Objectives
In examining the eyes and vision of preschool children, under the age of 4–5 years, the approach and method should be modified from the routine appropriate to older children and adults. It is likely that there will be much less cooperation on the part of the very young patient so clinicians need to use quick simple tests that can be applied for the short attention span. Precise measurements may not always be possible, and clinicians need to look for significant departures from normal.
A young child cannot give subjective symptoms and we rely on the observation and impressions of the parent. The family history will be very important. Children whose parents or siblings have strabismus or amblyopia are very much more likely to develop these conditions. Birth history is also important (e.g., prematurity, complications, birth weight).
This chapter is mainly concerned with the professional eyecare of young children, rather than with screening methods. No visual acuity test is likely to be adequate for screening by itself, even for refractive errors ( O’Donoghue, Rudnicka, McClelland, Logan, & Saunders, 2012 ): binocular vision tests and ophthalmoscopy are essential for detecting a wide range of potential problems ( Rydberg & Ericson, 1998 ). Regular eye examinations or vision screening throughout school life seems important as 70% of children who have significant ocular conditions go undetected by their parents or teachers ( Rose, Younan, Morgan, & Mitchell, 2003 ). Screening methods for preschool children are discussed on p. 175.
Unfortunately, many areas of the UK do not have good preschool or school screening programmes and, in the absence of these, optometrists should seek to examine all children (from neonates onwards) at routine intervals. This may not be necessary if there is a local screening programme that is thorough, properly audited, and repeated at periodic intervals to detect developing problems. However, even the best screening programme may fail to detect some anomalies and children with risk factors (family history, birth factors, symptoms) should always receive professional eyecare.
Active Pathology
Anomalies of binocular vision in children, as in adults, may be a sign of active pathology. The first responsibility of the practitioner is to investigate this possibility. It is particularly important to check for incomitancy, note the palpebral openings, and carry out careful ophthalmoscopy. Where there is any doubt, the patient will need to be referred for medical investigation before proceeding. It must be remembered that there are methods of investigation which are available in some hospitals but are seldom possible in primary eyecare practices.
Development of Vision
To assess the eyes and the vision of infants, it is important to know how vision normally develops from birth through infancy and childhood. Normal vision requires a good optical system with a focused image and good resolution. Optical resolution has also to be matched with a neural receptor system capable of good resolution and a neural image processing ability leading to perception. The perceptive level itself is very dependent on previous visual and other sensory experience. Obviously, this sensory experience cannot be present at birth. As experience grows, reflexes are reinforced and associations between different sensory input and experiences are formed. For example, it is clear that very early in life an infant learns to recognise the mother’s face and the meaning of different facial expressions. All visual functions are built as they are reinforced by experience acting on the anatomical and physiological systems which, although not complete at birth, mature early in life and allow the full potential of the visual system to develop.
The macular region of the retina is poorly developed at birth and both this and the visual cortex continue developing after birth. One would expect, therefore, that the spatial visual functions of neonates are significantly below the accepted norms for adults, and this is the case. It should be stressed that there is a wide variability in the development of visual functions and the figures given below are illustrative typical values from the literature. All aspects of visual development, normal and abnormal, have been reviewed ( Fulton, Hansen, Moskowitz, & Mayer, 2013 ). In contrast to major spatial resolution deficits seen in young infants, temporal resolution (e.g., flicker detection) is remarkably good ( Teller, 1990 ). Some theories account for this and other findings by considering the development of parallel pathways, including cortical and subcortical components ( Teller, 1990 ).
Contrast sensitivity at birth is far below adult levels and it improves rapidly over the first 6 months, achieving adult levels at the age of 3 months for low spatial frequencies, but taking more than 8 years to reach adult levels at high spatial frequencies ( Gwiazda, Bauer, Thorn, & Held, 1997 ). Faces are attractive to infants, and studies have shown that these are fixated at the age of 2 months but not at 1 month. Infants have some ability to discriminate between expressions at about 3 months and between faces at about 5 months.
Visual Acuity
The rate at which visual acuity develops in human infants depends on how it is defined and how it is investigated. Using the objective assessment of the visually evoked potential (VEP), it appears that the infant’s ability to resolve patterns improves from a level equivalent to about 6/38 at the age of 1 month to the equivalent of 6/15 at about the age of 6 months ( Teller, 1990 ). Another method, physiological optokinetic nystagmus (OKN), is mediated via a different nervous pathway to normal visual acuity. OKN methods are not in common clinical use and will not be covered here.
The most common approach to assessing visual acuity in infants is based on ‘preferential looking’ when the practitioner observes whether the infant turns the head or eyes to look at a grating or picture rather than a grey patch of equal size and luminance. If no preference is shown for the grating, it is assumed that it cannot be resolved, although it should be noted that this approach assesses extrafoveal vision. This method indicates that visual acuity is approximately equivalent to 6/180 to 6/90 at age 1 month, 6/90 to 6/36 at 3 months, and 6/60 to 6/18 at 6 months ( Appendix 2 ).
