Guidelines For Automated Preschool Vision Screening: A 10-Year, Evidence-Based Update

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

State of the Art

Guidelines for automated preschool vision screening:


A 10-year, evidence-based update
Sean P. Donahue, MD, PhD,a Brian Arthur, MD,b Daniel E. Neely, MD,c Robert W. Arnold, MD,d
David Silbert, MD, FAAP,e and James B. Ruben, MD,f on behalf of the AAPOS Vision Screening
Committee*

SUMMARY In 2003 the American Association for Pediatric Ophthalmology and Strabismus Vision
Screening Committee proposed criteria for automated preschool vision screening. Recent
literature from epidemiologic and natural history studies, randomized controlled trials of
amblyopia treatment, and field studies of screening technologies have been reviewed for
the purpose of updating these criteria. The prevalence of amblyopia risk factors (ARF)
is greater than previously suspected; many young children with low-magnitude ARFs do
not develop amblyopia, and those who do often respond to spectacles alone. High-
magnitude ARFs increase the likelihood of amblyopia. Although depth increases with
age, amblyopia remains treatable until 60 months, with decline in treatment effectiveness
after age 5. US Preventive Services Task Force Preventative Services Task Force guide-
lines allow photoscreening for children older than 36 months of age. Some technologies
directly detect amblyopia rather than ARFs. Age-based criteria for ARF detection using
photoscreening is prudent: referral criteria for such instruments should produce high
specificity for ARF detection in young children and high sensitivity to detect amblyopia
in older children. Refractive screening for ARFs for children aged 12-30 months should
detect astigmatism .2.0 D, hyperopia .4.5 D, and anisometropia .2.5 D; for children
aged 31-48 months, astigmatism .2.0 D, hyperopia . 4.0 D, and anisometropia .2.0 D.
For children .49 months of age original criteria should be used: astigmatism .1.5 D,
anisometropia.1.5 D, and hyperopia .3.5 D. Visually significant media opacities and
manifest (not intermittent) strabismus should be detected at all ages. Instruments that de-
tect amblyopia should report results using amblyopia presence as the gold standard. These
new American Association for Pediatric Ophthalmology and Strabismus Vision Screening
Committee guidelines will improve reporting of results and comparison of technolo-
gies. ( J AAPOS 2013;17:4-8)

A
See accompanying article on page 2 pproximately two decades ago, the first automated
a
Author affiliations: Department of Ophthalmology and Visual Sciences, Vanderbilt device for preschool vision screening became
University Medical Center, Nashville, Tennessee; bDepartment of Ophthalmology, commercially available. Although this was clearly
Queen’s University, Kingston, Ontario, Canada; cIndiana University, Department of a major breakthrough, attempts to compare data from var-
Ophthalmology, Indianapolis, Indiana; dOphthalmic Associates, Anchorage, Alaska;
e
Family Eye Group, Ephrata, Pennsylvania; fKaiser Permanente, Sacramento, ious validation studies proved difficult because there was
California no consensus for delineating which pathology this technol-
Financial conflict of interest: Dr. Donahue has been a consultant for several preschool ogy should detect or how the results of validation studies
vision screening companies, including Welch Allyn, MTI, Diopsys, iScreen, PediaVision,
and Plusoptix. Dr. Arnold has had instruments supplied by manufactures for the ABCD should be reported. In 2003 the Vision Screening Commit-
program. tee of the American Association for Pediatric Ophthalmo-
Supported by a grant for Research to Prevent Blindness from Vanderbilt University. Dr. logy and Strabismus (AAPOS) established guidelines for
Donahue receives support from the Coleman Endowed Chair in Ophthalmology and Visual
Sciences. the reporting of results from studies of automated vision
Presented at the 38th Annual Meeting of the American Association for Pediatric screening instruments.1 Because these technologies were
Ophthalmology and Strabismus, San Antonio, Texas, March 24-28, 2012. designed to identify children with strabismus, anisometro-
* A complete listing of the AAPOS Vision Screening Committee is provided in
e-Supplement 1. pia, and/or bilateral high magnitude refractive error, the
Submitted May 8, 2012. published guidelines primarily addressed the magnitude
Revision accepted September 25, 2012. of refractive error that was (by consensus) thought to put
Published online January 28, 2013.
Correspondence: Sean P. Donahue, MD, PhD, 2311 Pierce Ave, Nashville, TN 37232- a child at risk for the development of amblyopia—the
8808 (email: sean.donahue@vanderbilt.edu). “amblyogenic factors”, now called “amblyopia risk factors”
Copyright Ó 2013 by the American Association for Pediatric Ophthalmology and (ARFs).2
Strabismus.
1091-8531/$36.00 Since the publication of these guidelines, more data have
http://dx.doi.org/10.1016/j.jaapos.2012.09.012 become available about the prevalence of amblyopia risk

