Myopia Profile

Myopia Action Month now on offering FREE seminars, courses, resources and more: Get involved here.

Myopia Action Month now on: Get involved here.


Refraction challenges in children - what to prescribe?

Posted on December 1st 2020 by Connie Gan

Sponsored by


In this article:

How would you prescribe when you have an uncooperative child? Refraction challenges in children are addressed in this case.

Children can be tricky to refract. Depending on how cooperative the child is willing or able to be, practitioners may find themselves struggling to gain any meaningful subjective or objective refraction results. Refraction challenges in children are explored in this case that SK shared with the Myopia Profile community. This is not a case of a myope, but with more myopia management comes general paediatric management and hence cases such as these. What would you prescribe for a child with such findings?

SK Shared from another group I would love to read your input 5y old orthophoria, unresponsive to subjective refraction ( guess he means uncooperative patient ) and reported BCVA 0.6 OU ( colleague was questioning the child s capacity to give proper responses ) What glasses would you suggest in this pt . Cyclo wasn’t by atropine , I guess cyclopentolate 1% . It’s the child s first check Thank you very much in advance for any suggestions , I know there is too much info missing .. SK.jpg

It is useful to have a mental checklist of clinical considerations in prescribing for children.

1. Is there a risk of amblyopia?

CK With this prescription there is the possibility of amblyopia OU but more on the OD due to the higher cylinder…UTJ … I agree he/she likely have refractive ambylopia…

Based on the autorefractor results, the child has significant astigmatism in both eyes. According to Donahue et al, the refractive errors that place for children aged > 48 months at risk of amblyopia are >1.5 D astigmatism, >1.5 D anisometropia, and >3.5 D hyperopia.1 In this particular instance, there is a chance that this child may become amblyopic in both eyes, especially in the right, if this refractive error is not corrected.

2. Which refraction technique is best?

The first important factor for discussion is the technique of refraction. The child in this case is aged 5, and described as being 'unresponsive to subjective refraction'. The recommended refraction technique for a child of this age is retinoscopy, as described by both Optometry Australia and the American Optometric Association in their respective paediatric eye care guidelines. Optometry Australia recommend retinoscopy as the primary refraction technique for children aged from birth to 7 years. The American Optometric Association recommends retinoscopy is the preferable technique, with autorefraction best used as a starting point for other refraction techniques. If autorefraction is the only technique available, research indicates that its accuracy improves with cycloplegia,2 as was undertaken in this case. To read more about when to use cycloplegia in paediatric refractions, read our blog How to achieve accurate refractions for children. Let's explore the discussion that occurred around refraction techniques.

Autorefractor Vs. Retinoscopy

LM What are the ret findings for distance and near? This will give more a better idea of what the child can deal with in the specs rather than the auto findings.BB For me it depends on dry ret. My next step would be a return dry examination. I'm always happy with autorefractor for cyl axes direction not so for cyl amount. I'd expect to dry ret about +2.00/-3.00 and 2.00/-2.00 and would go roughly1.50/-2.50 and 1.5/-1.5 as a start point, look at acuity and build up cyl if it shows in over ret in about 3 months. Always worried about the perceptual and anisekonic comfort effect in going from no correction to full cyls first up myself.CH Coz you said the kid is ortho at dist. I would put in trial frame: +1.50/-3.50x15 +1.25/-2.25x168 Refine the sph and cyl while the kid look straight, especially the axis (high cyl 5* matter), make sure both vertical and horizontal ret is 90* each order and with same size. I trust my ret EO I would have done a dry retinoscopy prior for the sphere and wet retinoscopy after cyclo (especially if the child is non cooperative). Autorefractor gives you a good cylinder axis, but I often see more Hyperopia that’s elicited from wet retinoscopy. You can always do a 1.00 to 1.50 backoff of cyclo for sphere. I would probably correct fully for cyl but I wouldn’t prescribe full plus unless they are eso.UTJ Check the dry ret/book retinoscopy (one of oep class "just look retinoscopy" will give good tech. how to observe peds ret and reflex)… Trial frame what you may be prescribing - do quick check how they play around or work or even doing puppet touch game at different gaze… Check the reflex (book retinoscopy) I can always increase rx when needed however once patient adapt to the larger rx it will be difficult to let go. I never prescribe cyclo result and usually I consider final rx after 2nd visit to reconfirm for any fructuation in their accommodative system.

Many commenters concurred with guidelines from professional organizations, as indicated above, that dry or cycloplegic retinoscopy results would influence the decision on prescription. The comments suggest that autorefractors are handy for confirming the cylinder axis, but results for the degree of sphere or cylinder need to be refined by retinoscopy as the autorefractor is less accurate.

