To determine the most appropriate methods and criteria for vision screening, we used the five best screening tests (based on published data) on 712 junior- and senior-kindergarten children, then compared the results to the findings of a licensed optometrist who conducted a comprehensive eye exam of each child. To identify implementation issues we also screened more than 6500 kindergarteners in 30 communities across Ontario. These recommendations and sugggestions are based on our report of those studies plus additional analyses and our experience in the field.
The Ministry expects public health units to use three of the tests we tried. These are described below. A child passes screening only if he passes all three tests, so a shortcut might be to refer a child as soon as he fails one test or cannot complete a test. However, for data analysis the Ministry wants all of the tests done.
If a child is already wearing glasses, then there is no point to bothering with the third test below, the autorefractor, because that will merely confirm that he needs glasses. However, we still give him the first two tests, to check that his glasses are still suitable. .
With the Minstry's screening cut-offs, a child who fails screening stands a 50% chance of having a vision problem and a child who passes screening stands less than a 10% chance.
Most children with eye problems do not fail screening dramatically but some of them do. We think that screeners ought to watch out for those children and follow up as necessary to make sure that parents take them to an optometrist. If screening shows any of these results, an eye exam will almost always reveal a problem that needs treatment:
Crowded HOTV visual acuity test
The HOTV Letter Book with 50% Spaced Bars is the preschool equivalent of a conventional eye chart. It takes the longest time (three to seven minutes per child), so we found it sensible to buy two sets and have two screeners use this test simultaneously.
Each eye must be tested separately. The child is shown one letter at a time with bars around it. (Those "crowding bars" are essential: the version lacking them is too easy.) The child can either name the letter or, if he is shy or weak in English or doesn't know his letters, he can point to the same letter on a card. Testing continues with smaller and smaller letters.
Testing one eye at a time requires the other eye to be covered, but a kindergartener cannot be trusted to cover an eye completely with his own hand, especially when the letters get smaller and more difficult to read. For this reason eye patches are necessary. We like these:
The HOTV test requires good lighting on the screener's letter chart, which must be 10 feet from the child. We have done it in classrooms, closets, and stairwells.
Randot stereo test
The Randot Preschool Stereoacuity Test is a book the child looks at while wearing 3D glasses. On each page she searches for "the shape hiding in the snow", a shape that will pop out of the background if she has stereoscopic vision. If a child is sociable and has adequate English, then she should be able to tell the tester what she sees, else she can match the shape to one of four black-and-white shapes on the left side of the book. Testing continues through finer and finer degrees of stereo acuity, and typically takes one to two minutes. Correct answers are on the back of the book or the screener can see them if he wears 3D glasses. The test requires a place to sit beside the child and adequate light on the book. A Canadian source is www.topcon.ca.
The testing kit comes with one pair of child-sized 3D glasses. These are easily broken, so it's necessary to have spares on hand. Also, some children have large heads, so at least one pair of these glasses needs to be the adult size.
Screening autorefractors are light weight, hand-held devices that take a picture of the child's eyes. A light from the autorefractor strikes the child's eyes and is reflected back in a systematic way that indicates whether or not there is a refractive error (i.e., near-sightedness, far-sightedness, astigmatism). Using one takes less than a minute. The child need only sit still with open eyes and stare at the device.
Autorefraction needs only a small room or closet, but the light must be dim. When we have only one room to use for all of our tests, for autorefraction we erect a free-standing tent and drape black fabric over the top in lieu of a fly. To prevent tripping we cut the sill off the tent. The door we leave tied open.
At the time we began our research in 2014, the literature indicated that two autorefractors had the best reliability and validity: the Plusoptix A12C/R Mobile Binocular Autorefractor and the Welch Allyn "Spot" Visual Screener, Model VS100. Our data show the Plusoptix to be slightly more sensitive and slightly less specific, but their results are so highly correlated that we see no reason to choose one over the other. One member of our central research team prefers the Plusoptix and the other prefers the Spot.
Each of these units will convert their numerical measurements into "pass" or "fail". This reduces human error but the default criteria are unlikely to match the Ministry's specifications, so you will probably need to change them. Customized criteria for the Plusoptix must be obtained from the manufacturer.
To carry a photorefractor safely from one school to another requires purchasing a carrying case.
An autorefractor must be charged overnight and it must never be left where it can bake or freeze.
An experienced screener can screen eight to 10 five-year-olds per hour but children are not available continuously. When we need to screen a lot of children we find it more efficient to be able to handle them in larger batches. To do this we work with a team of five: two to give HOTV tests, one to give the stereo test, one to use the photorefractor, and one to marshal children and relieve the others.
The optimal number of screeners will depend upon the size of the school and the amount of travelling required. With 35 children to be tested and/or remote communities, we usually used two testers and occasionally just one. With 100 children we always used five.
Volunteers are a mixed blessing. One person may come on Tuesday and another person on Wednesday. Someone with years of experience screening children may want or revert to doing things his way, whether they fit the Ministry's protocol or not. We happily accepted the help of volunteers whenever we could but for quality control, we found it necessary to have one of our team on hand at all times.