Obviously, a Snellen type acuity measurement cannot be made until the child is older even if specially designed tests are used which employ pictures (see pp. 50–52). These tests suggest that 6/6 acuity is not achieved until over the age of 3 years. All these Snellen type tests, however, involve an element of form perception. At the simplest, the child must be able to recognise the difference between shapes. Form perception is developed later than simple resolution, so Snellen type measurements assess a more advanced form of vision than preferential looking tests (p. 177). Kay pictures overestimate acuity ( Anstice and Thompson, 2014 ) and other picture charts (e.g., Lea) are better for predicting letter charts ( Anstice et al., 2017 ). Lea symbols are designed along the Bailey-Lovie LogMAR principles ( Hyvarinen, Nasanen, & Laurinen, 1980 ), but still typically produce about one LogMAR line better than with Snellen ( Laidlaw et al., 2003 ). Most children can complete a recognition acuity task by the age of 4 years ( Anstice & Thompson, 2014 ). Monocular acuities are possible only in two-thirds of children aged 3–4 years and nearly all over 4 years ( Salt, Sonksen, Wade, & Jayatunga, 1995 ).
With computerised letter charts, testing is much simpler with single optotypes surrounded by crowding bars, and this approach is necessary for some younger children. However, this method produces less pronounced crowding than linear tests ( Anstice & Thompson, 2014 ), which should be used as soon as the child is able (p. 177). Some norms for various clinical visual acuity tests can be found in Table 3.1 and a guide for clinical use is given in Appendix 2 .
| Age (months) | Test | VA (Snellen equivalent) | Source |
|---|---|---|---|
| Newborn | Preferential looking | 6/360 to 6/120 | Stidwill (1998) |
| Unspecified | 6/240 | Ansons & Davis (2001) | |
| 1 | Preferential looking | 6/180 | Teller (1990) |
| Preferential looking | 6/360 to 6/90 | Fulton et al. (2013) | |
| 3 | Preferential looking | 6/120 to 6/30 | Fulton et al. (2013) |
| 4 | Preferential looking | 6/50 | Teller (1990) |
| Preferential looking | 6/120 to 6/30 | Stidwill (1998) | |
| 6 | Preferential looking | 6/30 | Teller (1990) |
| Preferential looking | 6/60 to 6/19 | Fulton et al. (2013) | |
| 9 | Preferential looking | 6/46 to 6/12 | Fulton et al. (2013) |
| 12 | Preferential looking | 6/24 | Teller (1990) |
| Preferential looking | 6/38 to 6/9 | Fulton et al. (2013) | |
| 12–17 | Cardiff cards preferential looking | 6/48 to 6/12 | Adoh & Woodhouse (1994) |
| 18 | Preferential looking | 6/30 to 6/7.5 | Fulton et al. (2013) |
| 18–23 | Cardiff cards preferential looking | 6/24 to 6/7.5 | Adoh & Woodhouse (1994) |
| 24 | Preferential looking | 6/12 to 6/9 | Teller (1990) |
| Preferential looking | 6/24 to 6/6 | Fulton et al. (2013) | |
| 24–29 | Cardiff cards preferential looking | 6/15 to 6/7.5 | Adoh & Woodhouse (1994) |
| 30–36 | Cardiff cards preferential looking | 6/12 to 6/6 | Adoh & Woodhouse (1994) |
| 36 | Preferential looking | 6/6 | Teller (1990) |
| Preferential looking | 6/12 to 6/5 | Stidwill (1998) | |
| Single optotypes | 6/6 | Ansons & Davis (2001) | |
| 36–48 | Single optotypes | 6/6 | Atkinson, Anker, and Evans (1988) |
| Crowded optotypes (Cambridge cards) | 6/12 to 6/9 | Atkinson et al. (1988) | |
| 36–48 | Crowded optotypes (Cambridge cards) | 6/6 | Atkinson et al. (1988) |
Refractive Error
The refractive error during the first year of life is very variable in most infants. At birth it is approximately +2.00 DS (dioptre sphere) (standard deviation = 2.00 DS). Hypermetropic astigmatism is present in 29% and myopia in 23% ( Cook & Glasscock, 1951 ). In many infants, high degrees of astigmatism are observed during the first year, but this is variable and usually disappears before the end of the first year. On average, hypermetropia decreases rapidly during the first year to a mean level of about +1.50 D at age 1 year, and then decreases slowly at the average rate of about 0.1 D per year, until the age of 10–12 years when the typical rate of change slows even more. Myopia over 5 D in children under the age of 10 years can be associated with systemic or ocular pathologies and it has been recommended that these cases are referred ( Logan, Gilmartin, Marr, Stevenson, & Ainsworth, 2004 ), preferably to a paediatric ophthalmologist.
