Glaucoma Screenings: Challenges and Failures

Introduction

Glaucoma is the second leading cause of blindness worldwide (cataracts are the leading cause). Over 8.4 million people will be bilaterally blind from glaucoma in 2010, and it is estimated that 12% of world blindness is caused by glaucoma.(1) Glaucoma is caused by a number of different eye issues that result in gradual vision loss due to damage to the optic nerve. Though high eye pressure is a risk factor for glaucoma, glaucoma may also occur in the presence of normal eye pressure. There are two main types of glaucoma in adults: primary open-angle glaucoma and closed-angle glaucoma. Open-angle glaucoma is the most common form and occurs when the intraocular pressure (the pressure inside the eye) rises because of a backup of fluid. Over time, as this pressure builds up, it damages the optic nerve and results in permanent and irreversable vision loss. Closed-angle glaucoma is more rare and occurs when the drainage canals of the eye get blocked causing a rapid increase in intraocular pressure.(2)

Since early detection and treatment are essential to preventing glaucoma and the accompanying vision loss, it is important to have routine eye exams by eye doctors. Assessment for glaucoma includes tonometry, a visual field test, and an eye exam by an eye doctor. Tonometry is a test that measures the intraocular pressure of the eye. The tonometry test can take measurements very quickly, but it is unable to detect glaucoma in those patients who do not have high intraocular pressure. This is problematic since over half of new glaucoma cases occur in people who do not exhibit high intraocular pressures.(3) During a visual field test, the patient is asked to look straight ahead and then indicate when lights can be seen in their peripheral vision. (4)

The Failures of Glaucoma Screening

Since many cases of glaucoma go unidentified and untreated, attempts have been made to create effective screenings for glaucoma, and these screenings are often conducted by laypersons and students. Even in developed countries, it is estimated that 50% of glaucoma sufferers remain undetected.(5) Though there are methods to test for glaucoma, glaucoma screenings have for the most part been ineffective and inaccurate in detecting the disease.

For example, the Student Sight Savers Program has implemented glaucoma screenings for over 41,000 people in the United States. The screenings include a questionnaire form, intraocular pressure measurement, and visual function assessment by frequency doubling technology (FDT). A research study entitled "Assessment of the Student Sight Savers Program Methods for Glaucoma Screening", published in the journal Ophthalmic Epidemiology, determined that of the Sight Savers’ three screening methods, the FDT test had the highest sensitivity, though it was only 58.1%. Sensitivity is the probability that a test gives a positive result when a disorder is present, while specificity is the probability that a test gives a negative result when the person tested is healthy.(6) Thus, a sensitivity measure of 58.1% means that nearly 50% of patients with glaucoma were not positively identified by the test.  The Sight Savers test also included IOP measurements with a non-contact tonometer, which demonstrated a sensitivity of only 22.1%.  When combining the three screening tests used by Sight Savers (FDT, IOP, and a questionnaire), the study determined that the sensitivity was only 88.6%, and the specificity was only 57.1%. Thus, the research study concluded that the Sight Savers screening test does not meet Prevent Blindness America standards for a screening test.  Prevent Blindness America’s criteria for minimum performance of a screening test is a specificity of 95% and a sensitivity of at least 85%. (7)

The low specificity and sensitivity of screening tests to find patients who have glaucoma is not unique to the Sight Savers study. A study dating back to 1991, which took place in Baltimore, similarly found that “even when a variety of combinations of factors or a multivariate predictive model is used, our ability to accurately classify persons into diseased and nondiseased groups was poor.” This led the researchers to conclude that “there is currently no test or combination of tests that provides a reasonable balance of sensitivity and specificity that would support the development and conduct of population-based screening programs for glaucoma.”(8)

Unfortunately, using imaging technology to screen for glaucoma, instead of FDT or tonometry, does not seem to be any more effective.  A study on the Heidelberg Retinal Tomography (HRT) in the Blue Mountains of Australia found that the test was 46% sensitive, and 91% specific.(9) The results predict that a screening program of 10,000 individuals with a 2% prevalence of open-angle-glaucoma (OAG) would detect only 80-90 out of the 200 affected, and would detect the presence of glaucoma in 390-880 normal individuals. Thus, using HRT would not be very cost effective or practical because it would result in the referral of too many normal individuals for evaluation and miss over half of those with glaucoma.(10) However, these results are less encouraging than most of the previous reports on the accuracy of HRT. This is largely due to the fact that the previous studies were based at clinics, whereas the more recent study used a random sample. “A general principle in evaluating a diagnostic test is that it will not perform as well in the “real world” as in a clinic setting. Tests generally perform better in the clinic, and glaucoma tests perform particularly well in a glaucoma-specialty clinic. The mix of glaucoma severity in a clinic is likely to be weighted to be more severe than in the population as a whole. Furthermore, clinic-based studies often compare perfectly normal volunteers with clearly glaucomatous patients, excluding suspects. The more the normal and affected groups differ, the less overlap, and therefore the greater the diagnostic accuracy. Therefore, it is inherently easier to distinguish between glaucoma and normal in most clinics than in a random sample of a population.” (11) Thus, until more accurate methods are developed to screen for glaucoma, “in the meantime, we should do our best to target high-risk populations, educate the eye care providing community to optimize glaucoma case finding during comprehensive eye examinations, and educate the community about glaucoma and the importance of periodic eye examinations to detect eye disease as recommended by the American Academy of Ophthalmology.”(12)