4 Journal of AAPOS
Volume 17 Number 1 / February 2013 Donahue et al 5

factors in young children, the age-dependent development amblyopia with spectacle treatment alone; this phenome-
of amblyopia in patients with amblyopia risk factors, and non is seen in children with anisometropic amblyopia13
the treatment of amblyopia with spectacles prior to the as well as in those with strabismic amblyopia.14 However,
initiation of active occlusion or penalization treatment. refractive adaptation is less likely to occur (or be complete)
The results of these studies have necessitated reassessment in children with deeper amblyopia; thus these children
of these guidelines. Likewise, technology has advanced, must be identified at a younger age.13 This and the afore-
and screening instruments are now available that detect ab- mentioned data with respect to the nature of amblyopia de-
normalities other than amblyopia risk factors. The purpose velopment in at-risk children suggest that preschool vision
of this article is to review the new evidence and adjust the screening devices should aim to detect only the greatest-
reporting guidelines appropriately. Although the new magnitude anisometropia at young ages, prior to when
guidelines remain consensus-based, the evidence support- amblyopia develops and becomes entrenched. A corollary
ing the new guidelines is significantly stronger than the is that refractive screening technologies should have high
evidence that supported the 2003 guidelines. specificity but low sensitivity to detect low-magnitude
Several prospective population-based studies have symmetric refractive errors in the youngest children.
confirmed that the prevalence of amblyopia in childhood Although detection of amblyopia at a younger age
is approximately 2%,3-5 a finding that is consistent with generally produces better treatment outcomes, new meta-
previous reports. However, the prevalence of ARFs is analyses have demonstrated that amblyopia treatment does
much greater than previously thought, probably in the not begin to decrease in effectiveness until approximately
neighborhood of 15% to 20%.6-8 Hence, it is clear that age 5 years.15 However, early detection of high-magnitude
the majority of children with amblyopia risk factors do refractive error may allow prevention of amblyopia in
not develop amblyopia, and this has been confirmed in some at-risk children and allow for treatment using refractive
a longitudinal follow-up study of children identified adaptation rather than active therapies at an age when ambly-
through vision screening.9 If the detection of decreased opia has not yet become entrenched. Also, earlier detection
vision and amblyopia are the goals of screening, then refer- may allow for treatment to be more cost effective by reducing
rals based on technology that detects risk factors will result the number of medical visits required for resolution.
in overreferrals. It is therefore imperative that updated Amblyopia screening should be viewed as a continuous
guidelines for detecting amblyopia risk factors propose process that occurs throughout visual development, begin-
levels that best separate those children who are most at ning in infancy. We anticipate that vision screening of
risk for developing amblyopia from those who are not. children will take place at several times during the forma-
The relationship between refractive error and the likeli- tive years rather than at one particular age; thus a high
hood of development of amblyopia is complex and sensitivity to detect mild amblyopia during a single
depends on the age of the child, the magnitude of refrac- screening is an unnecessarily expensive strategy if it is
tive blur, and other factors. For children up to 3 years of associated with a low positive predictive value. It is note-
age with anisometropia, the prevalence of amblyopia worthy that the US Preventative Services Task Force
appears to correlate with the magnitude of the anisometro- (USPSTF) now actively recommends vision screening at
pia.10 For those more than 3 years of age, however, the least once for children between 36 months and 5 years
prevalence of amblyopia remains relatively constant, but and specifically mentions photoscreening as an appropri-
the depth of amblyopia increases with age, and greater- ate screening technology.16 Although USPSTF guidelines
magnitude refractive errors seem solely to increase the consider the evidence in favor of screening children aged
depth but not the prevalence of amblyopia.11 Our updated 12-35 months to be “insufficient,” the invited commen-
guidelines lower the referral rate for young children by tary17 addresses this controversy and provides evidence
raising the threshold referral values. We recognize that to the contrary.
this will produce a corresponding decrease in sensitivity Older children ($5 years of age) have less time available
to detect low-magnitude refractive pathology (and proba- for treatment and may already have entrenched amblyopia.
bly mild amblyopia) but anticipate that it will minimally af- Thus, screening should be more sensitive in this age group.
fect the sensitivity to detect those high-magnitude Preference should be given to visual acuity measurement
refractive errors that are potentially most likely to lead to that uses crowded or surrounded optotypes (LEA symbols,
amblyopia. Because the prevalence of amblyopia increases HOTV chart, or Sloan), with monocular testing assured by
with the magnitude of anisometropic refractive error,11 au- patching, which allows the direct detection of impaired
thors are urged to report multiple levels of sensitivity for visual function. However, refractive error screening is
several magnitudes of refractive error above the thresholds also appropriate for those children who cannot cooperate
proposed here.12 with traditional screening, for high-volume field-based
Recommendations with respect to screening of screening, and for primary care settings in which tradi-
preschool children must also occur within the context of tional screening is either more challenging or less efficient
treatment using “refractive adaptation.” Many children, than automated screening (Sloan letters are preferred both
especially those having mild amblyopia, often have marked for screening and validation, but we recognize that most
improvement (and sometimes even resolution) of their providers use Snellen letters).

