See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/11534936

Referral rates for functional vision screening

among a large cosmopolitan sample of
Australian children

ARTICLE in OPHTHALMIC AND PHYSIOLOGICAL OPTICS · JANUARY 2002

Impact Factor: 2.18 · DOI: 10.1046/j.1475-1313.2002.00010.x · Source: PubMed

CITATIONS READS

44 83

4 AUTHORS, INCLUDING:

Barbara Junghans David Crewther

University of New South Wales Swinburne University of Technology
37 PUBLICATIONS 703 CITATIONS 242 PUBLICATIONS 2,138 CITATIONS

SEE PROFILE SEE PROFILE

Sheila Crewther
La Trobe University
249 PUBLICATIONS 2,156 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Sheila Crewther
letting you access and read them immediately. Retrieved on: 08 January 2016
Ophthal. Physiol. Opt. 2002 22 10±25

Referral rates for a functional vision screening

among a large cosmopolitan sample
of Australian children
Barbara Junghans1, Patricia M. Kiely2, David P. Crewther2 and Sheila
Gillard Crewther2
1
School of Optometry, The University of New South Wales, Sydney, New South Wales, Australia
2052, 2School of Psychological Science, La Trobe University, Bundoora, Victoria, Australia 3083

Abstract
The aim of this study was to investigate the incidence of functional vision problems in a large unselected
cosmopolitan population of primary school-age children and to investigate whether constant clinical
criteria for functional vision problems would be implemented by the practitioners involved in the
screening. Refractive errors, near point of convergence, stereopsis, strabismus, heterophoria and
accommodative facility were assessed for 2697 children (3±12 years) of varying racial backgrounds
living in Australia. The spherical component of the refractive error ranged from )7.75 to +9.50 D (mean
+0.54 D, ‹0.79) with a distribution skewed towards hypermetropia; astigmatism ranged from 0 to
4.25 D (mean )0.16 D, ‹0.35). There was a trend towards less hypermetropia and slightly more
astigmatism with age. Mean near point of convergence was 5.4 ‹ 2.9 cm, heterophoria at far and near
was 0.12 ‹ 1.58D exophoria and 1.05 ‹ 2.53D exophoria, respectively, 0.55% of children exhibited
vertical phoria at near >0.5D, accommodative facility ranged from 0 to 24 cycles per minute (cpm)
(mean 11.2 cpm, ‹3.7), stereopsis varied from 20 to 800 s (¢¢) of arc with 50% of children having 40¢¢ or
better. The prevalence of strabismus was particularly low (0.3%).
Twenty percent of the children were referred for further assessment based on criteria of one or more of:
stereopsis >70¢¢, accommodative facility <8 cpm, near point of convergence (NPC) >9 cm, near
exophoria >10D or near esophoria >5D, shift in eso or exophoria ³ 4D between distance and near,
astigmatism ³ 1 D, myopia more than )0.75 D, or hyperopia >+1.50 D.
Post-hoc analysis of the record cards seeking the reason for further assessment indicates that referrals
appear to have been based upon clinical intuition rather than on a set number of borderline or
unsatisfactory results.

Keywords: binocular vision, clinical pro®le, referral, refractive error, vision screening

and ocular motor problems make ocular or visual

Introduction
Vision is the core sensory input for humans throughout
life and particularly during early childhood (Atkinson,
2000). Disorders such as refractive errors, amblyopia
Received: 14 July 2000
Revised form: 6 June 2001
Accepted: 26 September 2001
Correspondence and reprint requests to: Barbara M. Junghans.
Tel.: +61 2 9385 4237; fax: +61 2 9313 6243
E-mail address: b.junghans@unsw.edu.au

Formerly of School of Optometry, UNSW.
de®cits the fourth most common childhood disability
and the most prevalent handicapping condition in
childhood (cited Ciner et al., 1998). However, as such
visual anomalies are not regarded as signi®cant health
problems, vision screenings are seldom included in
medical assessments (Thomson and Evans, 1999) despite
the evidence that systematically based ocular disease
often leads to the development of refractive error
(Nathan et al., 1985, 1986), further contributing to the
visual handicap.
Although there is considerable controversy among
ophthalmic authorities as to whether such visual han-
dicaps have any in¯uence on learning as a whole (Joint

