«MIST (Melbourne Initial Screening Test) Evaluation Studies February 1998 Part 1: Reliability and validity testing of the MIST Part 2: A negative ...»
The proportion of children able to be tested with a letter matching task varies with age.12, 13, 27 This would support the removal of the ‘inability to complete the gold standard examination’ as a failure criteria for children of this younger age group. A justification for passing the gold standard could be made on a judgment of those tests completed. The complexity of the task is the next consideration. It is suggested that cognition plays a greater part in crowded acuity than in single optotypes28 and other studies have reported fewer 3-year old than 4-year old children successfully completing a linear test.12, 13, 29 The acceptance of occlusion may be another indicator of the compliance level of children for VA testing, with studies showing a greater compliance with patching by 4-year old than 3-year old children.27, 28, 30 The level of VA accepted as normal in this age group must be decided. For vision screening, the accepted level of VA in Australia is 6/9.1 In relating the referral criteria to a gold standard for 3.5 to 4.5-year old children in particular, and accounting for the different testing distances, various levels have been cited in the literature. Some accept 3/6 as the acceptable level for either linear or single optotypes10, 23, 31, 32 others 3/4.511, 20, 30 and even 3/3,26, 33 referring any child with VA less than these levels.
Two studies set the strictest screening standards of less than 80 seconds using the Titmus stereo test,23, 34 one with older children. Earlier studies with Titmus tests reported that only 50% of children younger than 4 years gained stereoacuity better than 200 seconds of arc,35 and 83% of 4year olds gained 100 seconds or better.36 With the Randot E, Simons37 recommended a screening criteria of 250 seconds, though 168 seconds has been said to be the most sensitive threshold.10 The referral criteria for two screening programs using the Randot test have the acceptable level of stereoacuity at 200 seconds of arc, referring any children worse than this.21, 34 There is a reported This document is managed by the Department of Education and Early Childhood Development, Victoria (as of 27 August 2007) proportion of children who are unable to be tested with stereopsis tests, even in the presence of binocular vision, varying from 2.1%38 to 4%.10 Given these different findings, the gold standard criteria for this study was set at a pass level of better than 200 seconds of arc.
The aim of this study was to examine 200 children who had passed the MIST vision screening by the maternal and child health nurse, giving them a full orthoptic and ophthalmic examination in order to detect any ocular or visual problem that may have been missed by the screening program. Ethics approval was granted by the Human Ethics Committee, La Trobe University and the Department of Human Services Ethics Committee.
Participants Four testing centres from different regions of metropolitan Melbourne were selected – outer eastern, outer southern, inner Melbourne and inner western regions. This selection allowed a representative sample of both children and maternal and child health centres. All children who passed the screening test in these municipalities during the recruitment time were asked whether they were interested in participating in the project, a systematic sampling method. However, as interpreter facilities were not available, there was a necessary restriction to parents who had an appropriate level of english language. If parents expressed an interest, they were given an Information Form, then contacted by the researchers and an appointment was made for the assessment. All testing was performed within a few months of screening.
Procedures All clinical assessments were performed according to standard procedures and the data sheet protocol. In total, there were four ophthalmologists and five orthoptists who performed the assessments. At all sessions, each was ‘blind’ to the results of the other.
Prior to the assessment, each parent/guardian read and signed an Informed Consent Form. The first stage of the assessment consisted of the collection of general information. The orthoptic assessment consisted of
• cover test near (1/3 metre) and distance (6 metres)
• four dioptre test near and distance
• prism cover test near and distance
• ocular movement assessment
• convergence near point
• VA, right then left eye, at 3 metres. Acuity was performed initially with LogMAR reversible letter chart and matching card, followed by Sheridan Gardiner Single letters (SGS) if chart acuity was not possible. If LogMAR vision of 3/3.8 was not achieved with either eye, further testing was done with single letters
• fusional vergences, near and distance
• stereoacuity, Randot Stereotest After the orthoptic examination was complete, Cyclopentolate 1% was instilled in each eye. Each child was seen by the ophthalmologist 30 minutes later. The ophthalmic examination included macroscopic inspection, slit lamp examination, retinoscopy performed at the examiner’s standard working distance and indirect ophthalmoscopy.
