Journal of Pediatric Ophthalmology and Strabismus

Original Article 

Comparison of the Effect of Cycloplegia on Astigmatism Measurements in a Pediatric Amblyopic Population: A Prospective Study

Sunali Goyal, MD; Paul H. Phillips, MD; Mallikarjuna Rettiganti, PhD; Jeffrey M. Gossett, MS; R. Scott Lowery, MD

Abstract

Purpose:

To study the effect of cycloplegia on astigmatism measurements in pediatric patients with amblyopia.

Methods:

This was a prospective comparative clinical study. Participants 4 to 17 years old were recruited from the patient population at the Arkansas Children's Hospital eye clinic after informed consent was obtained. Autorefractor measurements were used to obtain values of refractive error in amblyopic and non-amblyopic patients before and after cycloplegia. The groups were subdivided into myopia and hyperopia and with and without underlying amblyopia. The refractive error was expressed as sphere, cylinder, axis of astigmatism, and spherical equivalent. The treatment effect was summarized as the mean difference (95% confidence interval) for each outcome.

Results:

No statistically significant difference was found on the axis and power of astigmatism before and after cycloplegia in the patients with amblyopia (P = .28 and .99, respectively).

Conclusions:

Non-cycloplegic autorefraction measurements may be considered safe for refining astigmatism power and axis in pediatric patients with amblyopia.

[J Pediatr Ophthalmol Strabismus. 2018;55(5):293-298.]

Abstract

Purpose:

To study the effect of cycloplegia on astigmatism measurements in pediatric patients with amblyopia.

Methods:

This was a prospective comparative clinical study. Participants 4 to 17 years old were recruited from the patient population at the Arkansas Children's Hospital eye clinic after informed consent was obtained. Autorefractor measurements were used to obtain values of refractive error in amblyopic and non-amblyopic patients before and after cycloplegia. The groups were subdivided into myopia and hyperopia and with and without underlying amblyopia. The refractive error was expressed as sphere, cylinder, axis of astigmatism, and spherical equivalent. The treatment effect was summarized as the mean difference (95% confidence interval) for each outcome.

Results:

No statistically significant difference was found on the axis and power of astigmatism before and after cycloplegia in the patients with amblyopia (P = .28 and .99, respectively).

Conclusions:

Non-cycloplegic autorefraction measurements may be considered safe for refining astigmatism power and axis in pediatric patients with amblyopia.

[J Pediatr Ophthalmol Strabismus. 2018;55(5):293-298.]

Introduction

Amblyopia affects 2% to 5% of all children. The key to management is early detection and treatment, which often includes spectacle correction of refractive error. Autorefractors are frequently used to obtain objective measurement of refractive error. Autorefraction is automated retinoscopy that determines the refractive error of each eye. This instrument is particularly useful when evaluating pediatric patients due to difficulty in obtaining reliable information from subjective refraction or retinoscopy. Kemper et al.1 determined that, for children ages 3 to 5 years, the sensitivity and specificity of monocular autorefraction for detecting amblyogenic pathology is 88% and 58%, respectively. Patient dilation is not typically done on every visit, and this can lead to errors in determining the refractive error. Previous studies have been conducted that compare cycloplegic and non-cycloplegic refraction and have mostly concluded that non-cycloplegic refractions are inaccurate due to patient accommodation.1–3 There is a greater risk of this error in pediatric patients, who have greater accommodative amplitudes. This effect can be more pronounced when using autorefraction instruments because fixation targets may stimulate accommodation. Of course, accommodation can affect spherical equivalent, but it is unknown whether accommodation affects magnitude and axis of astigmatism. Hence, knowledge of the change in refractive error status before and after cycloplegia is useful to determine the accuracy of non-cycloplegic autorefractions. It is also known that amblyopic eyes have deficient accommodation.1 However, no studies have specifically looked into whether cycloplegia affects magnitude of refractive error in amblyopic patients. To the best of our knowledge, none have addressed the change in axis and degree of astigmatism before and after cycloplegia in the subset of patients with amblyopia.

The purpose of the current study was to determine the change in axis and degree of astigmatism before and after cycloplegia in normal children and children with amblyopia. To our knowledge, this is the first study to address the change in axis and degree of astigmatism that occurs in pediatric amblyopic patients after cycloplegia using a table-mounted autorefractor.

Patients and Methods

This study was a prospective comparative study. Participants, aged 4 to 17 years, were recruited from the patient population at the Arkansas Children's Hospital eye clinic. The study was conducted in accordance with the tenets of the Declaration of Helsinki. A detailed informed consent was obtained at the time of each patient's clinic visit in accordance with the institutional review board.