10 ã 2002 The College of Optometrists


Organizational Policy Statement of the American Acad-
emy of Optometry, 1997; Committee on Children with
Disabilities: American Academy of Pediatrics, American
Academy of Ophthalmology, American Association for
Pediatric Ophthalmology and Strabismus, 1998), edu-
cational authorities in Australia regularly seek visual
assessments for children diagnosed as having learning
disabilities with the aim of lessening near visual
discomfort and increasing the likelihood of reading
practice and preventing the development of learning
associated behavioural problems. This practice is sup-
ported by the existence of signi®cantly increased bin-
ocular visual handicaps in children with mild to
moderate intellectual disability such as Fragile X,
Velo-Cardio±Facial syndrome and Down syndrome
(Hagerman and Cronister, 1996; Crewther et al., 1998;
Tsiaras et al., 1999; Cregg et al., 2001).
The majority of evidence of a positive association
between the e€ects of vision therapy and learning is
indirect and anecdotal, however, there are some well
designed experiments showing associations between
visual function and learning diculties, particularly
reading ability (Bennet et al., 1982; Grisham and
Simons, 1986; The 1986/87 Future of Visual Develop-
ment/Performance Task Force; Simons and Grisham,
1986; Simons and Gassler, 1988; Rosner and Rosner,
1994; Garzia and Franzel, 1996; Kulp and Schmidt,
1996). Signi®cant correlations have been shown between
reading accuracy, reading comprehension, reading rate
and numeracy and one or more of unaided distance
vision, binocular distance acuity, retinoscopy, conver-
gence, phoria, ®xation disparity and colour vision
(O'Grady, 1984). Meta-analysis has also shown signi®-
cant associations between reading performance and
hyperopia, exophoria at near, vertical phoria, aniso-
metropia and aniseikonia (Simons and Grisham, 1987).
Signi®cant correlation between reading problems and
poor fusional vergence reserves, convergence insu-
ciency and ®xation disparity (Simons and Grisham,
1987), binocular instability and low amplitude of
accommodation (Evans, 1998; Evans et al., 1999), and
accommodative facility (Hennessey et al., 1984) have
also been reported. The continuing association of subtle
impairments in visual function and reading disability in
the scienti®c literature is no longer controversial
(Fowler, 1991). Twenty years of research into reading
disability has established an extensive literature on
impaired M-pathway functions in children with devel-
opmental dyslexia (Lovegrove, 1996).
Although the type of vision screening and the choice
of tools to be included should be appropriate for the
age of the individual, and in the case of a child,
appropriate for their continuing educational needs,
there is little agreement which are the most e€ective
screening methods and little scienti®c evidence of the
ã 2002 The College of Optometrists
School screenings: B. Junghans et al. 11
e€ectiveness of some screening procedures (Ciner et al.,
1998).
Previous screenings have generally been performed on
predominantly Anglo-Saxon populations and examined
refractive errors within general populations of children
(Hirsch and Weymouth, 1991; Zadnik, 1997). However
ethnicity is a factor likely to a€ect population param-
eters for various ocular characteristics (Bear, 1991;
Zadnik, 1997), e.g. the prevalence of myopia in Hong
Kong for 6 years old is 30% (Lam and Goh, 1991)
compared with a prevalence of myopia of 4% for a
similar age group in San Francisco (Zadnik et al., 1993).
Certainly the increasingly cosmopolitan nature of cities
in the UK, USA and Australia and the change in visual
demands in a society with greater need for academic
persistence, has raised the question of whether the
parameters used in the past are still appropriate.
Criteria de®ning `normal' or `satisfactory' oculo-
visual parameters should be based on ranges of values
which allow adequate and comfortable behavioural
visual functioning at far and near, rather than on
boundaries based on the statistical normal distribution.
The concept of functional limits was applied to the study
of 12 000 young Nova Scotian children by Lam et al.
(1996) who used pre-determined criteria to select 162
Nova Scotian children as `normal' to establish standards
for childhood ophthalmological variables after screen-
ing 12 000 children aged 4.5±5.5 years. Inclusion criteria
were distance acuity £6/9)3; stereoacuity £702; NPC
£ 9 cm; refractive errors between )0.50 and +3.50 D;
astigmatism £1.25 D; anisometropia £1.00 D (a range
of refractive errors consistent with unaided visual acuity
of 6/9). However, this study did not address monocular
accommodative amplitude, or accommodative-conver-
gence relationships by using accommodative facility
tests at near.
The concept of referral from a screening requires the
determination of a ®nal pass/fail, which is obviously
more complex than for a single clinical test but has
seldom been considered in the literature. Children may
show a number or cluster of borderline results, each
insucient alone to clearly delineate a fail or to yield a
clear clinical diagnosis, but taken together indicating
that the vergence system and/or the accommodative
system is under stress when performing simple binocular
tasks. Yet even reported studies of single inadequacies of
the visual system are often inconsistent in their recom-
mendations because a battery of tests and associations
considered essential by one clinician, are not accepted as
the `gold standard' by others. For example, see Rouse
et al.'s (1998) discussion of the measurement and
analysis of convergence insuciency by di€erent inves-
tigators. In addition single disorders of the visual system
are often interpreted within a research context only after
having controlled for other parameters by applying a set
12 Ophthal. Physiol. Opt. 2002 22: No 1
of `normal' standards for entry into the study (e.g.
Rouse et al., 1998).
Although Ciner et al. (1998) have recently concluded
that `there is little agreement concerning the most
e€ective screening methods and limited recommenda-
tions of referral criteria', Bailey's review (Bailey, 1998)
states that the good predictive value of the Orinda Study
and its `Modi®ed Clinical Technique' (MCT) (Blum
et al., 1968) has gained broad acceptance. However, the
MCT does not address non-strabismic binocular dys-
function and requires ophthalmic-trained personnel. In
a recent development to address issues of time, cost±
bene®t, facilities required and training of personnel,
Thomson and Evans (1999) have moved towards imple-
mentation of a screening with some computer-based
elements with a sensitivity of 93.8% and speci®city of
96.1%: screening for symptoms via a questionnaire
completed by parents at home, computer-based mon-
ocular distance acuity, computer-based stereopsis and
normal Ishihara colour vision plates.
One aim of the current study was to examine
statistical population norms for refractive errors and
selected binocular parameters believed in Australia to be
associated with comfort in reading [stereopsis, hetero-
phoria, accommodative facility and near point of
convergence (NPC)] for a large random cosmopolitan
population. The screening was implemented as part of
the clinical curriculum for students of Optometry at the
University of New South Wales. Given the constraints
of the test battery used and the student±clinic setting,
the second aim was to analyse what `operational' criteria
are used by experienced clinicians to determine referral
with regard to binocular function during a large scale
screening, with a view to re®ning the screening protocol
and planning manpower for further screenings.
Methods
Ethics
Approvals were obtained from the Committees for the