Journal of AAPOS
6 Donahue et al Volume 17 Number 1 / February 2013

The original AAPOS guidelines were vague with respect Table 1. Amblyopia risk factors targeted with automated preschool
to the detection of strabismus, especially with respect to vision screening
incomitant (paretic, restrictive, and pattern) syndromes. Refractive risk factor targetsa
Intermittent exotropia and well-controlled deviations (eg,
Age, months Astigmatism Hyperopia Anisometropia Myopia
Superior Oblique Palsy, Monocular Elevation Deficiency,
Duane syndrome, and Brown syndrome) are neither typi- 12-30 .2.0 D .4.5 D .2.5 D . 3.5 D
31-48 .2.0 D .4.0 D .2.0 D . 3.0 D
cally associated with amblyopia development nor with .48 .1.5 D .3.5 D .1.5 D . 1.5 D
rapid loss of stereopsis; thus, they need not be detected
by modalities that seek to detect decreased binocular (re- Nonrefractive amblyopia risk factor targetsb
fractive) or monocular (amblyopia) visual acuity (such as All ages Manifest strabismus .8 PD in primary position
photoscreening or direct acuity testing). However, accom- Media opacity .1 mm
modative esotropia (a manifest strabismus on an accom- D, diopters; PD, prism diopters.
a
modative target at distance or near at any time during Additional reporting of sensitivity to detect greater-magnitude refrac-
a formal eye examination) is associated with amblyopia tive errors is encouraged.
b
For all ages.
development and degradation of stereopsis and should
be detected even though in its early stages it may be
intermittent. 2. Detection of Amblyopia Risk Factors Early in
The guidelines also are updated with respect to media Preschool Children (Age Group: 31-48 Months)
opacities, pupillary abnormalities, and eyelid abnormali- For older children who remain unable to have visual acuity
ties. Any media opacity greater than 1 mm in size is poten- assessed directly, the detection of lower magnitude ambly-
tially amblyopiogenic and should be detected with opia risk factors becomes more important, although
photorefractive screening. Isolated anisocoria does not symmetric bilateral moderate-magnitude astigmatic and
produce amblyopia and its association with ocular or hypermetropic refractive error probably remains unneces-
systemic pathology is exceedingly rare; hence it has been sary to detect or treat. Recommended targets are as follows:
removed from the list of amblyopia risk factors. Finally, astigmatism .2.0 D, hyperopia .4.0 D, and anisometro-
because nearly all amblyopia-related ptosis occurs in the pia .2.0 D. Refractive amblyopia risk factors that persist
setting of superimposed anisometropia,18,19 ptosis has toward the end of this age range are less likely to spontane-
been removed as an ARF. ously resolve3 and are more likely to be associated with am-
blyopia. The detection of higher magnitude anisometropia
(.3.0 D) should be highly sensitive for this age group
because it is nearly always associated with amblyopia that
Recommendations
continues to deepen over time.10
The following recommendations are summarized in
Table 1. 3. Detection of Amblyopia Risk Factors in Late
Preschool and Kindergarten Children (Age Group:
49-72 Months)
1. Detection of Amblyopia Risk Factors in Toddlers For children aged 49-72 months, amblyopia risk factors for
(Age Group: 12-30 Months) astigmatism, hypermetropia, and anisometropia are
For very young preverbal children, the detection of unchanged from the original guidelines. In this age range,
low-level refractive amblyopia risk factors should be highly moderate-magnitude astigmatism begins to produce
specific (ie, there should be very few false-positive refer- decreased visual function, and detection should probably
rals). The recommended target refractive magnitudes for occur during this time period. Detection of myopia begins
detection are as follows: astigmatism .2.0 D, hyperopia to become important in this age range because children
.4.5 D, and anisometropia .2.5 D. These targets are set begin to pay more attention to distance targets, and thus
at a higher level than for older age groups because children myopia of $ 1.5 D should be detectable.
with bilateral and symmetric refractive errors of this and
lesser magnitudes typically do not have functional im- 4. Detection of Amblyopia Risk Factors in
provement in visual behavior as a result of correction; as School-Aged Children (Age Group: >72 Months)
a result they are unlikely to wear their glasses. In addition, Most children older than 72 months of age are able to read
such refractive errors, when bilateral and symmetric, rarely a standard linear optotype eye chart and can be screened
cause significant bilateral ametropic amblyopia. The false- using this modality (see below).5 Exceptions are appropri-
negative cases that do occur can be captured at later ages, as ate for delayed children, those unable to read letters, chil-
recommended by guidelines of the American Academy of dren who are uncooperative with optotype-based visual
Pediatrics, through either repeated objective (ie, instru- acuity, and high-volume field screening. Further research
ment based) screening methods or during subjective acuity will determine whether objective technologies have greater
testing with minimal loss in function. utility compared with optotype-based screening for