All patients between 4 and 17 years old who had a planned dilation scheduled at the time of the study and were cooperative enough to obtain an autorefraction were included in the study. Amblyopia was defined as visual acuity of 20/40 or worse, with at least two lines of difference in visual acuity between the two eyes. Any organic pathology and amblyopia due to strabismus, anisometropia, or both were ruled out before inclusion. Patients with pseudophakia, aphakia, and poor cooperation were excluded from the study.

Information about the patient's age, gender, race, and ocular history (including known eye disease or previous surgery) were recorded. Dry autorefraction (measurement prior to cycloplegia), cycloplegic autorefraction, and pupil size after dilation were measured and documented. Cycloplegia was accomplished using the standard regimen of phenylephrine (2.5%), tropicamide (1%), and cyclopentolate (1%) (Alcon Laboratories, Inc., Fort Worth, TX) repeated three times at an interval of 10 minutes. The cycloplegic measurements were obtained 30 minutes or more after administration of these drops (wet measurements). All autorefractions were performed using the table-mounted Canon RK-3 autorefractor (Vision Systems Inc., Otawara, Japan). Refractive error was expressed as sphere, cylinder, axis of astigmatism, and spherical equivalent. Measurements were recorded for right and left eyes. The data were then entered into the database by data administrators.

Two subsets of patients were analyzed separately: hyperopia and myopia. Paired-sample t tests were used to compare “wet” outcomes versus “dry” among the myopia and hyperopia groups, overall and among those with and without amblyopia in each of the two groups. The treatment effect was summarized as the mean difference (95% confidence interval [CI]) for each outcome. All unadjusted summaries graphs were made with R software (version 3.2.2; R Foundation, Vienna, Austria). Within each subset (hyperopia and myopia) and for each of the four measurements (axis, cylinder, spherical equivalent, and sphere), a linear mixed model was fit to assess the difference in wet and dry measurements, adjusting for age, sex, race, and amblyopia. The mixed models included random intercept variance components for patient and eye because both wet and dry measurements were repeated on both eyes of the same patient. Least squares mean estimates (95% CIs) of wet and dry measurements are presented, as well as the estimated wet minus dry differences (with 95% CI). Mixed models were fit with SAS software (version 9.4; SAS, Inc., Cary, NC).

Results

A total of 38 patients with ages ranging from 4 to 17 years were included in the study (mean age = 8 years). Two patients (6 and 11) were excluded from the study because cycloplegic measurements were not obtained. Characteristics of the remaining 36 patients (72 eyes) are listed in Table 1. Wet and dry group means ± standard deviations are presented in Table 2 (with and without amblyopia).

Summary of Sample Characteristics

Table 1:

Summary of Sample Characteristics

Summaries of Wet vs Dry Measurements

Table 2:

Summaries of Wet vs Dry Measurements

A mixed-effects model approach was used to compare wet and dry eye measurements in the myopic and hyperopic subsets, adjusting for amblyopia, age, race, and sex (Table 3). The models included random effects for correlation of measurements on the same patient and eye. In the myopia group, no statistically significant difference was found in the axis or cylinder value before or after cycloplegia (P = .28 and .99, respectively). However, there was a statistically significant decrease in myopic sphere and resultant myopic spherical equivalent in wet versus dry measurements (P < .0001). No statistically significant difference was found in any of the hyperopic values (axis, cylinder, sphere, or spherical equivalent; P = .51, .26, .16, and .11, respectively).

Model-Based Estimates and Comparisons by Whether Myopic or Hyperopic

Table 3:

Model-Based Estimates and Comparisons by Whether Myopic or Hyperopic

Patients were divided again into four subsets based on hyperopic and myopic status, but this time presence or absence of amblyopia was added as an additional variable (Table 4). In the myopic subset of patients, no statistically significant difference was found in the axis or cylinder in wet versus dry measurements irrespective of the patient having underlying amblyopia (P = .49 and .29, respectively) or not (P = .34 and .09, respectively) (Figure 1). However, the myopic sphere and spherical equivalent values were statistically significantly decreased after cycloplegia measurements were done in patients with (P = .0022 and .0024, respectively) and without (P < .0001) amblyopia for both values. When evaluating the hyperopia, the axis and cylinder value remained unchanged before and after cycloplegia in patients both with (P = .91 and .084, respectively) and without (P = .42 and .51, respectively) amblyopia (Figure 1). Interestingly, there was a statistically significant increase in sphere (P = .0016) and spherical equivalent value (P = .0035) only in the hyperopic group that had underlying amblyopia.