use of humans in research at the University of New
South Wales (UNSW), Sydney, and La Trobe Univer-
sity, Melbourne, Australia. Permission to approach
schools to invite them to send entire classes to the
Vision Education Centre (VEC) (Junghans and Crew-
ther, 1992), School of Optometry, UNSW was obtained
from the NSW Department of Education and the NSW
Catholic Education Oce and School Principals. In
Victoria, permission to perform vision screenings on
children as part of other reading-related studies was
obtained from the Victorian Catholic Education Oce
and The State Directorate of School Education and
school Principals. In both states parents/guardians were
provided with an information sheet detailing the vision
screening and written consent was obtained prior to a
child's participation.
Subjects
A total of 2697 children aged 3±12 years participated
in the study (2612 from NSW, 85 from Victoria),
although not all were administered all tests. There were
1452 boys and 1195 girls and 50 children for whom
gender was not recorded. In both states there was
considerable variation in ethnic background of the
children but ethical constraints prevented us from
determining ethnic background. New South Wales
children came from primary schools in the Randwick
area of Sydney where 39.5% of people in Randwick
were reported to have been born overseas, with 90.7%
of these from 36 countries covering all continents (1996
Census data: The People of New South Wales: Statistics
from the 1996 Census, 1998). In this area recent
immigrants included 42.9% from eastern or middle-
eastern Europe, 22.2% from Asia and 13.7% from
western Europe.
Protocol
Although a comprehensive eye examination comprising
of age-appropriate tests by Australian standards was
performed on most children, the term `screening' is used
as parents were not present (see Bailey, 1998). In NSW
the screening was administered by ®nal year optometry
students under the supervision of optometrists with
considerable paediatric experience. The results presented
are those of the supervisor. The Victorian screenings
were administered in schools by one of two registered
optometrists (SGC and PMK). Protocols and instru-
mentation di€ered between the two sites only in that
‹2.00 ¯ippers and a Randot Stereo Test were used in
Victoria. Parents and teachers in NSW were asked to
complete a simple questionnaire prior to the screening,
regarding the presence of signs or symptoms of oculo-
visual problems (see Appendix).
The screening included visual acuity, cover test for
strabismus, motilities, saccades, pupil reactions, retino-
scopic determination of the refractive error, NPC,
heterophoria, stereopsis (Titmus or Randot) and
accommodative facility (the latter four tests were done
with the habitual correction unless a signi®cant change in
refraction had been found), colour vision and ophthal-
moscopy. The supervising optometrist was responsible
for the recommendation for referral for further assess-
ment. The recommended guidelines for determining
borderline and failed performance on each test are
shown in Table 1. These guidelines were determined by
two senior sta€ optometrists specialising in children's
vision, who drew on the literature and their own
ã 2002 The College of Optometrists
 School screenings: B. Junghans et al. 13

Table 1. Criteria used to determine degree of acceptability of oculo-visual parameters (exo ˆ exophoria, eso ˆ esophoria)
Criterion

Function Satisfactory Borderline Unsatisfactory

Near point of convergence <7.5 7.5±9 cm >9 cm

Phoria (distance) 3 m: 2 eso ± 4 exo 3 m: >3 eso, >5 exo 3 m: >5 eso, >8 exo
Phoria (near) 33 cm: 2 eso ± 8 exo 33 cm: >3 eso, >9 exo 33 cm: >5 eso, >10 exo
Near phoria minus distance phoria <4D ± ³4
Accommodative facility >8 cpm 7±8 cpm 8 years+ <7 cpm
8 years) <6 cpm
Stereopsis £ 70 s 80 s ³ 100 s
Colour vision 0±1 plates missed 2 plates missed
Rx sphere < +1.50 DS +1.50 DS >1.50 DS
< )0.50 DS )0.50 DS >) 0.75 DS
Rx cylinder < 0.75 DC )0.75 DC >)0.75 DC
Anisometropia 1.00 DS >1.00 DS

experience (one with a background in behavioural

optometry and the other a more traditional background).
Retinoscopy without cycloplegia was performed with
the child ®xating a distant target; in NSW these ®ndings
were re®ned by subjective refraction and visual acuity
determined with the refractive correction. Near point of
convergence was assessed using the push-up method to
determine break subjectively and objectively whilst
viewing a ®ne high contrast target (objective results
were recorded where there was disagreement). Hetero-
phorias were measured at distance (3 m) and near
(33 cm) using a Howell Phoria Card (Cyclopean Design,
Melbourne, Australia) which is an adaptation of a
principle attributed to Thorington (or Prentice) (Borish,
1975). Vertical phorias were measured at near only. The
Howell Phoria Card is oval to decrease fusional cues
from horizontal/vertical borders and measures from
10 exo to 9 eso at far, and 20 exo to 21 eso at near. The
black lines bordering the top and bottom edges of the
horizontally elongated yellow and blue target are
designed to provide optimal stimulation for foveal
blur-driven accommodation (E.R. Howell, personal
communication). Stereopsis was assessed using either
the Titmus Stereotest (¯y/animals/Wirt circles) measur-
ing to 40¢¢ (NSW) or the Randot Stereotest (shapes/
animal/circles) measuring to 20¢¢ (Victoria). Accommo-
dative facility was assessed over 1 min with ‹1.50 D
¯ippers (NSW) or ‹2.00 D ¯ippers (Victoria) while the
child binocularly ®xated a row of N6 letters held at
33 cm. In UNSW, ‹1.50 ¯ippers were used to reduce
the number of young children experiencing diculties
(an interpretation of the data from Scheiman et al.,
1988). In Victoria ‹2 D ¯ippers were used at 40 cm as a
more demanding task for younger children holding
books. A zero value was recorded for children unable to
see clearly through either of the lens powers. Colour
vision was tested using Ishihara plates (1979, 1980, or
1985 versions).
Statistical analyses
Data were analysed using Statistical Package for Social
Sciences (SPSS) for Windows software. Multivariate
interpretations were analysed by ANOVA. Incomplete
data sometimes occurred because of time constraints.
This has led to varying sample sizes for di€erent
parameters.
Results
Refractive error results include all children. The stra-
bismic children were excluded from the binocular vision
analysis. Correlations of poor ®ndings with the existence
of symptoms could not be determined accurately in the
current study due to lack of information provided on
parent/teacher questionnaires. 2490 children received
sucient testing to qualify for inclusion in the ®nal case
analyses. For 1845 children, data were available on the
four aspects of functional vision: NPC, accommodative
facility, stereopsis, and the di€erence between far and
near phorias.
Refractive error
Spherical component. The distribution of the means of
right and left eyes from each subject for the spherical
component of the refractive error (the more positive
meridian, not spherical equivalent) is skewed towards
hypermetropia (see Figure 1) with an overall mean of
+0.54 D, ‹0.79 (range )7.25 to +9.25 D, n ˆ 2697
children). Eighty-nine percent of eyes had spherical
components between 0 D and +1.50 D; 97% exhibited
spherical components between )1.75 D and +2.00 D;
5.3% of eyes were myopic ³0.50 D and 7.7% of eyes
were more than 1.50D hypermetropic. Only 1.1% of
eyes exhibited a myopic spherical component ³2.00D;
2.0% of eyes exhibited hypermetropia >+2.00 D. One

ã 2002 The College of Optometrists

14 Ophthal. Physiol. Opt. 2002 22: No 1

Figure 2. Distribution of the mean cylindrical component of the

refractive error from the two eyes of each child.