How accurate are autorefractors?

  • Choong and Wesemann showed that the results of autorefractors are accurate under cycloplegic condition whilst over-minussing under non-cycloplegic conditions.3,4 (autorefractors used: RETINOMAX K PLUS, CANON RF10, GRAND SEIKO WR5100K, RETINOMAX)
  • Hashemi showed that autorefractors tended to over-plus hyperopes and over-minus myopes when compared with cyloplegic retinoscopy. However, both methods corroborated and the difference was declared as clinically insignificant.5
  • Under cycloplegic conditions, the result of autorefraction and retinoscopy were similar for Indian children over 6 years of age in a study done by Guha et al. But, they included a caveat that the autorefraction result of children with mixed astigmatism, or under 6 years old, should be confirmed with retinoscopy.6 (autorefractor used: TOPCON KR-8900)
  • Prakabaran et al compared the results between a hand-held autorefractor, table-mounted autorefractor and retinoscopy. The table-mounted autorefractor gave similar results to retinoscopy whereas hand-held auto refractors gave slightly different results to table-mounted auto refractor and retinoscopy.7 (table-mounted autorefractor: CANON FK-1, hand-held autorefractor: RETINOMAX)

These studies indicate that autorefraction is almost accurate as retinoscopy when cycloplegia is used, but is unlikely to substitute for retinoscopy in children under 6 years of age, or those with high refractive error.

The American Optometric Association's Clinical Practice Guideline on Pediatric Eye and Vision Examination (pg 23) states that "autorefraction may be used as a starting point for subjective refraction, but not as a substitute for it; however, retinoscopy, when performed by an experienced clinician, is more accurate than automated refraction for determining a starting point for non-cycloplegic refraction." This was advice for school aged children, from age 6 to 18 inclusive. In clinical practice, it is ideal to confirm the refractive error with retinoscopy. Non-cycloplegic retinoscopy can also give you binocular vision insights as well - such as near lag of accommodation with near retinoscopy.

3. Full versus partial correction - which to prescribe?

Team Full Correction

NP I would prescribe full cyclo at this stage…JD …It is highly likely that the child will do better with full Rx (as long as he is not rendered exo) as the stimulus to accommodate accurately is missing with poor VADB Give the lot. No question.BY Prescribe in full…AY Give the full cyclo RxLA If you trust the autorefractor, give the full amount. Cyclo Ret obv much better. Why would you reduce cyl for adaptation? 5 yr olds adapt just fine. And an over plussed person can see clearly at close distances = no amblyopia An underplussed person may not accommodate accurately, so may not see clearly at any distance = amblyopia What's more important, clear DV or fixing amblyopia?

There was a group of practitioners supporting full correction for a few reasons:

  • That it would be better for the child's accommodation and yield better acuity to avoid amblyopia
  • That children can adapt to high cylinder corrections quite easily so there is no need to start with a partial script.

Team Partial Correction

CK With this prescription there is the possibility of amblyopia OU but more on the OD due to the higher cylinder. The cylinder thus is the only thing that really needs correction. I would give the full cylinder and a +1.00 of hyperopia. There is enough accommodation for the rest.CK generally yes, but the full cyclo might cause severe blur at distanceAA If we absolutely must go off autorefractor, I don't see any issue giving +2.00/-4.00x15 (Ks would be nice though) in regards to hyperopia. The best vision sphere on that is plano so there's no forced relaxation of accommodation. Maybe go +1.75/-4.00...again I'd check out Ks, as well as autos pre cyclo.SA I would prescribe half the hyperopia and give the full cyl and monitor closelyWYQ I would try full cyclo and give half of the cyl at this stage.BY …If you're worried about dist blur then cut back 0.25. If that's a nidek printout then they're quite accurate with cyclo 45mins.PC I would probably take off 0.50 from both the sphere and the cyl for adaptationJD …If you modify the sphere/cyl, you need to do so in a way that keeps the circle of least confusion on the retinaBB Removing 0.5 of cyl for first time adaption purposes in kids is fairly common practice in Australia built up over time. For CLC purposes sure we should then only be removing 0.25 sphere but in reality removing 0.50 or even more is probably no big issue given the accommodative capacity of the young eye will readily then place that CLC at the retina in a plus refraction. In other countries the considered norm is full cyclo RX from day one. I guess we were instilled with a softly softly approach to take PX comfort into consideration, while still ending up at the same final Rx

On the other hand, the majority of the commenters would prefer to prescribe a slightly reduced correction for both the hyperopia and astigmatism. A 0.50D reduction in sphere or cylinder is mentioned most frequently in the comments. There was mention of reducing the prescription such that one maintains the circle of least confusion (i.e. reducing 0.25DS for every 0.50DC change).

