Diagnostic assessment of CVI in children is an interdisciplinary task, which needs close cooperation between the disciplines of (neuro-)ophthalmology, neuropaediatrics, developmental neuropsychology, orthoptics as well as early intervention and special education specialists. Observations reported by parents, brothers and sisters and other reference persons about the child’s visually guided behaviour such as grasping for toys, visual exploration, fixation, responses to display of toys and other objects and play behaviour can provide significant indications concerning spared and affected visual capacities. In principle, it has to be acknowledged that when assessing children, who do not possess sufficient language competence and, therefore, cannot report or comment on what they can see and why vision is problematic, behaviour-based methods are used. Even for older children with CVI, it might be difficult to report in detail what they can see and can make out, in terms of useful vision. Language development may be delayed, and naming may be impaired because of semantic problems or because visual accuracy is (was) not sufficient to learn reliable associations between visual stimuli and their labels. Delayed language development may also make understanding of instruction and giving the proper verbal response difficult for the child. In these cases, behaviour-based assessment methods should be given preference. The reliable assessment of visual capacities, and their impairments, respectively, supposes at least sufficient cognitive, in particular attention, motor capacities and motivation, including visual curiosity. It is important to note that responses to visual stimuli, or performance in a visual task, which cannot be replicated, should not be taken for granted as contributing to a diagnosis.

Apart from the conventional criteria for tests, validity, reliability and objectivity, the so-called ecological validity plays an essential role in both the diagnostic assessment and treatment of functional impairments. Ecological validity comprises three dimensions, the nature of the setting, stimuli and response, and signifies (1) the degree to which the observed and recorded behaviour/performance reflects the behaviour/performance that a subject shows/possesses in the ‘real world and (2) the degree to which findings based on standardised test instruments can be generalised (or extended) to natural, everyday life settings (Schmuckler 2001). Of course, ecological validity varies between functional domains and thus assessment instruments. The diagnosis of homonymous hemianopia, for example, possesses a rather low ecological validity, because it does not include any information concerning the degree of compensation and thus the associated visual impairment. In contrast, impaired visual search performance will allow a good prognosis concerning a child’s behaviour in everyday life conditions, because reduced accuracy tells us that the child will omit targets and reduced speed indicates that the child needs much more time for the visual search process, with limited search time also resulting in neglecting parts of the actual scene.

For the acquisition of valid, significant and reliable diagnostic data as well as for comparisons in follow-up (longitudinal) assessments of the same child and between children, it is essential to keep testing conditions comparable, i.e. to assess vision in standard conditions. Standardisation includes lighting conditions, equipment of the testing room (friendly atmosphere, but not too many pictures and furniture, because the child may be distracted), type and duration of stimulus presentation and verbal and nonverbal interaction with the child. Any effective means of stimulation of the child’s motivation by the investigator is welcome, but this may also be too much for the child to cope with, due to information overload with visual, auditory and tactile stimuli; therefore, ‘less is sometimes more’. In addition, it is important to consider breaks as often and as long as required. New visual stimuli may increase attention and enhance motivation; it seems sensible, therefore, not to show the same stimulus too often successively, because the child may show habituation to this stimulus and thus fail to respond (Fantz 1964). Results, which are obtained under conditions that do not guarantee examination, can be used for preliminary screening, but not for diagnostic reports.

The examination of the various functional capacities of the central visual system should always seek not only impaired visual capacities but also those that are spared. The so-called positive and negative picture of performance is not only important to understand the quality and degree of visual disability but also for the planning of therapeutic interventions and the benefit of these interventions. The use of standardised assessment tools is, therefore, recommended; in addition, data from children with normal development of the same age, gender and socio-economic status should be available for comparison.

There is still a lack of standardised diagnostic tools. Many assessment measures that are used in developmental research cannot easily be translated into diagnostic practice; on the other hand, proven tests, for example, to assess visual acuity or contrast vision, are not very helpful for the evaluation of (the degree of) visual disability in individual everyday life conditions of the child, i.e. ecological validity is rather low (Colenbrander 2009, 2010; Boot et al. 2010; see Sect. 7.​3) when other visual disabilities are evident. The verification of spared manifest visual capacities often meets with difficulties, because assessment standards designed for typical children may have limited salience for the child whose visual system has developed differently. Here, a quasi-experimental procedure may be helpful: one factor is manipulated in a systematic and reproducible way, for example, the use of coloured instead of ‘black’ symbols and forms may be used to estimate visual acuity, because coloured stimuli can have higher salience for the specific child (exceptions: congenital colour deficiencies or cerebral dyschromatopsia). Visual acuity can also be estimated by means of moving stimuli that differ in size, because accuracy of smooth pursuit eye movements depends crucially on visually acuity (Atkinson and Braddick 1979; see Sect. 6.5.4). This quasi-experimental approach may be especially helpful if ‘normal’ assessment procedures do not produce useful results. Such deviations from the standardised assessment should, however, always be documented. In addition, it should be clearly understood that such procedures do not possess the same diagnostic quality in terms of validity, although reliability and objectivity can, however, be achieved by using this type of approach in a standardised form. Diagnostic assessment tools presented in this chapter, which have been standardised, accord with the outcome from developmental studies in children with normal development (e.g. Granrud 1993; Slater 1998a, b). Finally, it should be added that measurements of visual functions and capacities should be repeated by the examiner, and in doubtful cases by another examiner; this can increase the reliability of measurements considerably and thus also the validity of the test used (Mash and Dobson 2005).