Figure 1. Distributions of the mean spherical component of the

refractive error from the two eyes of each child.
in 14 of all children assessed (7.1%) already had been
prescribed spectacles, and one of ®ve (20.9%) were
recommended to return to their practitioner for a review
of their prescription as a change of at least 0.50 dioptres
appeared necessary.
There was a high correlation (p < 0.0005) between the
values for mean spherical component for right and left
eyes at all ages (see Table 2), with only 1.4% of 2697
children showing anisometropia >1.00 DS. There was a
trend towards less hypermetropia with age.
Astigmatism. The distribution of the average cylindri-
cal component for right and left eyes (n ˆ 2697) has a
mean of )0.16 D, ‹0.35 (range 0 D to )4.25 D) (see
Figure 2). Seventy-one per cent of eyes exhibited
negligible astigmatism (i.e. £0.25 DC) and 27%
showed astigmatism between )0.25 D and )1.0 D.
There was moderate to high correlation between the
amount of astigmatism in right and left eyes for all age
groups (see Table 2). Although the older age groups
appeared to demonstrate more astigmatism, there was
only low correlation with age (r ˆ )0.041, p < 0.05)
and the magnitude was below the minimum step used
clinically.
Of the 1680 (28.5%) eyes with astigmatism >0.25 D
in the current study, approximately 55% exhibited
with-the-rule astigmatism (WTR: negative axis within
30° of the horizontal), nearly 8% exhibited oblique
astigmatism (axis 30±60° or 120±150°) and the remain-
ing 38% exhibited against-the-rule astigmatism (ATR:
axes 60±120°). Over 80% of children with measurable
astigmatism in both eyes (n ˆ 589 children) showed
similar axes in the two eyes (r ˆ 0.724, p < 0.01).
48.9% had binocular WTR astigmatism; 0.9% had
binocular symmetrical (mirror-image) oblique astig-
matism; 33.2% had binocular ATR astigmatism; 6.1%

Table 2. Spherical and cylindrical components of the refractive error and correlation between right and left eyes
Sphere Cylinder

Age RE DS LE DS RE DC LE DC
(years) (Mean ‹ S.D.) (Mean ‹ S.D.) Correlation (r) (Mean ‹ S.D.) (Mean ‹ S.D.) Correlation (r)

3 0.85 ‹ 0.38 (12) 0.83 ‹ 0.36 (12) 0.898* )0.02 ‹ 0.07 (12) )0.02 ‹ 0.07 (12) 1.00*
4 0.59 ‹ 0.52 (123) 0.62 ‹ 0.54 (123) 0.850* )0.15 ‹ 0.25 (123) )0.13 ‹ 0.21 (123) 0.619*
5 0.64 ‹ 0.53 (348) 0.66 ‹ 0.61 (347) 0.743* )0.17 ‹ 0.39 (348) )0.15 ‹ 0.30 (347) 0.712*
6 0.56 ‹ 0.59 (221) 0.57 ‹ 0.57 (221) 0.883* )0.15 ‹ 0.32 (221) )0.15 ‹ 0.29 (221) 0.690*
7 0.64 ‹ 0.79 (330) 0.64 ‹ 0.72 (330) 0.796* )0.15 ‹ 0.27 (330) )0.13 ‹ 0.34 (330) 0.469*
8 0.60 ‹ 0.91 (471) 0.59 ‹ 0.80 (470) 0.853* )0.16 ‹ 0.35 (471) )0.16 ‹ 0.37 (470) 0.741*
9 0.52 ‹ 0.80 (214) 0.58 ‹ 0.93 (212) 0.873* )0.16 ‹ 0.38 (214) )0.14 ‹ 0.36 (212) 0.845*
10 0.45 ‹ 0.80 (280) 0.48 ‹ 0.87 (279) 0.929* )0.15 ‹ 0.33 (280) )0.14 ‹ 0.35 (279) 0.710*
11 0.36 ‹ 0.86 (532) 0.41 ‹ 1.02 (531) 0.875* (530) )0.18 ‹ 0.36 (532) )0.16 ‹ 0.37 (531) 0.712* (530)
12 0.37 ‹ 0.77 (166) 0.42 ‹ 0.91 (166) 0.916* )0.25 ‹ 0.50 (166) )0.21 ‹ 0.41 (166) 0.843*
all 0.53 ‹ 0.76 (2697) 0.55 ‹ 0.82 (2691) 0.859* (2690) )0.17 ‹ 0.35 (2697) )0.15 ‹ 0.35 (2691) 0.711* (2690)
Numbers in brackets indicate number/pairs of eyes; *denotes signi®cant correlation: p < 0.0005.

ã 2002 The College of Optometrists

 School screenings: B. Junghans et al. 15

had WTR in one eye and ATR in the other. The
remaining 10% had asymmetrical oblique axes in the
two eyes.
Near point of convergence
Mean values for NPC for di€erent age groups are shown
in Table 3 and the distribution is shown in Figure 3. The
NPC (break) ranged from 2 cm (`to the nose') to 30 cm
(mean 5.4 cm, ‹2.9, n ˆ 2524). Analysis of the variance
across all age groups indicated that there was no main
e€ect for NPC to vary with age, although there was a
mild trend for NPC to be more remote with increasing
age. However, 17.2% of children had NPC ³7.5 cm,
11% exhibited values ³9 cm and 2.6% of children
exhibited values ³12 cm.
Heterophoria and strabismus
Only eight children (0.30% of 2697 children) were
reported to have strabismus (one speci®ed as intermit-
tent). The distributions for distance horizontal hetero-
phoria, near horizontal heterophoria and the di€erence
between the two are shown in Figures 4±6. Distance
horizontal phorias showed a leptokurtotic distribution
with a mean of 0.12D exo, ‹1.58 (see Table 3) with
range 10D exo to 20D eso, n ˆ 2239. 1.0% of the children
were >5D exo at 3 m and 0.3% >8D exo; 27.3% dem-
onstrated some degree of eso; however, only 0.5%
were >5D eso. The mean phoria at near was 1.05D exo,
‹2.53 (n ˆ 2282) (see Table 3) with a range from
22D exo to 21D eso. At near, 11.9% children displayed
>3D eso; 1.3% of children had exo >9D, and 0.5% had
>5D eso.
Five children of the 896 (0.55%) tested for vertical
phoria at near were found to have latent deviations
greater than 0.5D. Only one of these was classi®ed as
strabismic (9D hyper).
The mean di€erence between far and near horizontal
phorias was relatively exo (0.92D ‹ 2.34, n ˆ 2126) (see
Table 3) with a range from 20 more exo to 20D more eso
(Figure 6). The number of children exhibiting a far to
near exo shift ³4 was 212 (9.9%), and 23 (1.1%) showed
an eso shift >4. An eso shift from far to near of any
magnitude was found in 16.7% of children. There was
no signi®cant change in phoria measures at far or near
with increasing age.
Twenty-nine children exhibited horizontal phorias
³10D exo at near (1.3% of 2280), and 33% of these also
revealed an NPC greater than 10 cm. Of the 212 (10%
of 2126) children with a distance to near exo shift ³4, 60
(20.3%) exhibited an NPC ³7.5 cm (2.1% of total).
14.2% of myopes ³)0.50 D were eso at near.
Stereopsis
The distribution of stereopsis ®ndings from 2697 chil-
dren is shown in Figure 7. The median value for
stereopsis was 40¢¢. In this situation the median is more
representative of central tendency than the mean
because of the skewed non-parametric demand of the
tasks in the Titmus ¯y and Randot tests. However,
26.9% of children showed stereopsis less than 70¢¢.
Notably, ®ve of the eight children listed as strabismic
recorded some level of stereopsis (2 at 400¢¢, 1 at 200¢¢, 1
at 100¢¢ and 1 at 40¢¢). In addition, four of the ®ve
children with signi®cant vertical phoria (all around 2D)
recorded between 40¢¢ and 140¢¢ stereopsis. There was a
mild improvement in stereopsis with age (p ˆ 0.0194).
Accommodative facility
The mean values for the accommodative facility of the
NSW children using ‹1.50 D ¯ippers are shown in
Table 3, and the distribution of these is shown in
Figure 8. The addition of the results of the 79 Victorian

Table 3. Mean values for near point of convergence (NPC), distance phoria, near phoria and the difference between the near and the distance
phorias, and accommodative facility for different age groups
Age Accommodative
(years) NPC (cm) Distance phoria (D) Near phoria (D) Near±distance phoria (D) facility (cpm)