Nearly three-quarters of children with esotropia and/or amblyopia have a ‘significant’ refractive error (myopia, hypermetropia ≥+2.00 D, anisometropia ≥1.00 D, astigmatism ≥1.50 DC [dioptre cylinder]) and children with these refractive errors have a one in four chance of developing strabismus and/or amblyopia ( Bishop, 1991 ). Hypermetropic children are at higher risk of developing accommodative esotropia if there is a positive family history of esotropia, subnormal random dot stereopsis, or hypermetropic anisometropia ( Birch, Fawcett, Morale, Weakley, & Wheaton, 2005 ).
Leat (2011) reviewed prescribed criteria for infants and children and recommended the criteria reproduced in Table 2.12 and Appendix 2 .
Uniocular Fixation
Normally, the peripheral retina is well-developed at birth, but the central five degrees of the retina is at best partially functional at birth. Hence, in the first few weeks of life precise foveal fixation is unlikely, but fixation of suitable targets may take place at nonfoveal retinal locations. The tendency to fixate new objects increases during the first 3 months of life. The fixation reflex requires reinforcement by active vision if it is to develop normally. If the system is faulty in some way, this may prevent normal development of central fixation and therefore normal acuity.
The fixation reflex does not become firmly established until later and if anything impedes it during the first 2–3 months of life, central fixation can easily be lost. This period of 2 or 3 months of rapid maturation is known as the critical period for fixation. In comparison, the critical period for acuity development is 2 or 3 years. The critical period is followed by a further interval during which the system can easily break down: the plastic period. Central fixation can still be lost up to the age of 3 years if anything disturbs the system.
Occasionally there are abnormalities in the foveal nervous system which are present from birth, but these account for only a very small number of eyes with fixation failure. Most loss of fixation arises from the lack of a central image in a strabismic eye or, occasionally, from a very blurred image. If either of these is present during the critical period there will be a failure of development of normal fixation and acuity and, unless treatment is given before the end of the plastic period, it is unlikely that central fixation can ever be achieved. The longer the strabismus or the blurred vision is left untreated, the less chance there is of ever achieving central fixation with full acuity. This emphasises the need for early detection and treatment. It has also been shown that if an eye is occluded for a significant period (e.g., ptosis or cataract) during the critical period, this impedes acuity development (deprivation amblyopia). If the occlusion occurs before the age of several months, central fixation will also be lost.
Blinking Reflex, Vestibulo-Ocular Reflex, Saccadic and Pursuit Eye Movements
At birth, a blinking response to bright lights should be present ( Mehta, 1999 ). The vestibulo-ocular reflex is present (in full term infants) by the seventh day ( Mehta, 1999 ). Saccades are readily apparent in neonates but tend to be small and are relatively unresponsive to novel stimuli in the periphery. By the second week of life, small saccadic eye movements can reliably direct the line of sight towards a peripheral target and after the second month, large single saccades occur. Although this resembles the situation in adults, adult levels of saccadic accuracy are still being reached at 7 months of age ( Harris, Jacobs, Shawkat, & Taylor, 1993 ). Compared with adults, saccadic latencies are prolonged in infants, preschool children, and possibly even older children.
Pursuit eye movements are present in neonates, but are brief, intermittent, and frequently interspersed with saccades. Parents and clinicians should be able to detect the behavioural sign of infants fixating and following targets of interest by the age of 2 months. The visual system becomes better at pursuing faster targets beyond the first 10–12 weeks. In infantile esotropia, there is usually crossed fixation so that fixation occurs with the right eye for objects in the left field and with the left eye for objects in the right of the field. Under these circumstances, the pursuit reflex may develop normally in each eye for half the field. In the other half, the eye may not follow correctly if the other eye is covered. This will give the appearance of a lateral rectus palsy. As the crossed fixation is a form of alternating strabismus, it usually allows the development of good acuity.
Fusion, Vergence, and Stereopsis
Rudimentary binocular alignment without cosmetically noticeable strabismus is often present at birth, but true bifoveal fixation probably does not occur until the age of about 2–3 months. Occasional (<15% of the time) neonatal misalignments of the visual axes are common and usually innocuous in the first month of life but should become much less common in the second month ( Horwood, 2003b ). These are most often convergent, probably reflect the normal development of vergence control, and only require referral if they worsen after 2 months or if there is an intermittent deviation at 4 months ( Horwood, 2003a ).