Journal of AAPOS
Volume 17 Number 1 / February 2013 Donahue et al 7

Table 2. Reporting guidelines for nonrefractive vision screening References


instruments
1. Donahue SP, Arnold RW, Ruben JB, AAPOS Vision Screening
Report sensitivity and specificity to detect Committee. Preschool vision screening: What should we be detect-
ing and how should we report it? Uniform guidelines for reporting
1. Visual acuity \20/30 results of preschool vision screening studies. J AAPOS 2003;7:
2. $3 lines intraocular visual acuity difference 314-16.
3. Manifest strabismus 2. Matta NS, Singman EL, Silbert DI. Performance of the Plusoptix vi-
sion screener for the detection of amblyopia risk factors in children.
J AAPOS 2008;12:490-92.
3. Multi-ethnic Pediatric Eye Disease Study Group. Prevalence of
amblyopia and strabismus in African American and Hispanic children
school-aged children.20 The USPSTF has specifically en- ages 6 to 72 months the Multi-ethnic Pediatric Eye Disease Study.
dorsed the use of photoscreening modalities for the detec- Ophthalmology 2008;115:1229-36.
tion of amblyopia risk factors in the 3- to 5-year-old age 4. Friedman DS, Repka MX, Katz J, et al. Prevalence of amblyopia and
group.16 strabismus in white and African American children aged 6 through
71 months the Baltimore Pediatric Eye Disease Study. Ophthalmo-
logy 2009;116:2128-34.
5. Detection of Amblyopia and Decreased Visual 5. Pai AS, Rose KA, Leone JF, et al. Amblyopia prevalence and risk
Acuity Using Traditional (Optotype-Based) factors in Australian preschool children. Ophthalmology 2012;119:
Screening 138-44.
6. Borchert M, Tarczy-Hornoch K, Cotter SA, Liu N, Azen SP,
Traditional optotype recognition screening remains a via- Varma R, MEPEDS Group. Anisometropia in Hispanic and African
ble option for cooperative patients who can read linear let- American infants and young children the Multi-ethnic Pediatric Eye
ters because it allows the direct detection of decreased Disease Study. Ophthalmology 2010;117:148-53.
visual acuity. Nonetheless, it is time-consuming and often 7. Multi-Ethnic Pediatric Eye Disease Study Group. Prevalence of
difficult until children are well into their elementary school myopia and hyperopia in 6- to 72-month-old African American and
Hispanic children: The Multi-ethnic Pediatric Eye Disease Study.
years. When such children are screened, detection of mon-
Ophthalmology 2010;117:140-47.
ocular visual acuity \20/30, as specified by joint American 8. Fozailoff A, Tarczy-Hornoch K, Cotter S, et al., Writing Committee
Academy of Pediatrics/AAPOS guidelines, should be the for the MEPEDS Study Group. Prevalence of astigmatism in 6- to
standard. The use of stereopsis testing in isolation remains 72-month-old African American and Hispanic children: The
poorly validated and is not considered here. Multi-ethnic Pediatric Eye Disease Study. Ophthalmology 2011;
118:284-93.
9. Colburn JD, Morrison DG, Estes RL, Li C, Lu P, Donahue SP.
6. Detection of Amblyopia and Decreased Visual Longitudinal follow-up of hypermetropic children identified during
Acuity Using Instruments Other than preschool vision screening. J AAPOS 2010;14:211-15.
Photoscreeners and Autorefractors 10. Donahue SP. Relationship between anisometropia, patient age,
and the development of amblyopia. Am J Ophthalmol 2006;142:
Because many children with ARFs never develop ambly- 132-40.
opia, the development and validation of screening tech- 11. Leon A, Donahue SP, Morrison DG, Estes RL, Li C. The
niques that can detect children with amblyopia or age-dependent effect of anisometropia magnitude on anisometropic