3 4.2 ‹ 2.1 (11) ± ± ± 9.3 ‹ 0.6 (3)

4 4.6 ‹ 3.1 (106) 0.15 ‹ 1.41 (62) )0.74 ‹ 2.17 (72) )0.89 ‹ 2.43 (62) 11.4 ‹ 3.6 (33)
5 4.7 ‹ 2.2 (310) )0.20 ‹ 1.28 (304) )1.24 ‹ 2.05 (267) )1.04 ‹ 2.18 (225) 11.3 ‹ 3.4 (197)
6 5.0 ‹ 2.4 (208) )0.15 ‹ 1.36 (179) )1.07 ‹ 2.09 (196) )0.86 ‹ 2.03 (178) 11.0 ‹ 3.2 (175)
7 5.5 ‹ 3.2 (310) )0.15 ‹ 1.46 (309) )1.42 ‹ 2.94 (332) )1.19 ‹ 2.50 (308) 10.8 ‹ 3.5 (303)
8 5.3 ‹ 2.7 (443) )0.20 ‹ 1.49 (376) )0.90 ‹ 2.32 (382) )0.76 ‹ 2.23 (363) 11.1 ‹ 3.8 (427)
9 5.4 ‹ 2.9 (200) )0.03 ‹ 2.01 (175) )0.83 ‹ 2.79 (183) )0.82 ‹ 2.40 (172) 11.2 ‹ 3.7 (199)
10 5.6 ‹ 2.7 (263) )0.16 ‹ 1.57 (189) )0.93 ‹ 2.72 (189) )0.72 ‹ 2.25 (185) 11.0 ‹ 4.1 (259)
11 6.0 ‹ 3.1 (518) )0.03 ‹ 1.83 (468) )0.93 ‹ 2.63 (485) )0.92 ‹ 2.53 (461) 11.3 ‹ 3.7 (494)
12 6.5 ‹ 3.5 (155) )0.21 ‹ 1.29 (177) )1.12 ‹ 2.52 (176) )0.93 ‹ 2.24 (172) 11.8 ‹ 3.7 (156)
All 5.4 ‹ 2.9 (2524) )0.12 ‹ 1.58 (2239) )1.05 ‹ 2.53 (2282) )0.92 ‹ 2.34 (2126) 11.2 ‹ 3.7 (2370)
Numbers in brackets refer to the number of children.

ã 2002 The College of Optometrists

16 Ophthal. Physiol. Opt. 2002 22: No 1

Figure 3. Distribution of values for the near point of convergence

(cm). Figure 6. Distribution of values for the difference between distance
phoria and near phoria (D). A negative value indicates that the near
phoria is relatively more exophoric than the distance phoria.

Figure 4. Distribution of heterophoria values at distance (D) (neg-

ative values indicate exophoria, positive values indicate esophoria).
Figure 7. Distribution of stereopsis values (¢¢arc).

Figure 5. Distribution of heterophoria values at near (D) (negative

values indicate exophoria, positive values indicate esophoria). Figure 8. Distribution of values for accommodative facility (cpm).

ã 2002 The College of Optometrists


children (for whom ‹2 ¯ippers were used) does not
alter the values. The mean values for NSW and Victoria
did not di€er signi®cantly from each other, being
11.2 ‹ 3.7 cpm (range 0±24 cpm, n ˆ 2370) and
10.2 cpm, ‹4.1 (range 0±24 cpm, n ˆ 79), respectively,
with an overall mean of 11.2 cpm.
Of the 416 children who had NPC ³ 7.5 cm and for
whom accommodative facility was also measured, 41 or
9.8% had accommodative facility <7 cpm (1.8% of the
2303 who had both functions measured). Of the 72
children with signi®cant near phoria (>10 exo or
>4 eso) only ®ve children (6.9%) rated poorly on
accommodative facility (<7 cpm).
Colour vision
One hundred and seventy-nine of 2330 (7.68%) children
failed more than two colour plates on the Ishihara Test.
Analysis of groupings of functional vision problems
The frequency with which children failed one or more of
the test of binocular vision is presented in Table 4. Of
the 1845 children who completed all four binocular tests,
39.2% (724 children) demonstrated inadequate function
on one of the tests: 17% exhibited unusual phoria shifts
from distance to near (11.3% a relative convergence,
5.7% a relative divergence of ³ 4), 13.3% had stereopsis
³80¢¢, 6.8% had binocular accommodative facility below
7 cpm, and 2.2% had a NPC > 10 cm.
Two hundred and ten children (11.4% of 1845) failed
to meet clinical criteria for two aspects of functional
vision: stereopsis was poor in 74.9% of these cases (158),
vergence errors from far to near were present for 63.6%
(134) of these cases, and accommodative facility was
poor in 43.1% (91) of the 210 cases. Only 2.1% of the
1845 (n ˆ 39) performed poorly on three tests of
binocularity; of these 51.3% (n ˆ 20) showed poor
accommodative facility, poor stereopsis, and excessive
vergence shifts from far to near, whilst almost a further
quarter (23.1%, n ˆ 9) manifested the same cluster of
problems except that the vergence change was excess-
ively exophoric. A poor performance on all four
binocularity tests was found in only 0.3% of the 1845
Table 4. Failure rates on near point of convergence, heterophoria,
stereopsis and accommodative facility for the 1845 children
assessed for all four binocular parameters
Number of tests failed n %
None 867 47.0
One 724 39.2
Two 210 11.4
Three 39 2.1
Four 5 0.3
ã 2002 The College of Optometrists
School screenings: B. Junghans et al. 17
children (n ˆ 5). In summary, 978 children (53.0% of the
1845) demonstrated between one and four poor results
on the battery of four tests of functional vision.
Decision to take further action
In all cases the decision to refer was made by the
supervising optometrist. This decision was based on the
complete clinical pro®le of each child; that justi®cation
of the reasoning was not recorded. Reports sent to
parents recommending further assessment classi®ed
problems in broad layperson's terms, e.g. `focusing',
`eye co-ordination', or `ocular health', rather than giving
a speci®c diagnosis. Thus, the referral data shown in
Table 5 represent a post hoc appraisal of the record
cards for the 495 children referred (19.9% of 2490).
Uncorrected anisometropesP1.00 DS made up 3.0%
of those referred. Poor binocular co-ordination alone
was one reason for referral (42.0% of 495, 8.35% of
2490). However, 68 children who were referred for
signi®cant refractive error also showed mildly impaired
binocularity on at least one test and 116 referred for
refractive error showed unsatisfactory binocularity on
at least one test. In particular, of those who were
signi®cantly anisometropic or astigmatic (n ˆ 98) there
were 78 (80%) who had at least one poor aspect of
functional vision. For 44 children the reason for referral
was not clear from the record card, and for 11 (0.4% of
the total number of children seen) there was ocular
pathology.
Table 6 summarises the ®ndings for the 2490 children
considered to have received a full `screening'. Of these,
177 (7.1%) were already wearing spectacles, 37 of whom
(20.9% of those already wearing spectacles) were in need
of a re®nement to their refractive correction of at least
0.50 D. A further 34 children were receiving care: three
apparently for monitoring of low myopia, 12 for
monitoring hypermetropia P 1 D, seven for previous
pathology (representing 0.3% of 2490), 10 probably for
binocular anomalies (deduced either by the child
describing the prescription of `eye exercises' or from
their poor performance at our screening) and two
children reported being examined for colour vision
defects.
In summary, further assessment was recommended
for 19.9% of the total of 2490 children screened with
8.47% reporting being under care already.
Discussion
The data presented in the current study re¯ect real
population norms as the population is non-self-selecting
as opposed to norms established from a clinical prac-
tice. This study aimed to report not only the parameters
of individual tests for a group of children with a large
18 Ophthal. Physiol. Opt. 2002 22: No 1

Table 5. Analysis of ocular parameters for

% of 495
all children deemed to require further
Parameter n children
assessment
Already in spectacles 37 7.47
No previous Rx given 458 92.53
Refractive error Hyperopia > 1.50D 81 16.36
Hyperopia of 1.50D 21 4.24
Myopia > 0.75D 102 20.61
Myopia of 0.50D 15 3.03
Anisometropia > 1.00 DS 37 7.47
Astigmatism > 0.75 DC 44 8.89
Astigmatism of 0.75 DC 18 3.64
with binocular vision problems 164 33.13
Binocular vision problems 1 borderline result 92 18.59
2 borderline results 46 9.29
1 unsatisfactory result 94 18.99
2 unsatisfactory result 29 5.86
1 borderline + 1 unsatisfactory 60 12.12
>2 borderline or unsatisfactory 50 10.10
Pathology, colour vision or unclear 68 13.73

