC-Norfolk cohort, a community based cross sectional study of a UK population.1 With knowledge of both the distribution of IOP and the frequency of glaucoma, the authors determined that the diagnostic capability of IOP alone to detect glaucoma is poor. This conclusion is important because “Glaucoma and suspected glaucoma combined account for the sixth largest share of National Health Service (NHS) outpatient attendances. . .”.1 Since guidelines from the National Institute for Health and Care Excellence2 describe one of the features of chronic open angle glaucoma as “IOP in either eye exceeding 21 mm Hg,” referrals solely on the basis of IOP may be contributing to resource overuse.

Glaucoma, an optic neuropathy with a characteristic appearance of the optic disc and progressive vision loss, is a leading cause of blindness worldwide.3 The authors’ findings of a prevalence of 3-4% in a population aged 48 years or more, a preponderance of open angle glaucoma, and increasing prevalence with age have been frequently seen in other, similar studies.45 Readers should be aware that the population studied was almost entirely white. Open angle glaucoma is much more prevalent in people of African descent,6 and angle closure glaucoma is more prevalent in people of Asian descent, particularly those from China.7

That vision loss from glaucoma can be reduced by treatment to lower the IOP and that glaucoma is largely asymptomatic until there is noticeable damage to the optic nerve, provide the rationale for screening for the presence of glaucoma in the community or in a primary care setting.

The designation “glaucoma” is now reserved for people with damage to the optic nerve, not a particular level of IOP. Screening strategies for glaucoma could potentially include: historical and demographic information, as glaucoma is more prevalent in older people, those of African origin, and those with a first degree relative with glaucoma8; IOP measurement, as the prevalence of glaucoma increases as the IOP increases9; imaging of the optic disc, retinal nerve fibre layer, and macula, because all these structures change with progressive optic nerve damage; and tests of visual function, particularly the visual field test.

Historically, IOP has been a common screening tool for glaucoma for several reasons. Firstly, the higher the IOP, the more likely it is that someone has glaucoma or will develop glaucoma. Secondly, much evidence has shown that a raised IOP is not only associated with glaucoma, but can cause glaucoma damage. Lastly, the measurement of IOP is simple, can be performed by non-specialists, and produces a single number that does not require sophisticated interpretation.

The deficiencies of IOP as a standalone screening tool for glaucoma, however, have been known for decades. Chan and colleagues’ data from the EPIC-Norfolk cohort reaffirm that no IOP threshold reliably separates those with optic nerve damage from those without. We now know that approximately 50% of people with glaucoma (about 76% in the current study) have an IOP of less than 21 mm Hg and so would be missed by screening; furthermore, approximately 10% of people without glaucoma have an IOP of at least 21 mm Hg.9 Many of the latter will never develop the condition and would be overdiagnosed and potentially overtreated after screening.10

The authors conclude that “The large number of people with confirmed glaucoma and intraocular pressure under the threshold for ocular hypertension (21 mm Hg) reinforces the weakness of reliance on this for detection of glaucoma,” but do not provide alternative solutions. Ideally, a screening test would actually detect the disease rather than simply identifying one of the risk factors. Two possible options are imaging the optic nerve and retina to detect any structural damage, or testing the function of the optic nerve. Both remain problematic.

Optical coherence tomography can rapidly image the optic nerve and retinal nerve fibre layer to identify structural damage. Although one preliminary clinic based report showed the high diagnostic accuracy of optical coherence tomography for diagnosing glaucoma, the test has yet to be evaluated for screening11 and the associated costs presently precludes widespread use in community or primary care settings. An analysis of data from the National Health and Nutrition Examination Survey revealed that functional testing of the optic nerve with frequency doubling technology perimetry was not sensitive or specific enough for population based screening.12 Unfortunately, currently no adequate screening test for glaucoma exists.1314

Instead of attempting to detect individual eye diseases it may be more cost effective to screen for all leading causes of vision loss at the same time: cataract, glaucoma, age related macular degeneration, and diabetic retinopathy. For example, a study of screening for diabetic retinopathy using a non-mydriatic camera with remote evaluation of images also found a glaucomatous appearing optic nerve in 10% of screened eyes, and age related macular degeneration in 9%.15 Researchers could usefully investigate combined approaches to screening in future studies. Reliable and timely detection of glaucoma, a common and treatable cause of blindness, is a public health priority.


  • Research, doi: 10.1136/bmj.j3889
  • Competing interests: I have read and understood the BMJ policy on declaration of interests and declare the following: I have advised Ivantis and Endo-Optiks on trials of potential treatments for glaucoma.

  • Provenance and peer review: Commissioned; not peer reviewed.


View Abstract