strabismus directly would be a major advance (Table 2). amblyopia severity. J AAPOS 2008;12:150-56.
The Pediatric Vision Scanner detects the absence of foveal 12. Donahue SP, Johnson TM, Ottar W, Scott WE. Sensitivity of
photoscreening to detect high-magnitude amblyogenic factors.
fixation as a harbinger of strabismus and amblyopia and is
J AAPOS 2002;6:86-91.
potentially an example of one such instrument.21,22 13. Cotter SA. Pediatric Eye Disease Investigator Group, Edwards AR,
Electrophysiologic testing of acuity or foveation might Wallace DK, Beck RW, et al. Treatment of anisometropic amblyopia
have similar advantages.23 Reporting results from these in children with refractive correction. Ophthalmology 2006;113:
instruments using the 2003 reporting guidelines would 895-903.
14. Cotter SA, Edwards AR, Arnold RW, et al., Pediatric Eye Disease
be futile because these instruments are not designed to de-
Investigator Group. Treatment of strabismic amblyopia with refrac-
tect refractive error in the absence of a resultant decrease in tive correction. Am J Ophthalmol 2007;143:1060-63.
visual acuity. Instead, results from testing of these and sim- 15. Holmes JM, Lazar EL, Melia BM, et al., Pediatric Eye Disease Inves-
ilar instruments should report sensitivity and specificity to tigator. Effect of age on response to amblyopia treatment in children.
detect manifest strabismus, and $3 lines of interocular Arch Ophthalmol 2011;129:1451-7.
acuity difference. 16. US Preventive Services Task Force. Vision screening for children 1 to
5 years of age: US Preventive Services Task Force Recommendation
In conclusion, further advances in technology will statement. Pediatrics 2011;127:340-46.
invariably force a reassessment of the preferred means of 17. Donahue SP, Ruben JB, American Academy of Ophthalmology;
detecting children who have amblyopia or other causes of American Academy of Pediatrics, Ophthalmology Section; American
decreased visual acuity. Similarly, advances in our knowl- Association for Pediatric Ophthalmology and Strabismus; Children’s
edge regarding the natural history of refractive error in Eye Foundation; American Association of Certified Orthoptists. US
Preventive Services Task Force vision screening recommendations.
children, and risk factors for amblyopia development will
Pediatrics 2011;127:569-70.
also force a reassessment of these guidelines. We actively 18. Oral Y, Ozgur OR, Akcay L, Ozbas M, Dogan OK. Congenital
encourage continued research in these areas and look for- ptosis and amblyopia. J Pediatr Ophthalmol Strabimus 2010;47:
ward to further revision of these guidelines. 101-4.

Journal of AAPOS
8 Donahue et al Volume 17 Number 1 / February 2013

19. Srinagesh V, Simon JW, Meyer DR, Zobal-Ratner J. The association 22. Louden SE, Rook CA, Nassif DS, Piskun NV, Hunter DG. Rapid,
of refractive error, strabismus, and amblyopia with congenital ptosis. high-accuracy detection of strabismus and amblyopia using the
J AAPOS 2011;15:541-4. pediatric vision scanner. Invst Ophthalmol Vision Sci 2011;52:
20. Salcido AA, Bradley J, Donahue SP. Predictive value of photoscreening 5043-8.
and traditional screening of preschool children. J AAPOS 2005;9:114-20. 23. Simon JW, Siegfried JB, Mills MD, Calhoun JH, Gurland JE. A new
21. Nassuf DS, Piskun NV, Hunter DG. The Pediatric Vision Screener III: visual evoked potential system for vision screening in infants and
Detection of strabismus in children. Arch Ophthal 2006;124:509-13. young children. J AAPOS 2004;8:549-54.

Journal of AAPOS

You might also like