Table 6. Summary of children either refer-

Number of % of 2490
red for, or presenting with, signi®cant
children complete cases
visual problems
Referred refractive error alone 52 2.09
Referred refractive error and binocular vision anomaly 167 6.71
Referred binocular vision anomalies alone 208 8.35
Referred other reason (pathology, colour vision only, 68 2.73
or unclear from record card)
Subtotal 495 19.88
Already under care, not wearing spectacles 34 1.36
Already under care, wearing appropriate spectacles 140 5.62
Total 669 26.86

and diverse ethnic base, but also to investigate reasons
for referral and examine the referral rate due to
abnormal binocular function on a particular set of tests
which are intrinsically likely to a€ect academic per-
formance.
The screening was carried out by optometry students
who were within 6 months of graduation. The fact that
this screening was part of a programmed school
excursion might easily have led to boredom and
distractions, and hence more readily revealed anomalies,
or, it may have led to subtleties being missed as a result
of increased visual interest and attention engendered by
the surroundings. In addition, the battery of tests was
designed to re¯ect Australian practice in the early 1990s,
and the needs for our students of optometry to
experience paediatrics. The battery was not targeted as
`the optimum' screening on a cost±bene®t basis,
although some of the conclusions might have applica-
tion to the design of such a test battery.
The non-English speaking background of most of the
families in this study diminished the usefulness of
presenting symptoms as part of a problem-based
approach to case analysis and may have impacted on
referral rates. There was heavy reliance on class teachers
to identify the child's symptoms. This does not detract
from the results presented, but emphasises the need for
school screenings in multi-racial settings to be wisely
planned and properly marketed to ensure sound out-
comes.
Refractive errors
Refraction and age. A relatively predictable change in
refractive status occurs during childhood (Hirsch and
Weymouth 1991). A similar result has been found in the
current investigation.
Spherical component. One in every 14 children seen was
already wearing a refractive correction, but one of ®ve
of these in our opinion were no longer adequately
corrected. Eighty-nine percent of the children screened
aged between 3 and 12 years demonstrated a spherical
component for the refractive error (expressed in negative
cylinder form) between 0 and +1.50 D, 2% showed
hypermetropia greater than +2.00 D, 5.3% were ³ 0.50
myopic and only 1.1% showed myopia ³ 2.00 D.

ã 2002 The College of Optometrists


Morgan and Kennemer (1997) analysed photorefrac-
tion photographs of 14 000 US primary-school children
of which 4.6% displayed signi®cant anomalies (hyper-
opia ³2.5 D, myopia ³1.0 D, anisometropia ³1.0 D,
media opacity ³1.0 mm in diameter, ocular misalign-
ment ³5°/10 prism dioptres), and a further 9.7% showed
possibly signi®cant ®ndings (same signs but lesser
degree). Of these, 4.5% were myopic, 3.1% anisome-
tropic and 2.9% hypermetropic. Lam et al. (1996)
studying children in Nova Scotia, reported mean values
of +0.69 D, ‹0.53 for the spherical component
(+1.01 D using negative cylinder form). Ninety-six
percent of eyes were between +0.07 and +2.07 (using
negative cylinder form); indicating a skewing towards
hyperopia. Moreover, 48.1% of children had no meas-
urable di€erence between the two eyes for the spherical
component, while 98.1% exhibited less than 0.75 D
di€erence. In comparison, our ®nding that 97% of
errors lie between )1.75 D and +2.00 D indicates a
considerable myopic shift in our population which may
be attributable to the high number of Asian children in
our study.
Other Australian studies of non-selected children have
included assessment of refractive errors. Walters
(1984a,b) studied a group of 5597 Victorian rural
children aged 4±14 years (mean 10.4 ‹ 1.2 years) from
predominantly Caucasian low socio-economic back-
grounds. Refractive errors ranged from high plus to
high minus, however, 90% showed retinoscopic ®ndings
between )0.25 and +1.25 D (mean +0.39 D, ‹0.89).
Similarly, Amigo et al. (1976), studying rural underpri-
vileged children from NSW (n ˆ 1166), found 91% with
retinoscopic ®ndings between 0.00 and +1.25 (mean
approximately +0.5 D). By contrast, Robbins and
Bailey (1975) studied a relatively high socio-economic
group of children (n ˆ 1243) aged 3±12 years, and found
that only a total of 6.6% were myopic ³0.50 D and only
9.0% were hyperopic ³1.50 D, which raises the question
whether the di€erent ethnic background for our popu-
lation is as signi®cant a refractive problem as one might
have anticipated.
Astigmatism. The degree and orientation of astigmatism
are important components of refractive error, although
many studies mask the degree of astigmatism by
reporting only the equivalent sphere (Hirsch and Wey-
mouth, 1991). In the current study, 71% of all eyes
exhibited negligible astigmatism and only 2% of all
children had astigmatism greater than 1.00 DC. This
supports the observations that over 72% of eyes of
children exhibit no astigmatism (Hirsch, cited in Lyle,
1991), and that over 90% have less than 1.00 D of
astigmatism (Lam et al., 1996). Similarly, only 8% of
non-clinically sourced Australian children aged 5±
14 years showed greater than 0.50 D of astigmatism
ã 2002 The College of Optometrists
School screenings: B. Junghans et al. 19
(Amigo et al., 1976). The tendency for an increase in
astigmatism with age (0.23 D cyl from age 3±12) in our
group of children was similar to that observed by Hirsch
(cited in Lyle, 1991) who noted that 81% of 6 year olds
but only 72% for 12 year olds exhibited negligible
astigmatism. Like the study of Lyle (1991), the current
study found that more eyes of school-age children (55%)
exhibit with-the-rule refractive astigmatism than
against-the-rule astigmatism. However, a recent study
by Chua et al. (2001) reports that in Singaporean school
children aged 7±9 years almost 94% have with-the-rule
astigmatism if astigmatism is present.
Functional vision
As this project comprised of an one-o€ screening of each
child, the data portrays the status of near vision
captured as a re¯ection of each child's habitual vision,
though, for the few for whom a signi®cant (>0.50 D)
change in refractive error was detected, the functional
tests were carried out through the new prescription.
Therefore, 1.4% of the results presumably showed less
binocular problems than had been operating in the
general population, though this may not necessarily be
the case (Dwyer, 1992).
It might be argued that the battery of tests selected for
use at VEC UNSW does not represent the best selection
of tests to detect unstable binocularity for near vision,
however, at the time of the design of the screening this
was not the primary aim.
In retrospect for example, it could be argued that the
aligning prism required to eliminate any ®xation dis-
parity on a Mallett unit (associated phoria) as predom-
inantly used in the UK should be measured in preference
to dissociated phoria. Wick and London (1987) argued
for the use of ®xation disparity testing, amongst other
tests, to elucidate which component of binocularity
contributes most to visual distress. Yekta et al. (1989)
also found a relationship between near point symptoms
and associated phoria using a modi®ed Mallett unit,
rather than between dissociated heterophoria and symp-
toms. They studied a sample of 187 Iranians aged 10±
65 years visiting an optometry clinic, although in that
study almost no ®xation disparity (0.09 ‹ 0.78¢) was
found in those aged under 20 years. By comparison,
Sheedy and Saladin (1977) found that dissociated
heterophoria was a better discriminator than ®xation
disparity for esophores with symptoms amongst optom-
etry students, although their later study favoured the use
of ®xation disparity (Sheedy and Saladin, 1978). The
clinic in the School of Optometry, UNSW, did not use
®xation disparity as a tool, as Alexander (1990) found
that there was a 7 min of arc calibration error on the
Sheedy Fixation Disparometer. Furthermore, it is well
known that ®xation disparity may be present with an
20 Ophthal. Physiol. Opt. 2002 22: No 1
absence of symptoms (Jenkins et al., 1989). This clouds
the usefulness of ®xation disparity as a screener in a
population of children for whom the symptomatology is
unclear and when time is a factor. We cannot comment
on the ecacy of the tests chosen for the screening, as
our cohort of children have not been followed to
determine whether the recommendations to refer were
satisfactory or not.
However, it is hoped that the discussion which follows
on each of elements of the screening, plus the analysis of
the clusters of abnormal results along with the decision
to refer or not, helps in shaping decisions of how to
handle mass vision screenings which target functional
vision.
Strabismus. The presence of strabismus in this cohort of
children was found to be extremely low (0.3%) com-
pared with other studies, e.g. 1±2% (Ciner et al. 1998)
up to even 8% (The 1986/87 Future of Visual Develop-
ment/Performance Task Force, 1988). A probable
explanation is that our supervising clinicians are reluct-
ant to apply the term `strabismic' if children with
misalignment of the visual axes as indicated by the cover
test are able to compensate sucient to demonstrate
stereopsis of 100¢¢ arc or better.
Stereopsis. The median value for stereopsis in the
current study using the Titmus (contoured) Stereotest
was 40¢¢ of arc (range 20±800¢¢). 73.1% of children had
stereopsis of 70¢¢ or better, which is considered normal
by Simons (Simons 1986) and Lam (Lam et al. 1996),
and better than normal by Osipov (Osipov 1996). Lam
et al. (1996) found that 83.3% of children achieved 50¢¢
or better using the Randot Stereotest (circles) which is a
local test measuring down to 20¢¢. However, in the main
we used the Titmus Fly test which measures only down
to 40¢¢. Thus there is a possibility that the Titmus Fly
test may underestimate stereopsis in children. The values
of 20¢¢ in the current study were obtained for Victorian
subjects where the Randot test was used.
Near point of convergence. In the current study 82.8%
of children could maintain single vision as close as
7.5 cm using an accommodative target and 89% of
values for NPC (break) were less than 10 cm. Walters
(1984a) similarly found that 89% of schoolchildren have
NPC £ 8 cm. However, Hayes et al. (1998) found that
85% of elementary school children who had already
passed a Modi®ed Clinical Technique screening had
NPC £ 6 cm when taken as an average of three repeti-
tions using a target on nearpoint rule, with a mean of
approximately 4 cm. They concluded that a cut-o€ of
6 cm should be used for NPC break in a clinic setting,
but that a looser 6±10 cm pass/fail criterion should be
used in screening situations where the coexistence of
symptoms often cannot be clearly established. They
noted the relevance that the input of accommodative
vergence plays in the choice of target, and hence NPC. A
possible reason that our ®ndings yielded a more remote
average NPC of 5.7 cm is that this test was one of a
large battery of tests in the screening (and often
performed after other tests), whereas in Hayes et al.
(1998) normative study of NPC, the attention of the
clinician and subjects was more clearly on the accurate
conduct of the test.
Heterophoria. The data presented here represent the
largest evaluation of horizontal heterophoria using the
distance (3 m) and near (33 cm) Howell Phoria Cards
on children, and concur with the unpublished data
collected by Howell himself (personal communication).
Distance lateral phoria was marginally exophoric, and
near was less than 1.5 more so. The results were not
signi®cantly di€erent from those of Walline et al. (1998),
who used subjective evaluation of the cover test on 1495
children in kindergarten, second and ®fth grades. Both
studies have shown that nearly all primary-school
children are orthophoric sofar, and that there is a slight
exo but more variable shift for near. As in Walline's
study, there was no evidence of signi®cant change of
distance phoria across age groups, and we found also
that 14% of myopes are esophoric at near. In another
Australian study, the majority of eyes of school-aged
children exhibited a low heterophoria: between 1 eso
and 4 exo at near, with the peak of distribution also at
1 exo (Walters, 1984a).
A signi®cant increase in exophoria when viewing at
near is central to the diagnosis of convergence insu-
ciency according to Duane (reviewed Rouse et al.,
1998). Overall 9.9% of children in the present study
manifest a near exophoric shift ³ 4. These ®gures
compare well with those of Rouse et al. (1999), who
found 7.4% of white children 9±13 years old and greater
proportions of black, Hispanic or Asian children
(although few children were of Asian origin) presented
with low suspect CI (near exophoria plus one of:
excessive relative exophoria at near, a remote NPC or
poor fusional reserves). However, only 2.1% of 2078
children in our study also exhibited a near point of
convergence ³ 7.5 cm, which falls between the divisions
of low and high suspect convergence insuciency by
Rouse et al. ¢s criteria. This appears to be substantially
lower than the ®gure of 8.7% for white children of
high suspicion for having CI in Rouse's study, and
considering the high numbers of Asians in our sample,
very di€erent to their observation that Hispanics and
Asian are even more likely to have CI. Rouse et al.
(1998) reported a frequency of 8.4% for `de®nite'
convergence insuciency in New York (428, 8±12 years
olds), but a frequency of only 4.9% in Southern
ã 2002 The College of Optometrists