Commenters’ concern with partial spherical correction is that it may adversely affect the child’s accommodation whereas the concern for partial correction in cylinder is it might allow amblyopia to become more embedded. The main reasons for partial correction are for easier adaptation and that a full cycloplegic prescription may blur distance vision. Susan Cotter’s paper suggests that an optometrist’s decision on how much plus to prescribe in childhood hyperopia often considers the child’s visual function with the prescription, and these include allowing for robust function of accommodation, vergence, stereopsis and near vision.8

What the literature says:

  • Prescribe full cycloplegic correction for hyperopes with esotropia. If esotropia is absent, slight under-correction can be prescribed.9-12 The purpose of under-correction is to allow some accommodation.9 The child does not need full correction for good vision10 and under-correcting can act as stimulus for emmetropization.11,12
  • No under-correction is needed for astigmatic correction.9,10 Children are able to adapt to high cylinder power correction.9 For first correction, mild under-correction can be given for adaptation.10
  • Harvey et al showed that best corrected acuity improved from 20/51 (6/15) to 20/40 (6/12) after 4 months of spectacle correction in 3-5 year old kids with astigmatism. Both myopic/mixed and hyperopic astigmatism groups had similar improvement. Astigmatic correction of at least 2DC was given full correction in 3 to <4 year olds and at least 1.50DC for children 4 years and older.13
  • Roch-Levecq et al showed gradual improvement in visual motor integration related scores in preschooler after 6 months of spectacle correction.14 Ametropia correction was determined by cycloplegic retinoscopy as bilateral hyperopia ≥ 4 D in 3 to 5-year olds, astigmatism ≤ -2 D in 3-year olds and ≤ -1.50 D in 4- and 5-year olds, or a combination of both, and emmetropia as ≤2 D and ≥ -1 D in both eyes.

In summary, giving mild under-correction in hyperopia and astigmatism is an acceptable game plan. However, if the patient has esotropia, it is important to prescribe the full hyperopic correction from cycloplegic results; or if adaptation is a significant concern, start with an initial reduced prescription, building up to the full prescription. To read more about thresholds for prescribing in young children (up to 6 years of age), read our blog entitled How to manage the very young myope.

4. When to follow up

NP … follow up 1 month, 3 months and 6 months. Check VA with Lea, Patti Pics etc. May need to go for amblyopia therapy at a later stage depending upon the improvement in VA.JD ...and of course cover testUTJ Then i follow up in 1 to 3months depends casw by case to close monitor and check their visual acuity if its possible.

As noted in the original post, the child was reported to only achieve a binocular best corrected acuity of 0.6 Decimal, which is a little worse than 20/32 or 6/9.5. Hence, close follow up is required to monitor improvement and manage any emergent amblyopia. During the follow up, we would need to assess:

  • Visual acuity - to monitor amblyopia
  • Binocular vision status - to rule out any strabismus influencing gains in acuity, especially any esotropia/ high esophoria in this case if partially corrected
  • Keratometry/topography (where possible, due to age) - to monitor corneal as compared to refractive astigmatism.

Take home messages:

  1. In measuring paediatric refractions, we can use objective measures like retinoscopy and autorefraction results for patients where subjective measure are not possible. The research and consensus indicates, though, that retinoscopy is the preferable technique.
  2. Most colleagues agree to slightly under-correcting hyperopes and even astigmatic corrections by around 0.50D. Typically we would not under-correct myopes, although guidelines for prescribing in myopic children up to 6 years of age indicate that this may be suitable up until age 4 for myopes, but full correction after age 4 is preferable.

Further reading on prescribing for children

Meet the Authors:

About Connie Gan

Connie is a clinical optometrist from Kedah, Malaysia, who provides comprehensive vision care for children and runs the myopia management service in her clinical practice.

Read Connie's work in many of the case studies published on Connie also manages our Myopia Profile and My Kids Vision Instagram and My Kids Vision Facebook platforms.

About Kimberley Ngu

Kimberley is a clinical optometrist from Perth, Australia, with experience in patient education programs, having practiced in both Australia and Singapore.

Read Kimberley's work in many of the case studies published on Kimberley also manages our Myopia Profile and My Kids Vision Instagram and My Kids Vision Facebook platforms.

This content is brought to you thanks to unrestricted educational grant from

Back to all articles

Enormous thanks to our visionary sponsors

Myopia Profile’s growth into a world leading platform has been made possible through the support of our visionary sponsors, who share our mission to improve children’s vision care worldwide. Click on their logos to learn about how these companies are innovating and developing resources with us to support you in managing your patients with myopia.