Snellen Versus ETDRS

Visual acuity (VA) testing is an essential part of a comprehensive eye exam – but it is also often used in ophthalmology clinical trials as an efficacy and/or safety outcome. Nevertheless, there are multiple methods by which VA can be evaluated (1). The Snellen VA chart is the most common VA testing method performed in clinical settings (2), but the Early Treatment of Diabetic Retinopathy Study (ETDRS) charts are largely used in research; indeed, they are required by the US FDA for measurement of visual outcomes in clinical trials. In this latter regard, the ETDRS Best-Corrected Visual Acuity (BCVA) methodology is the gold standard (3,4).

Eye care providers often rely on clinical trial outcomes to make decisions, making it all the more important for us all to understand the key differences between Snellen and ETDRS VA testing:

• ETDRS charts are more precise at measuring VA than Snellen charts (5); Snellen charts do not test VA between the 20/100 and 20/200 lines, whereas ETDRS charts do capture VA measurements between those lines (4)
• ETDRS charts enhance testing standardization and accuracy by including more lines than Snellen charts, which are evenly spaced (0.1 log unit line steps; Logarithm of minimum angle of resolution [logMAR]). This greater accuracy with ETDRS charts results in increased testing time. Lim and colleagues report that mean reading times for normal subjects on the ETDRS chart are 34.65 seconds versus 18.67 second on the Snellen chart (6).
• Unlike Snellen, ETDRS always has five letters on each row, and the spaces between letters is logarithmic and always equidistant (2).
• ETDRS testing is conducted at four meters (13 feet) or at one meter (3.28 feet; used with subjects who have decreased VA), while Snellen testing is normally tested at 20 feet or a simulated 20 feet, hence the 20-foot denominator in Snellen VA (2,7).  
• ETDRS VA is typically calculated based on a scoring system of the total number of correctly read letters between both the 4-meter and 1-meter charts, while Snellen VA is based on the smallest line the subject reads correctly (2,7).
• ETDRS is more reproducible than Snellen; Shamir and colleagues report a mean test retest variability (TRV) of 0.107 with ETDRS versus 0.232 with Snellen (lower TRV is better) (2,5).
• ETDRS charts typically take up more space and are costlier than Snellen charts given that ETDRS charts require specific lighting and multiple charts (2,5).
• Unlike Snellen, ETDRS BCVA testing requires a standardized trial frame refraction that can be time consuming without proper training or experience (7). Further, ETDRS trial frame refractions require use of a specific ETDRS refraction chart that is exclusively used for ETDRS refractions (7). Additionally, the ETDRS charts that are used for determining BCVA cannot be used for ETDRS refraction and vice versa (7).
• ETDRS charts can produce better VA than Snellen charts; Kaiser reports an average difference of -0.13 LogMAR with ETDRS charts compared with Snellen charts (3). This is equivalent to 6.5 more letters read and more than one line better with ETDRS (7). This better VA measurement with ETDRS increases as the subjects’ VA worsens; Kaiser specifically reports that those with VA worse than 20/200, the difference between Snellen and ETDRS is on average -0.20 LogMAR or two full lines better with ETDRS (3).

The notion that the Snellen VA is not as accurate as ETDRS testing should give pause to clinicians and researchers; therefore, when deciding which VA chart is best, there are several factors to consider. If eye care professionals are conducting clinical research with BCVA as an efficacy outcome, ETDRS will likely be required, and you will also be required to be able to test at four meters and at one meter when subjects have the potential for abnormal vision. You will be required to have a dedicated ETDRS exam lane with proper lighting and equipment, and you will be required to be ETDRS-certified as an examiner. The FDA considers 15 ETDRS letters (three ETDRS lines) to be the minimal clinically important difference (MCID) in ophthalmology trials, and because ETDRS tends to produce more precise VA measurements, investigators will only want to use ETDRS to determine if a clinical trial intervention is efficacious (3,8). Alternatively, Snellen charts may be acceptable on rare occasions in clinical trials when VA is only a safety outcome.

Snellen may also be the best option under other conditions. For one, there are certain circumstances where Snellen VA is required, such as in a US Department of Transportation physical (9). Additionally, if there is no clinical trial being conducted and limited space or portability is a concern, Snellen is probably a better option (2,5). Further, if time is a concern, such as in obtaining an eye exam screening VA or an uncorrected VA on a patient before a refraction in a non-research setting, Snellen VA is probably the best option. However, if time is not a substantial factor and the patient does not have ocular disease, ETDRS and Snellen can be used interchangeably in a clinical setting (10).

If an eye care provider is treating patients with ocular disease and is not planning on conducting any clinical research in their clinic, given that Snellen is less costly and a faster test, Snellen VA may still be a better choice if patient fatigue, time, or costs are a concern (5). However, in such a case, examiners should be consistent with their testing methods, especially in a multi-doctor practice. Clinicians that only use Snellen should be mindful that if they refer patients for eye care in clinics and sites that do conduct clinical research, ETDRS BCVA improvements from clinical trial interventions, especially in those with poorer BCVA, may not be fully appreciated when the patient returns to their referring Snellen-only clinic, especially if the improvements are made between 20/100 and 20/200 lines, which Snellen does not measure (4). Further, if an eye care provider in a clinic with only Snellen VA charts is using Snellen VA to justify recommending that a patient consider a clinical trial at another site, if the patient is seemingly on the border of eligibility with Snellen VA testing, the patient may actually be above or below the ETDRS BCVA threshold for enrollment. Regardless, if providers believe patients could be eligible and that the patient can benefit from a clinical trial, they should educate their patients on their options and refer such patients to enrollment sites for more information and potential care. For a list of active clinical ophthalmology trials see clinicaltrials.gov.

For organizations that desire to participate as a site in clinic trials, establishing a BCVA testing lane is essential. Some studies may also require advanced ETDRS BCVA testing training, which is study specific. This advanced training is typically accomplished with the assistance of a specialized BCVA ETDRS certification program. Though the requirements for conducting accurate ETDRS BCVA testing and refractions may seem daunting, because ETDRS testing is more specific than Snellen VA testing, because it requires more thorough staff education, because it requires a dedicated room, and because it require specialized equipment, these BCVA testing certification programs can simplify the learning process and streamline the implementation a clinical trial. BCVA testing services can train and certify staff members and evaluators in-person, they can assist in the setup of a BCVA testing room, certify an ETDRS BCVA testing room, and facilitate future recertification of staff and ETDRS BCVA testing rooms. As a result of a ETDRS BCVA certification program, sites can optimize their ability to participate in ophthalmology clinical trials and will possess the confidence and skills necessary to accurately score tests.