Conjugate eye movements may or may not occur in neonates, although convergence may not occur for 2 months. One view ( Schor, 1993 ) is that tonic, proximal, and accommodative vergence are present at birth, but fusional vergence develops later (at about 4 months), possibly in line with improved visual acuity. A clinical implication of this is that infants are unlikely to show a vergence response to horizontally orientated prisms until fusional vergence develops. One would expect the development of sensory fusion to be closely interlinked with motor fusion, and this appears to be the case. Several measures of cortical binocularity confirm that sensory fusion is, on average, first found at the age of 3–4 months ( Birch, Shimojo, & Held, 1985 ). The ultimate goal of binocularity is stereoacuity and it is not surprising that research techniques have shown that stereoacuity is initially absent and then develops abruptly and rapidly from the age of about 3–4 months ( Birch et al., 1985 ). Clinically, stereopsis is detected at different times using different tests and these are discussed later in this chapter.
Cortical cells require an input from both eyes if they are to become the ‘binocularly driven’ cells of the normal adult system. This must occur during the critical period. A strabismus or occlusion of one eye will prevent this bifoveal stimulation and, if not checked during the critical period for this function, binocular vision may never be possible as the binocularly driven cells will not develop but will become monocular cells. Nelson’s (1988a) review on the ‘risk of binocularity loss’ concluded that ‘relative plasticity’ was at its maximum between the ages of 1 and 3 years. The plasticity then reduced rapidly at first (to 50% of its maximum at age 4 years) and then more gradually, to 20% of its maximum at age 6 and about 10% at age 8. Obviously, ‘infantile esotropia syndrome’ ( Chapter 15 ) with its onset under the age of 6 months needs early referral. If such patients are not seen until they are over a year old, it is very unlikely that anything other than a cosmetic operation will help. Indeed, the prognosis for achieving binocular vision in infantile esotropia is never very good.
Accommodation
Accommodation is probably present from birth but is initially (until the age of about 3 months) inaccurate and principally operative over a short range (from about 20–75 cm). It is thought the main constraints on accommodative function in infants are attention and detection of the blur signal. Under ideal conditions of attention, accommodative function varies from one infant to another ( Hainline, 2000 ), but is probably good enough to give them the acuity that their sensory system can resolve ( Aslin, 1993 ).
Examination
During the examination, young children usually sit on the carer’s knee, where they will feel more secure in otherwise strange surroundings. The parent or carer may be able to help in eliciting the child’s attention when required, or in steadying the head. Give time for the patient to get used to the situation while you are taking the history and symptoms from the parent or carer. Try to relate to the child in a friendly way at appropriate moments before carrying out any ‘tests’. Where possible, tests are presented as games to be played. A third adult in the room may be helpful for holding test cards and fixation targets for distance vision. Do not darken the room unless absolutely necessary, and then it is better to adjust the lighting slowly. It is sometimes advantageous to wear informal clothing, avoiding clinical white coats. Picture books may be useful and small attractive toys to hold attention for fixation are necessary, preferably ones that make a noise to give reinforcement. It is also useful to have toys in the reception area, and ideally a children’s area with a small table and chairs as this helps children to feel at home from the beginning of their visit. For toddlers, it is useful to have a car booster seat to place on the consulting room chair.
Methods and Equipment
For preschool children, the practitioner must decide the minimum test results they need to obtain and the most direct way of gathering this information. It is important to be quick and to frequently change tests to maintain interest. If a strabismus is suspected, the priorities are to address the following questions:
- 1.
Is binocular vision present?
- 2.
Are there any signs of a strabismus?
- 3.
Is the unaided vision the same in the two eyes?
- 4.
Is the refractive error normal for the age?
- 5.
Is the corrected acuity normal for the age?
It may not be possible to answer all these questions for every child. With others, much more can be done in addition. A lot will depend on the level of cooperation of the young patient. Patience and more than one visit may be required. The following procedures are typical of those that can be used in preschool children. The order of testing will depend on the child’s cooperation and the practitioner’s personal preference.
Vision and Visual Acuity
A variety of clinical tests allow the practitioner to measure uncorrected vision and visual acuity in children of any age, although monocular testing is particularly difficult at about 1–2 years ( Shute, Candy, Westall, & Woodhouse, 1990 ). For infants, preferential looking cards are usually the best method of assessment ( Fig. 3.1 ), and a single presentation is used when the looking behaviour is clear, with a maximum of three presentation when equivocal ( McCulloch, 1998 ). Classic grating pattern preferential looking cards, such as the Keeler acuity cards or Lea gratings paddles, are required for infants below the age of about 6 months. After this age, children become bored with these tests ( Teller, 1990 ) and, especially over the age of 1 year, children usually respond to the more interesting vanishing optotype cards (Cardiff Acuity Test; Adoh & Woodhouse, 1994 ). The Cardiff test is not good at detecting amblyopia ( Geer & Westall, 1996 ), so crowded optotype tests should be used as soon as the child is capable. Like any visual acuity test, the Cardiff test will not detect all cases of refractive error ( Howard & Firth, 2006 ).