In addition to having low specificity and sensitivity, glaucoma screenings are problematic because they may unintentionally hinder people from seeking comprehensive eye exams by eye doctors. The intention of glaucoma screenings is to identify those with glaucoma who may not be under the care of an eye doctor, though we know from the literature that the screening methods do not meet published quality standards.  However, even if a community member truly doesn’t have glaucoma, they could have other eye conditions that need to be diagnosed and treated by an eye doctor, such as refractive error, cataracts, diabetic retinopathy, etc.  If a community member is told that their screening determined that they do not have glaucoma, they may be under the incorrect impression that they received an eye exam and therefore they would unfortunately believe that they don’t need to seek comprehensive eye care by an eye doctor.(13)  It is important that the comprehensive eye exam educational message be promoted to the community.   A cursory screening will not provide the community member with the essential comprehensive exam, diagnosis, or treatment that they need.  However, cursory screenings with medical equipment provide an illusion of comprehensive eye care.  Why should the community member go to an eye doctor if they were already “examined” with medical equipment at the glaucoma screening?  If the community members have not recently been to an eye doctor (or have never been to an eye doctor), they may not understand that there is a difference between the glaucoma screening and an actual comprehensive eye examination by an eye doctor.  The screening can therefore unintentionally reinforce and create new patient barriers to care, and can cause harm. Therefore, rather than implementing glaucoma screenings, which have been identified by research studies to be ineffective, it is instead important to promote regular eye exams by eye doctors.

In addition to providing the illusion that a comprehensive eye exam may not be needed, glaucoma screenings are also ineffective because of the lack of follow up that occurs after glaucoma signs are detected. This is largely because when glaucoma is caught early on, people do not exhibit symptoms, so they might not believe they have the disease, or they think that treatments won’t be effective. A study conducted by Mansberger on a community-based screening program using FDT found that about 1/3 of those who screened positive did not go for an evaluation by an eye care provider within six months.(14) Another study conducted in 2005 found that 20% of those who are diagnosed with glaucoma do not return for follow up care. “The unfortunate reality is that even when people learn about the risk of glaucoma, even if they have a printout that says they’re at risk, and even though they’re provided free health care and free transportation, a significant proportion do not follow up.”(15) Thus, it is necessary to educate the population to make sure that those at risk for glaucoma return for follow up care and understand the importance of taking their medications, even if they are asymptomatic.

“Screening programs must have an infrastructure to ensure that those who screen positive have a definitive exam and receive the treatments they need.” (16)(17) Additionally, unless the screening sensitivity is confirmed at 100%, those who screen negative must not be told that they do not have glaucoma. They, too, need an eye exam by an eye doctor.

Additional Considerations

Conducting population-based mass screenings for glaucoma is expensive and largely inaccurate.  In addition, recent studies have shown that conducting self-recruited population screenings by advertising about the risks and symptoms of glaucoma are also not effective. A study concluded that “providing information on the risk factors for glaucoma during the pre-publicity for such an event does not necessarily result in an above-average prevalence of glaucoma among self-recruited participants in non-population-based screening, as is the intention.”(18) Thus, a less expensive and more accurate approach is to provide public glaucoma educational sessions for a target population with a high risk for glaucoma.  One such high risk group is relatives of people with known glaucoma cases, who have a 10-fold increased risk of glaucoma. It is estimated that there are 100,000 family members of present glaucoma patients who also have undiagnosed glaucoma in the United States. Thus, ophthalmologists should actively ensure that those with a family history of glaucoma especially need to have regular comprehensive eye exams by eye doctors.(19)

Footnotes

(1) Quigley, H., and Broman, A. “The number of people with glaucoma worldwide in 2010 and 2010.” British Journal of Ophthalmology. 90. (2006): 262-267. Accessed on 1 October 2010. <http://bjo.bmj.com/content/90/3/262.full>

(2) “What is Glaucoma?” The Glaucoma Research Foundation. Accessed on 1 October 2010. <http://www.glaucoma.org/learn/what_is_glaucom.php>

(3) Cook, C., et. al. “Cataract and Glaucoma Case Detection for Vision 2020 Programs in Africa. An Evaluation of 6 Possible Screening Tests.” Journal of Glaucoma. 18.7 (2009): 557-562. Accessed on 6 October 2010. <http://journals.lww.com/glaucomajournal/Abstract/2009/09000/Cataract_and_Glaucoma_Case_Detection_for_Vision.10.aspx>