California and considered that the di€erence was the
result of the di€erent ethnic proportions in the two
groups with a greater proportion of African American
(53%) and Hispanics (23.3%) in New York than in
Southern California (3.8 and 9.2%, respectively). In
non-clinical populations of children aged 9±13 years
in three states in northern America Rouse et al. (1998)
found that only 4.2% were de®nite CI (near exophoria
and ³ 4 exophoric from far phoria, NPC > 7.5 cm
break, poor fusional reserves), 8.8% were low suspects
and 8.4% were high suspects. There was a signi®cant
ethnic di€erence with respect to classi®cation of CI (9%
were Afro-American, 5% Hispanic and a little over
1% Asian-Paci®c). Using the same classi®cation, only
2.1% of the children in the current study are de®nitely
CI and 5.9% are low CI. These di€erences may re¯ect
the di€ering ethnic diversities of the populations (there
are few Afro-American or Hispanics in Australia but
many Asians), or it may be because of the di€erent age
range in the two studies.
Accommodative facility. Accommodative facility is a
measure of the continuing ability of accommodation to
change rapidly in response to stimulation and inhibi-
tion when vergence is ®xed. In the current study using
‹1.50 ¯ippers the mean value for accommodative
facility was 11.2 ‹ 3.7 cpm (2330 children) with a
range from 0 to 24 cpm. Low values of accommodative
facility have been associated with symptoms related to
near-point asthenopia in young children (Siderov,
1990). Mean binocular accommodative facility (‹2 D
¯ippers, polarised acuity suppression target) has been
reported to be 7.7 ‹ 5.2 cycles per minute (cpm) for
subjects aged 18±30 with normal binocular vision
(Zellers et al., 1984), whereas much lower results have
been reported for over 500 children by Scheiman et al.
(1988) who found 5 ‹ 2.25 cpm binocularly for the 8±
12 years olds and 3 ‹ 2.5 binocularly for the 6 years
olds when the children were required to read out
several letters as con®rmation that sucient clarity had
been achieved. McKenzie and colleagues used similar
conditions to those of Zellers et al. (1984) and found
that 26% of a group of 140 children aged 8±12 years,
failed accommodative facility when the criterion was
8 cpm or better (McKenzie et al., 1987). The mean for
binocular accommodative facility in the current study
was signi®cantly higher than the 7±8 cpm using
‹2.00 D ¯ippers noted by McKenzie et al. (1987),
although only a high contrast black and white near
letter chart with no suppression check was used in our
screening.
To clarify the interpretation of our data, an addi-
tional study was conducted to compare binocular
accommodative facility for three di€erent targets: black
and white letters (unpolarised), the Bernell vectographic
ã 2002 The College of Optometrists
School screenings: B. Junghans et al. 21
acuity/suppression slide (VO/9), and a bar-reader with a
di€erent age-appropriate word scrolled into view after
each ¯ip. The bar reader consisted of a 2-mm wide
vertical bar placed approximately 30 mm in front of a
square box with a 15 ´ 6 mm window which allowed
scrolling of single words drawn from a list of 16 words
taken from the composite Dolch and Hillerich Basic
Sight Word list each of six letters and printed in upper
case ®ve point New York font with a space between
each letter. Only children aged 9±12 years (n ˆ 72) who
had passed all aspects of the VEC examination other
than accommodative facility were included. Testing was
over 1 min for each target, with randomised order of
targets.
Analysis of variance across the three versions of the
accommodative facility test indicated that the order of
presentation of the tests was insigni®cant, as was the
di€erence between binocular accommodative facility
with ‹1.50 ¯ippers using a polarised target (mean
8.0 ‹ 3.5 cpm) and a bar reader (mean 8.8 ‹ 3.1 cpm).
However, the simple black and white target without
suppression checks resulted in a mean facility of
11.0 ‹ 3.6 cpm which was signi®cantly di€erent
(p < 0.05) from the other two versions of the test
(Fisher LSD protected Student t-test). This demon-
strates that targets without a suppression check lead to
less demand on the patient to perform the task under
conditions of true binocularity, therefore resulting in
measures of accommodative facility which more closely
resemble monocular values. Thus we believe that a
comparison between accommodative facility testing
using targets with and without suppression checks, with
due regard for a learning e€ect, may be useful in
assessing the degree to which a child suppresses under a
visually demanding task.
It is not possible to ascertain why only 68% of
children had all four functional tests completed,
although there is a trend for fewer of the youngest
children to be in this group (especially those under the
age of 6 years). One could speculate that results of
qualitative observations during the screening drove the
inclusion of near point testing for some, but resulted in a
`reasoned' omission for others.
The current study has shown that 53% of the children
who completed the battery of four functional tests had
at least one poor ®nding. However, only 13.8% had at
least two problems and 2.4% at least three. This result is
considerably higher than the prevalence of any form of
binocular defect reported by Walters (1984a), who
found that only of 36.7% failed one binocular vision
test in a group of probably less racially diverse Austra-
lian children undergoing a similar range of tests.
However, Walters used much more stringent criteria
for accommodative facility. The association of binocular
dysfunction with refractive errors in the present study
22 Ophthal. Physiol. Opt. 2002 22: No 1
was similar to that found by Dwyer (1992) in a clinic-
based setting.
Decision to take further action. A retrospective analysis
seeking why children were referred is presented in
Tables 4 and 5. Despite 53% of children having failed
at least one aspect of functional vision testing, another
8.8% requiring correction of signi®cant refractive error
and 2.6% requiring counseling for defective colour
vision, only 19.9% of the 2490 were deemed to have
failed the screening and require further examination.
The most straightforward decision to refer is for those
requiring refractive correction (8.8% referrals), yet some
children were handled contrary to the stated guidelines.
Regarding analysis of vision at near, when to refer is less
clear. Should one borderline result alone demand
further assessment? As there was no pattern as to
whether a certain number of borderline results in a
referral, it appears that decisions to refer have been
based on clinical intuition.
The addition of children already under care but not
requiring immediate attention, brought the number who
failed the screening guidelines to 26.86%, which is
similar to the 20% reported for males and 26% for
females after MCT examination of a whole primary
school by Coleman (1970), and to the 29.2% found by
Robbins and Bailey (1975) in a large study of children
aged 3±11 years (also using the MCT), and the 28.1%
found by Dwyer (1983) using an expanded version of the
MCT that included accommodative-convergence tests.
In some instances (approximately 2%), it was not
possible to infer from the record card exactly why the
child was referred for further assessment. This highlights
the diculties for the supervising optometrist in articu-
lating the reason why a consultation has been brought to
a particular conclusion, given that the parent is not
present, and, as is the case in most screenings, the
optometrist has not himself carried out all of the testing.
Indeed, when constructing the Portsea Modi®ed Clinical
Technique (Dwyer, 1983) for use at summer camps, it
was decided that it was more appropriate to refer based
upon the clinician's opinion that `the child would bene®t
from full clinical assessment' rather than on speci®c pre-
determined criteria. One particular diculty encoun-
tered for our screening was the poor ability of parents of
non-English speaking backgrounds to convey symptoms
and history via a simple questionnaire. Hence, decisions
to refer have been made in the light of an unreliable
understanding of the child's background.
Hoppe (1997) notes that epidemiological considera-
tions drive optometric clinical decision-making di€er-
ently when a number of tests have been carried out in
parallel as occurs in a screening, rather than when
testing is based upon problems elicited during history
and symptom taking. The latter drives a case-appro-
priate test set with far less risk of false negative
outcomes, but which might include items that are never
considered for a screening on a cost±bene®t basis. As
children in the current study were lost to follow-up, it
has not been possible to review the results of further
assessment as regards the relative numbers of false
positives and negatives. However, the supervising
clinicians responsible for the referrals may have devi-
ated from the original set of tests and moved towards a
more case-appropriate eye examination using evidence-
based criteria derived from their own experience. This
could have resulted in the mismatch between the VEC
guidelines for referral and the actual pattern of
referrals.
Conclusion
The current project presents oculo-visual characteristics
(including refractive and functional vision measures) for
a large group of urban multicultural Australian children
aged 3±12 years.
By undertaking analysis of the large VEC data set as a
whole and referring to population norms from the
literature, it has been possible to establish problems in
delivery of vision care during the screening which would
not have been apparent by review of individual records.
One example is the di€erent accommodative facility
results, which led to an examination of the actual
protocol used. Another is the role of the attending
optometrist.
As the highly regarded MCT does not address non-
strabismic binocular disorders, the current study
attempts to establish the suitability of a set of referral
criteria regarding binocular vision. In their current
form, the set of VEC referral guidelines do not match
those used by the supervising clinicians when given
freedom to use their own judgment. The lack of a clear
pattern guiding decisions to refer in this study suggests
that, when working within a limited time frame,
experienced clinicians ultimately decide on the basis of
clinical intuition. This lack of a `referral formula' has
implications for the stang of screenings of school
children aged 5±12 years, namely that optometrists
conversant with the appraisal of the subtleties of near
point visual performance must be in attendance and
have sucient time themselves to be one-to-one with
each child. However, our data do not allow an evalu-
ation of the ecacy of this clinical intuition in detecting
visual anomalies.
Acknowledgements
The work of Year 4 students School of Optometry
UNSW Nuong Le and Cathy Seto on the comparison of
the three methods of accommodative facility testing is
ã 2002 The College of Optometrists