Gold Standard Criteria The criteria for failing a gold standard examination in this study of younger children were defined after adaptation of those by De Becker et al.21
• VA of 3/4.8 (-3) or less in one or both eyes, (LogMAR 0.24) or
• VA of 3/4 (-2) or less in one or both eyes, SGS
• a difference in VA of two lines or more between eyes (LogMAR difference 0.18)
• stereoacuity of 200 seconds of arc or less This document is managed by the Department of Education and Early Childhood Development, Victoria (as of 27 August 2007)
• any constant or intermittent heterotropia
• refractive error of
0.75 dioptres or more of myopia
3.50 dioptres or more of hypermetropia
1.50 dioptres or more of astigmatism
1.00 dioptres or more of anisometropia
• any anomaly judged significant enough to require follow-up.
All data from the examinations was coded and entered into a computer statistical program (Statview) for analysis. VA was recorded on the data sheets as the number of letters correctly matched on each line, an interpolated LogMAR score. Single optotype acuity was coded according to the number of letters correctly matched.
Over the recruitment period, Parent Contact Forms were received from the nurses for 296 children, of whom 201 (100 males and 101 females) participated in the examination. The mean age of the children at testing was 45.7 months (3 years, 9.7 months) (SD 2.93 months) with a range from 40 to 55 months. The 201 children were each seen at one of four testing centres, by one of five orthoptists and one of four ophthalmologists. The numbers assessed by each can be seen in Table 7.
Family History From the family history, 61 of the children (30%) had one or both parents who wore glasses.
Seventeen (8.4%) of the children had a relative with a known strabismus and another 14 (6.9%) reported a family member with a ‘lazy eye’.
Ocular deviation Only five of the children were noted to have an exophoria for distance fixation (mean deviation of 5.2 prism dioptres). Thirty-eight children were shown to have some latent deviation for near fixation, an exophoria noted in 35 (mean 6.1 prism dioptres) and an esophoria noted in three (mean 5.3 prism dioptres). One child (#162) was found to have an intermittent small esotropia.
Convergence near point (CNP) was normal in all children, with a mean CNP of 3.9 cm (SD 0.96), ranging from 1 to 7 cm.
Binocular function All 201 children demonstrated a binocular response with the 4-dioptre prism test and so were considered to show no central scotoma.
An acceptable prism fusion vergence range for near fixation was obtained from all children, with the mean convergence break point being 38.7 prism dioptres (SD 6.46) and the mean divergence 16.1 (SD 2.78), similar to that previously reported.35 Prism fusion vergence range for distance fixation was obtained on all children except one, mean convergence 16.6 (SD 4.34) and mean divergence 11.8 (SD 2.32).
This document is managed by the Department of Education and Early Childhood Development, Victoria (as of 27 August 2007) One-hundred-and-sixty-six children (82.5%) obtained a stereoacuity result of at least 70 seconds of arc. The minimum standard defined for this current study was stereoacuity of better than 200 seconds of arc, 95% achieved this. Ten children (5%) did not obtain this level, four obtained 200 seconds of arc and two obtained 400 seconds of arc. This left four children from whom a result was not obtained, due to a lack of comprehension of the test. All but one of these 10 children were demonstrably binocular with good prism fusion vergences. The child previously described (#162) with an intermittent esotropia gained stereoacuity of 400 seconds.
Visual acuity The preferred test of VA was the 3 metre LogMAR chart with matching letters. If the children were unable to cope with this complex task, the simpler SGS was used. Using the chart, VA of at least 3/3.8 (≥3 letters) (LogMAR ≤ 0.14) was achieved in the right eye by 121 (61%) and in the left eye by 119 (59%) of the children. The gold standard level of VA to be achieved in this study was 3/4.8 (≥3 letters) (LogMAR ≤0.24), and the number of children achieving this level increased significantly to 162 (81%) in the right eye and 160 (80%) in the left eye.