(4) Mansberger, S. “Should We Be Screening for Glaucoma?” Review of Ophthalmology. Accessed on 1 October 2010. <http://www.revophth.com/index.asp?page=1_14683.htm>

(5) Holló, G., Kóthy, P., Géczy, A., and Vargha, P. “Health anxiety in a non-population-based, pre-publicised glaucoma screening exercise.” Eye. (2010): 699-705. Accessed on 1 October 2010. <http://www.nature.com/eye/journal/v24/n4/full/eye2009131a.html>

(6) Lagrèze, W. “Vision Screening in Preschool Children. Do the Data Support Universal Screening?” Deutsches Arzteblatt International. 107. (2010): 28-29. Accessed on 1 October 2010. <http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2915478/>

(7) Salim, S., et. al. “Assessment of the Student Sight Savers Program Methods for Glaucoma Screening.” Ophthalmic Epidemiology. 16 (2009); 238-242. Accessed on 1 October 2010. <http://informahealthcare.com/doi/pdf/10.3109/09286580902863023>

(8) Tielsch, J., et. al. “A Population-based Evaluation of Glaucoma Screening: The Baltimore Eye Survey.” American Journal of Epidemiology. 134.10 (1991):1102-10. Accessed on 1 October 2010. <http://aje.oxfordjournals.org/content/134/10/1102.full.pdf+html>

(9) Lagrèze, W. “Vision Screening in Preschool Children. Do the Data Support Universal Screening?” Deutsches Arzteblatt International. 107. (2010): 28-29. Accessed on 1 October 2010. <http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2915478/>

(10) Maul, E., and Jampel, H. “Glaucoma Screening in the Real World.” Ophthalmology. 117.9 (2010). Accessed on 1 October 2010. <http://www.sciencedirect.com/science?_ob=MImg&_imagekey=B6VT2-50XBWWG-5-1&_cdi=6278&_user=483702&_pii=S0161642009012871&_origin=search&_coverDate=09%2F30%2F2010&_sk=998829990&view=c&wchp=dGLzVlz-zSkzV&md5=a5166123fcd5399220259b3ec56a3341&ie=/sdarticle.pdf>

(11) Maul, E., and Jampel, H. “Glaucoma Screening in the Real World.” Ophthalmology. 117.9 (2010). Accessed on 1 October 2010. <http://www.sciencedirect.com/science?_ob=MImg&_imagekey=B6VT2-50XBWWG-5-1&_cdi=6278&_user=483702&_pii=S0161642009012871&_origin=search&_coverDate=09%2F30%2F2010&_sk=998829990&view=c&wchp=dGLzVlz-zSkzV&md5=a5166123fcd5399220259b3ec56a3341&ie=/sdarticle.pdf>

(12) Ibid.

(13) Salim, S., et. al. “Assessment of the Student Sight Savers Program Methods for Glaucoma Screening.” Ophthalmic Epidemiology. 16 (2009); 238-242. Accessed on 1 October 2010. <http://informahealthcare.com/doi/pdf/10.3109/09286580902863023>

(14) Friedman, D. “Issues in Screening for Glaucoma.” Ophthalmic Epidemiology. 14 (2007): 101-102. Accessed on 1 October 2010. <http://www.v2020la.org/pub/PUBLICATIONS_BY_TOPICS/Glaucoma/Issues%20in%20screening....pdf>

(15) Mansberger, S. “Should We Be Screening for Glaucoma?” Review of Ophthalmology. Accessed on 1 October 2010. <http://www.revophth.com/index.asp?page=1_14683.htm>

(16) Friedman, D. “Issues in Screening for Glaucoma.” Ophthalmic Epidemiology. 14 (2007): 101-102. Accessed on 1 October 2010. <http://www.v2020la.org/pub/PUBLICATIONS_BY_TOPICS/Glaucoma/Issues%20in%20screening....pdf>

(17) Mills, R. “Glaucoma Screening: The Value is in the Details.” Accessed on 1 October 2010. <http://www.v2020la.org/pub/PUBLICATIONS_BY_TOPICS/Glaucoma/Glaucoma%20Screening.....pdf>

(18) Holló, G., Kóthy, P., Géczy, A., and Vargha, P. “Health anxiety in a non-population-based, pre-publicised glaucoma screening exercise.” Eye. (2010): 699-705. Accessed on 1 October 2010. <http://www.nature.com/eye/journal/v24/n4/full/eye2009131a.html>

(19) “A More Proactive Approach is Needed in Glaucoma Care.” Archives of Ophthalmology. 123 (2005): 1134-1135. Accessed on 1 October 2010. <http://archopht.ama-assn.org/cgi/content/full/123/8/1134>