Registration of VEPs facilitates determination of the quality and speed of processing of visual stimuli, e.g. patterned black-and-white stimuli, which are transmitted from the retina to the primary visual cortical area (striate cortex, Brodmann area 17, V1), where the electrical responses are generated (Barnikol et al. 2006). Latency and amplitude of the first positive wave (P1) are the most commonly used parameters to evaluate the electrophysiological responses. However, it must be recognised that the amplitude and form of the VEP depend crucially on fixation accuracy and concentration upon the stimulus pattern during the recording, because the major contribution to the amplitude of P1 stems from the central visual field (Cannon 1983). Furthermore, even if the form, amplitude and latency of P1 are optimal for analysis, a fundamental problem remains. These parameters characterise the electrophysiological response to a defined visual stimulus, which means that it is a kind of correlate of vision, but is neither a measure of true vision nor an assessment of visual capacity, e.g. visual acuity or contrast sensitivity (Arroyo et al. 1997). Fortunately, there exist highly positive correlations in children (and adults) between the amplitude of P1 responses to a pattern reversal checkerboard or grating stimulus and visual acuity and spatial contrast sensitivity, respectively, if an adequate range of spatial frequencies is used (0.1–30 cycles/degree; Ridder 2004; Lenassi et al. 2008; Mackay et al. 2008; Iyer et al. 2013). Pattern-evoked visual potentials have also been proved to possess high sensitivity and specificity in children with suspected functional visual acuity loss (Hamilton et al. 2013), in preterm children with subtle visual dysfunction (O’Reilly et al. 2010) and in children with spastic cerebral palsy (Costa and Ventura 2012). In addition, pattern-evoked visual potentials may also be useful in longitudinal assessment of visual function in children with CVI (Watson et al. 2007), but besides parallel courses of PL, visual acuity and VEP development, disparities in PL and VEP can also be found (Lim et al. 2005). Unfortunately, however, a sufficient quality of P1 can occur in the context of cerebral blindness. Nearly normal P1 responses after pattern stimulation have been recorded in adults with bilateral destruction of V1 and resulting severe visual disability, characterised by a visual acuity 1.3logMAR (<0.05). Apparently, a smaller number of spared neurons in V1 is required for generating P1 than are needed to ‘see’ the same stimulus pattern and discriminate pattern components (Celesia et al. 1982; Wygnanksi-Jaffe et al. 2009). Thus, the preserved P1 does not always allow a valid conclusion to be drawn about visual acuity or contrast vision; the only correct conclusion that can be drawn is that there is good transmission of visually triggered retinal signals to V1 and that there is still some neuronal activity in this visual cortical area, which taken positively could provide evidence for persisting cortical visual function. Despite the methodological and interpretational difficulties, VEP recordings may be useful in children with severe forms of CVI (van Genderen et al. 2006). In some cases, it is the only way to get diagnostic information about the morphological and physiological ‘state’ of the visual system (Taylor and McCulloch 1992), which in addition, allows a prognosis concerning the development of visual acuity in preterm children and in children with CVI (Atkinson et al. 2002a; Watson et al. 2010). Measurements can also be used as a guide to the dimensions of habilitational and educational materials to evaluate for responses and use. However, VEP vernier acuity appears more affected in subjects with CVI and is more associated with behavioural measures of grating acuity than is VEP grating acuity (Skoczenski and Good 2004; Watson et al. 2009). It should be added here that flash-evoked visual potentials are not useful to assess cortical visual function, because they are mainly generated by subcortical structures of the visual system and thus do not reflect acuity-equivalent visual processing (Schroeder et al. 1989).

In this section, assessment methods are presented for visual and oculomotor functions and capacities, which play a prominent role in visual perception and its development and which can be assessed with high validity and reliability: visual field, spatial contrast sensitivity and visual acuity, colour and form discrimination, localisation of visual stimuli, stereopsis, object and face perception and oculomotor functions including visual exploration and visual search. Thus, a comprehensive set of visual and oculomotor tools should not solely comprise visual field, visual acuity, contrast sensitivity and eye movements, although such a tool set can serve as a valid basic standard with high availability and simple use by practitioners (Powers 2009). Emphasis is on children in the first 3–4 years; for older children, assessment standards exist or neuropsychological assessment methods can be applied, which are also used with young adults (see Fischer and Cole 2000; Hyvarinen 2000; Topor and Erin 2000; Zihl 2011; Lezak et al. 2012). In older children who show severe multiple disabilities, particularly in cognition, the tests described can also be used. It should be mentioned that not all tests are sufficiently standardised and for some tests data from normal children for comparison are missing.

In cases of suspected CVI, it appears reasonable and useful to follow a standard scheme. After making the first contact with the child, including systematic observation of the child’s behaviour (which often gives the first indication of visual difficulties), the medical history, head and body posture, eyelids, poisition of the eyes and possible spontaneous eye movements (e.g. spontaneous nystagmus and gaze dependant nystagmus) are elicited. After checking the ability of the child to participate sufficiently in the assessment, anterior parts of the eye, pupil function, accommodation, vergence, ocular motility (saccades and pursuit eye movements) and eye, head and grasping movements to a simple visual stimulus are examined. Systematic assessment of visual and oculomotor functions and capacities is then performed. Figure 6.1 shows the various diagnostic steps in the visual and oculomotor assessment in suspected CVI in a schematic diagram; the proposed diagnostic procedure is also intended to help with differentiating primary and secondary visual and oculomotor dysfunction.