gratefully acknowledged. SGC's salary was funded by
National Health and Medical Research Council `A
temporal mechanism for the acuity de®cit in strabismic
amblyopia' 1994±96; PMK's salary was funded by
NHMRC `A role for the outer retina in the control of
axial refractive error' 1997±99 and by Australian
Research Council `Psychophysiology of Transient Pro-
cessing in Reading' 2000. We also thank Patricia
Munoz, Bernice Webster and Murray Lawson for their
assistance.
References
Alexander, J. A. (1990) Calbration of the Sheedy ®xation
disparometer. Optom Vis Sci. 67, 90.
Amigo, G., McCarthy, A. and Pye, D. (1976) Visual charac-
teristics of an underprivileged group of Australian children.
Aust. J. Optom. 59, 188±197.
Atkinson, J. (ed.) (2000) The Developing Visual Brain. Oxford
University Press, New York.
Bailey, N. R. (1998) Assessing the predicative ability of the
test-positive ®ndings of an elementary school vision screen-
ing. Optom. Vis. Sci. 75, 682±691.
Bear, J. C. (1991) Epidemiology and genetics of refractive
anomalies. In: Refractive Anomalies. Research and Clinical
Implications (eds T. Grosvenor and M. C. Flom) Butter-
worth-Heinemann, Stoneham, MA, pp. 57±80.
Bennet, G. R., Blondin, M. and Ruskiewicz, J. (1982)
Incidence and prevalence of selected visual conditions. J.
Am. Optom. Assoc. 53, 647±656.
Blum, H., Peters, H. B. and Bettman, J. W. (1968) Vision
Screening for Schools 1968 The Orinda Study. University of
California Press, Berkeley, CA.
Borish, I. M. (1975) Phorometry. In: Clinical Refraction.
Professional Press, Chicago, IL, USA, pp. 816±818.
Chua, W. H., Saw, H. M., Wu, C. Y. et al. (2001) Astigmatism
and myopia in Singapore schoolchildren. Invest Ophthalmol
Vis Sci. 42, S389.
Ciner, E. B., Schmidt, P. P., Orel-Bixler, D., Donson, V.,
Maguire, M., Cyert, L., Moore, B. and Schultz, J. (1998)
Vision screening of pre-school children: evaluating the past,
looking toward the future. Optom. Vis. Sci. 75, 571±584.
Coleman, H. M. (1970) An analysis of the visual status of an
entire school population. J. Am. Optom. Assoc. 41, 341±347.
Committee on Children with Disabilities, American Academy
of Pediatrics, American Academy of Ophthalmology,
American Association for Pediatric Ophthalmology and
Strabismus (1998) Learning disabilities, dyslexia, and vision:
a subject review. Pediatrics 102, 1217±1219.
Cregg, M., Woodhouse, J. M., Pakeman, V. H., Saunders, K.
J., Gunter, H. L., Parker, M., Fraser, W. I. and Sastry, P.
(2001) Accommodation and refractive error in children with
Down syndrom. cross-sectional and longitudinal studies.
Invest. Ophthal. Vis. Sci. 42, 55±63.
Crewther, S. G., Kiely, P. M., Kok, L.-L. and Crewther, D. P.
(1998) Anomalies of genetic development as predictors of
oculo-visual abnormalities in Velo-Cardio±Facial syndrome.
Optom. Vis. Sci. 75, 748±757.
ã 2002 The College of Optometrists
School screenings: B. Junghans et al. 23
Dwyer, P. S. (1983) The Portsea Lord Mayor's children's camp
vision screening: a rationale and protocol for optometric
screening. Aust. J. Optom. 66, 178±185.
Dwyer, P. S. (1992) The prevalence of vergence accommoda-
tion disorders in a school-age population. Clin. Exp. Optom.
75, 18±20.
Evans, B. J., Patel, R., Wilkins, A. J., Eperjesi, F., Speedwell,
L. and Du€y, J. A. (1999) Review of the management of
3323 consecutive patients seen in a speci®c learning dicul-
ties clinic. Ophthal Physiol. Opt. 19, 454±466.
Evans, B. J. (1998) The underachieving child. Ophthal Physiol.
Opt. 18, 153±159.
Fowler, M. S. (1991) Binocular control in dyslexics. In: Vision
and Vision Dyslexia ± Vision and Visual Dysfunction, Vol. 13
(ed. J. F. Stein) Macmillan Press, London, pp. 141±147.
Garzia, R. P. and Franzel, A. S. (1996) Refractive status,
binocular vision and reading achievement. In: Vision and
Reading, Ch 6 (ed. R. P. Garzia) Mosby, St. Louis, MO,
111±131.
Grisham, J. D. and Simons, H. D. (1986) Refractive error and
the reading process: a literature analysis. J. Am. Optom.
Assoc. 57, 44±55.
Hagerman, R. and Cronister, A. (1996) Fragile X Syndrome:
Diagnosis, Treatment and Research. John Hopkins Univer-
sity Press, Baltimore.
Hayes, G. J., Cohen, B. E., Rouse, M. W. and De Land, P. N.
(1998) Normative values for the near point of convergence
of elementary schoolchildren. Optom. Vis. Sci. 75, 506±512.
Hennessey, D., Iosue, R. A. and Rouse, M. W. (1984) Relation
of symptoms to accommodative infacility of school-aged
children. Am. J Optom Physiol Opt. 61, 177±183.
Hirsch, M. J., Weymouth, F. W. (1991) Prevalence of refractive
anomalies. In: Refractive Anomalies. Research and Clinical
Implications (eds T. Grosvenor and M. C. Flom) Butter-
worth-Heinemann, Stoneham, Boston, MA, pp. 21±25.
Hoppe, E. (1997) Use of epidemiology in clinical decision
making. In: Clinical Decision Making in Optometry (eds
E. R. Ettinger and M. W. Rouse) Butterworth-Heinemann,
Boston, MA, pp. 65±99.
Howell, E. R., Brown, B. and Bailey, I. L. (1980) The Stimulus
to Accommodation by the Human Eye. Proceedings of the
Aust. Physiol. Pharmacol. Soc. May 1980. (by Personal
Communication).
Jenkins, T. C. A., Pickwell, L. D. and Yekta, A. A. (1989)
Criteria for decompensation in binocular vision. Ophthal
Physiol. Opt 9, 121±125.
Joint Organizational Policy Statement of the American Acad-
emy of Optometry and the American Optometric Associ-
ation. (1997) J. Am. Opt Assoc 68, 284±286.
Junghans, B. M. and Crewther, S. G. (1992) The vision
education centre: a multi-level educational tool. J. Optom.
Education 17, 83±86.
Kulp, M. T. and Schmidt, P. P. (1996) E€ect of oculmotor and
other visual skills on reading performance: a literature
review. Optom. Vis. Sci. 75, 682±691.
Lam, C. S. Y. and Goh, W. S. H. (1991) The incidence of
refractive errors among school children in Hong Kong and
its relationship with the optical components. Clin. Exp
Optom. 74, 97±103.
24 Ophthal. Physiol. Opt. 2002 22: No 1
Lam, S. R., LaRoche, G. R., De Becker, I. and Macpherson,
H. (1996) The range and variability of ophthalmological
parameters in normal children aged 4, 1/2±5, 1/2 years.
J. Pediatr. Ophthalmol. Strabismus 33, 251±256.
Lovegrove, B. (1996) Dyslexia and a transient/magnocellular
pathway de®cit. the current situation and future directions.
Aust. J. Psychol 48, 167±171.
Lyle, W. M. (1991) Astigmatism. In: Refractive Anomalies.
Research and Clinical Implications (eds T. Grosvenor and
M. C. Flom) Butterworth-Heinemann, Stoneham, Boston,
MA, pp. 146±155.
McKenzie, K. M., Kerr, S. R., Rouse, M. W. and DeLand, P.
N. (1987) Study of accommodative facility testing reliability.
Am. J. Optom. Physiol. Opt. 64, 186±194.
Morgan, K. S. and Kennemer, J. C. (1997) O€-axis photore-
fractive eye screening in children. J. Cataract Refract. Surg.
23, 423±428.
Nathan, J., Kiely, P. M., Crewther, S. G. and Crewther, D. P.
(1985) Disease-associated visual image degradation and
spherical refractive errors in children. Am. J. Optom.
Physiol. Opt. 62, 680±688.
Nathan, J., Kiely, P. M., Crewther, S. G. and Crewther, D. P.
(1986) Astigmatism occurring in association with pediatric
eye disease. Am. J. Optom. Physiol. Opt. 63, 497±504.
O'Grady, J. (1984) The relationship between vision and
educational performance. Aust. J. Optom. 67, 126±140.
Osipov, G. I. (1996) Dynamic stereovisometry. Vestn. Oftal-
mol. 112, 35±37. (abstract only, article in Russian).
Robbins, H. and Bailey, I. L. (1975) The organisation and
®ndings of a school vision screening service. Aust. J. Optom.
58, 392±401.
Rosner, J. and Rosner, J. (1994) The relationship between
ametropia and elementary school achievement. Br. J.
Optom. Disp. 2, 8.
Rouse, M. W., Borsting, E., Hyman, L. et al. (1999) Fre-
quency of convergence insuciency among ®fth and sixth
graders. Optom Vis Sci. 76, 643±649.
Rouse, M. W., Hyman, L., Hussein, M. and Solan, H. (1998)
Frequency of convergence insuciency in optometry clinic
settings. Optom. Vis. Sci. 75, 88±96.
Scheiman, M., Herzberg, H. and Frantz, K. (1988) Margolies.
Normative study of accommodative facility in elementary
school children. Am. J. Optom Physiol. Opt. 65, 127±124.
Sheedy, J. E. and Saladin, J. J. (1977) Phoria, vergence, and
®xation disparity in oculomotor problems. Am. J. Optom
Physiol. Opt. 54, 474±478.
Sheedy, J. E. and Saladin, J. J. (1978) Association of
symptoms with measures of oculomotor de®ciencies.
Am. J. Optom Physiol. Opt 55, 670±676.
Siderov, J. (1990) Improving interactive facility with vision
training. Clin. Exp. Optom. 73, 128±131.
Simons, K. (1986) Stereoacuity norms in young children. Arch.
Ophthalmol. 28, 84±92.
Simons, H. D. and Gassler, P. A. (1988) Vision anomalies and
reading skill: a meta-analysis of the literature. Am. J. Optom
Physiol. Opt. 11, 893±904.
Simons, H. D. and Grisham, J. D. (1986) Vision and reading
disability: research problems. J. Am. Optom Assoc. 57, 36±41.
Simons, H. D. and Grisham, J. D. (1987) Binocular anomalies
and reading problems. J. Am. Optom Assoc 58, 578±587.
The People of New South Wales: Statistics from the, 1996
Census. (1998) Ethnic A€airs Commission of NSW, NSW
Government, Sydney, 149±151.
The 1986/87 Future of Visual Development/Performance
Task Force. (1988) The ecacy of optometric vision
therapy. J. Am. Opt Assoc 59, 95±105.
Thomson, W. D., Evans, B. (1999) A new approach to vision
screening in schools. Ophthal Physiol. Opt. 19I, 196±209.
Tsiaras, W. G., Pueschel, S., Keller, C., Curran, R. and
Giesswein, S. (1999) Amblyopia and visual acuity in children
with Down's syndrome. Br. J. Ophthal. 83, 1112±1114.
Walline, J. J., Mutti, D. O., Zadnik, K. and Jones, L. A. (1998)
Development of phoria in children. Optom. Vis. Sci. 75,
605±610.
Walters, J. (1984a) Portsea modi®ed clinical technique: results
from an expanded optometric screening protocol for
children. Aust. J. Optom. 67, 176±186.
Walters, J. (1984b) Portsea modi®ed clinical technique: eval-
uation of an expanded optometric vision screening protocol.
Aust. J. Optom. 67, 212±220.
Wick, B. and London, R. (1987) Analysis of binocular visual
function using tests made under binocular conditions.
Am .J. Optom Physiol. Optics. 64, 227±240.
Yekta, A. A., Pickwell, L. D. and Jenkins, T. C. A. (1989)
Binocular vision, age and symptoms. Ophthal Physiol. Opt 9,
11±20.
Zadnik, K. (1997) The Glenn A. Fry award lecture (1995).
Myopia development in childhood. Optom. Vis. Sci. 74,
603±608.
Zadnik, K., Mutti, D. O., Friedman, N. E. and Adams, A. J.
(1993) Initial cross-sectional results from the Orinda
Longitudinal Study of Myopia. Optom. Vis. Sci. 70,
750±758.
Zellers, J. A., Alpert, T. L. and Rouse, M. W. (1984) A review
of the literature and a normative study of accommodative
facility. J. Am. Optom. Assoc. 55, 31±37.
ã 2002 The College of Optometrists
 School screenings: B. Junghans et al. 25

Appendix.
Teacher's checklist
Children with vision problems frequently show recognisable signs. Could you please tick against any of the problems below which apply to this
child.
Pupil's name: ________________
1. Eyes look red or sore _____
2. Screws up eyes when looking in the distance _____
3. Book held close when reading or sits awkwardly _____
4. Complains of: Headaches _____
Dif®culty seeing clearly in the distance _____
Blurring of vision while reading or writing _____
Seeing double _____
Eye burning or itching during or after close work _____
Parent's checklist
Children with vision problems frequently show recognisable signs. Could you please tick against any of the problems below which apply to this
child.
Pupil's name: ________________
1. Eyes look red or sore _____
2. Screws up eyes when looking in the distance? _____
3. Book held close when reading or sits awkwardly? _____
4. Has to sit very close to the television? _____
5. Complains of: Headaches _____
Dif®culty seeing clearly in the distance _____
Blurring of vision while reading or writing _____
Seeing double _____
Eye burning or itching during or after close work _____
6. Is the child taking regular medication? If so, what?
______________________________________________
7. Is your child already receiving vision care?
If so, from whom?____________________________________________

ã 2002 The College of Optometrists