In the right eye, of the 39 children who did not achieve chart acuity of 3/4.8, 33 achieved a single acuity level of 3/3 and six achieved a level of 3/4. This meant that all of the 201 children gained either 3/4.8 linear acuity or 3/4 single optotype acuity in the right eye. In the left eye, of the 41 children who did not achieve 3/4.8, 36 achieved a single optotype acuity level of 3/3 and four achieved a level of 3/4. This meant that 200 of the 201 children gained either 3/4.8 linear acuity or 3/4 single acuity in the left eye.
One child did not achieve a satisfactory level of VA. This child (#115) achieved an unequal VA result of 3/4.8 (2) (LogMAR 0.26) in the right eye and 3/9.5 (3) (LogMAR 0.54) in the left eye with the linear test, and 3/3 (1) and 3/6 being achieved in the right and left eyes respectively with the SGS.
Visual acuity difference Of the children achieving a result with the chart, only two children had an acuity difference of two lines or greater. One child (#115) described previously had a difference of two lines and one letter on the chart, and seven letters on single acuity. Another child had a difference of two lines on the chart, but achieved equal VA of 3/3 with the single optotypes, so was considered to have passed the gold standard using the single letters.
Refractive error The refractive error was recorded as the retinoscopy result calculated for the working distance, using minus cylinders, with no calculation being made for the effect of the cyclopegia. For the 201 children the mean spherical error was +1.18 dioptres (SD 0.703) in the right eye, ranging from 0.00 to +3.25 and +1.17 dioptres (SD 0.685), ranging from 0.00 to +3.00 in the left. The mean astigmatic error was 0.17 dioptres (SD 0.286) in the right eye and 0.16 dioptres (SD 0.263) in the left eye, both ranging from 0.00 to 1.5 dioptres. All children passed the gold standard for spherical equivalent refractive error. Four children failed the gold standard criteria, with an astigmatic error of 1.50 dioptres in one eye (#6, #19, #115, #158).
The mean anisometropic difference between the two eyes was 0.07 dioptres (SD 0.119), ranging from 0.00 to 0.75 dioptres. All children passed the gold standard for anisometropia.
Ophthalmic assessment On macroscopic inspection, only two children were reported to have an ocular anomaly, one with a small right lid chalazion, the other with a slight right ptosis. On slit lamp examination three children were noted to have some observable differences. One child had a small area of focal endothelial changes in the peripheral cornea, two other children were noted to have epiblepharon, one mild and one minimal. On dilated ophthalmoscopy, two children were noted to have physiological variations, one with physiological cupping of the discs and another with buried This document is managed by the Department of Education and Early Childhood Development, Victoria (as of 27 August 2007) optic nerve head drusen. None of these conditions were considered clinically significant, nor to require any further follow-up.
Failure of the gold standard criteria In this age group it is expected that a small number of children will be unable to complete the full gold standard assessment due to factors such as the level of cognitive development and ability to maintain concentration on the tasks for the length of time required.12, 13, 27 It is therefore necessary to interpret the results to the gold standard assessment in this light, with the incompletion of any test viewed in association with the other results obtained.
Only five children (2.5%) were unable to complete the examination, each missing one test. For each of these children confirmation was available from all other tests, both objective and subjective, that ocular status and visual function were normal. Measurement of distance prism fusion vergence was not obtained on one child, but the child was orthophoric for both near and distance fixation, with near prism vergence, stereopsis and VA all normal. Measurement of stereoacuity was not obtained on four children, all of whom were orthophoric with normal prism vergences, VA and refraction, but were noted to have either a poor understanding of english or no comprehension of the test. All of these children were considered to have passed the gold standard examination.