Evidence-based eye examinations

1.1.1

Visual acuity should not be measured with Snellen charts

The principal reason for using a Snellen chart nowadays would appear to be tradition or habit. Visual acuity (VA) charts using the logMAR system are now widely available and are much superior to Snellen charts, providing VA measurements that are twice as repeatable and over three times more sensitive to interocular differences in VA (see section 3.1.1 ). ^, VA can still be recorded and reported to others in the more familiar Snellen format of 6/6, 20/20, or 1.0.

1.1.2

The von Graefe prism dissociation test does not deserve its widespread use

The faith in this test has been such that a publication in a leading international optometry journal pronounced it as the gold standard test of heterophoria assessment and used it to assess the usefulness of the cover test. The gold standard in this area should be the cover test and not the von Graefe as it is objective and not reliant on subject responses or subject to prism adaptation and subsequent studies have shown it to be far more repeatable than the von Graefe prism dissociation test, which has been shown to be unreliable. The study should have used the cover test as the gold standard and they would then have reported the limitations of the von Graefe prism dissociation test. Several studies have now reported that the modified Thorington test is much simpler, faster, and more reliable than the von Graefe prism dissociation test, and if an additional test to the cover test is required, it should be the modified Thorington.

1.1.3

Lengthy tentative reading addition tests are not necessary

A final example of tests used because of tradition in phoropter-based refractions is the use of tentative reading addition tests of binocular cross-cylinder followed by negative and positive relative accommodation (NRA-PRA). These are lengthy tests given that the only study to have properly investigated the various tentative addition tests indicated that asking patients their age provided more accurate information than either of those two tests. Using an ingenious study design, Hanlon et al. examined patients who returned to an optometry practice because they were dissatisfied with the near vision in their new glasses. Each patient’s reading addition was determined using four methods (age, binocular cross-cylinder, NRA-PRA balance, ¹ / ₂ amplitude of accommodation). The review (recheck) examination also determined whether the near addition in the glasses the patient disliked was too low or too high. The percentage of adds for each tentative add test that gave the same result as the incorrect add or worse (higher than an improper add determined too high or lower than an improper add determined as too low) was calculated. They reported that the simplest and quickest test, asking the patient their age, accounted for the fewest errors (14%). The other techniques gave errors in 61% (binocular cross-cylinder), 46% (NRA-PRA), and 30% ( ¹ / ₂ amplitude) of cases. This suggests that the tentative addition should simply be based on patient age. Subsequent research suggests that the tentative addition estimate can be further improved by considering both the patient’s age and working distance and/or symptoms with their current near correction (see 4.13.1 ). There is little need for the lengthy binocular cross-cylinder followed by NRA-PRA.

1.2.1

PubMed, Google Scholar, and international optometry research journals

If you wish to review the literature, two very useful free access databases are PubMed ( www.pubmed.com ; provided by the US National Library of Medicine) and Google Scholar. They both include the abstracts or summaries of papers from all the main optometry and ophthalmology research journals. Questions from clinicians on optometric internet/e-mail discussion groups can often be fully answered by a quick PubMed or Google Scholar search that can provide a much stronger level of evidence than anecdotal suggestions from colleagues based on one or two patient encounters.

Full access to one or more of the international optometry research journals is provided by membership of various professional bodies: Ophthalmic & Physiological Optics (College of Optometrists, UK), Optometry & Vision Science (American Academy of Optometry), Clinical & Experimental Optometry (Optometry Australia, New Zealand Association of Optometrists, Hong Kong Society of Professional Optometrists and the Singapore Optometric Association), Journal of Optometry (Spanish General Council of Optometrists) , Contact Lens & Anterior Eye (British Contact Lens Association) , African Vision & Eye Health (South African Optometric Association), the Chinese Journal of Optometry & Ophthalmology , and the Canadian Journal of Optometry . Some professional bodies also have their own library and provide full access to a wide range of online optometry and ophthalmology journals.

1.2.2

How do you assess the usefulness of optometric tests?

The usefulness of optometric tests is typically assessed by either comparing the test against an appropriate gold standard and/or assessing its repeatability and/or its discriminative ability. For example, a test that is being used as an objective measure of subjective refraction should be assessed by how closely the results match subjective refraction results and new tonometers are assessed by their agreement with the results of Goldmann Applanation Tonometry (GAT, although this is not ideal when GAT has flaws, see section 7.7.1 ).

1.2.3

How to assess tests that are part of the subjective refraction

The use of subjective refraction as a gold standard assessment of refractive error has meant that there has been little or no comparison of the various methods used in subjective refraction. Previous studies have tended to compare the various tests against each other. For example, West and Somers compared the various binocular balancing tests and found that they all gave similar results and concluded that they were therefore all equally useful. Johnson et al. reported a similar finding when comparing subjective tests for astigmatism. However, the size of the differences found with the different balancing and astigmatic tests in these studies would likely have led to symptoms and patient dissatisfaction with some of the refractive prescriptions if they had been worn, so that the conclusions seem incorrect and the study design poor. An inventive but under-utilised approach is to use some measure of patient satisfaction or dissatisfaction as the gold standard. For example, Strang et al. compared the refractive corrections provided by subjective refraction and autorefraction by randomly allocating glasses in a double-blind protocol. Subjects wore each prescription for 2 weeks and completed a questionnaire that assessed visual performance and ocular comfort following each period of wear.