Model-Based Estimates and Comparisons by Whether Myopic/Hyperopic and Whether Patient Has Amblyopia

Table 4:

Model-Based Estimates and Comparisons by Whether Myopic/Hyperopic and Whether Patient Has Amblyopia

Scatter plot showing correlation of (A) axis and (B) cylinder before and after cycloplegia in patients with myopia and hyperopia with and without amblyopia.

Figure 1.

Scatter plot showing correlation of (A) axis and (B) cylinder before and after cycloplegia in patients with myopia and hyperopia with and without amblyopia.

Discussion

Determining refractive error in the pediatric age group is influenced by accommodation and hence is a concern in the pediatric population. It is still debatable as to how accurately noncycloplegic refraction can be compared and used in clinical settings. Previous studies have opined that non-cycloplegic refraction in the pediatric age group may be unreliable.4–6 Most studies have found overestimation of myopia and underestimation of hyperopia in undilated refractions due to the role of accommodation.7,8 However, pediatric ophthalmologists often face challenges when performing cycloplegia because of patient cooperation, extended visits, and patient and family desire to avoid eye drops. The traditional methods of non-cycloplegic and cycloplegic retinoscopy are valuable tools to estimate refractive errors in children, but require trained examiners and cooperative patients. Many studies have evaluated the effect on cycloplegia of refractive error measurements in children using autorefractors. However, most studies used handheld autorefractors. Handheld autorefractors may not be as accurate as table mounted autorefractors, but nonetheless are useful in special challenging situations.

In the current study, a table-mounted autorefractor with a fogging mechanism was used for all measurements. Our study concurred with previous studies regarding falsely high values of myopia in non-cycloplegic measurements. However, in contrast to previous studies, our study did not find any significant difference in hyperopia measurements. Significant uncorrected astigmatism can lead to amblyopia.9,10 Astigmatism values of more than 1.50 diopters and oblique astigmatism increase the incidence of amblyopia, hence the need for reliable tools to diagnose amblyopia accurately.10,11 The most interesting finding in the current study was that across the study population there was no significant change in axis of astigmatism or value of cylindrical power with or without cycloplegia, which may suggest that patients with pure astigmatism may benefit from noncycloplegic autorefraction. This result gives pediatric ophthalmologists the assurance to not only detect astigmatism (quantity and axis) with confidence, but also to tweak and adjust glasses in children using non-cycloplegic autorefractors and thus avoid the need to dilate them at each visit. The current study did find that patients with amblyopia and hyperopia as the underlying refractive error were underestimated if no cycloplegia was used. The likelihood of fewer patients may explain this trend.

The information provided by the current study would benefit pediatric patients by facilitating more accurate spectacle prescriptions with the least amount of diagnostic testing. The study has the obvious limitations of a small sample size and patients not classified according to age or accommodative amplitude. We believe further multicenter studies in this area may be beneficial for further utilization of this information with more confidence.

References

  1. Kemper AR, Margolis PA, Downs SM, Bordley WC. A systematic review of vision screening tests for the detection of amblyopia. Pediatrics. 1999;104:1220–1222.
  2. Hu YY, Wu JF, Lu TL, et al. Effect of cycloplegia on the refractive status of children: the Shandong children eye study. PLoS One. 2015;10:e0117482. doi:10.1371/journal.pone.0117482 [CrossRef]
  3. Cordonnier M, Dramaix M. Screening for refractive errors in children: accuracy of the hand held refractor Retinomax to screen for astigmatism. Br J Ophthalmol. 1999;83:157–161. doi:10.1136/bjo.83.2.157 [CrossRef]
  4. Zhao J, Mao J, Luo R, Li F, Pokharel GP, Ellwein LB. Accuracy of noncycloplegic autorefraction in school-age children in China. Optom Vis Sci. 2004;81:49–55. doi:10.1097/00006324-200401000-00010 [CrossRef]
  5. Hopkins S, Sampson GP, Hendicott P, Lacherez P, Wood JM. Refraction in children: a comparison of two methods of accommodation control. Optom Vis Sci. 2012;89:1734–1739. doi:10.1097/OPX.0b013e318277182c [CrossRef]
  6. Fotedar R, Rochtchina E, Morgan I, Wang JJ, Mitchell P, Rose KA. Necessity of cycloplegia for assessing refractive error in 12-year-old children: a population-based study. Am J Ophthalmol. 2007;144:307–309. doi:10.1016/j.ajo.2007.03.041 [CrossRef]
  7. Lan W, Zhao F, Lin L, et al. Refractive errors in 3–6 year-old Chinese children: a very low prevalence of myopia?PLoS One. 2013;8:e78003. doi:10.1371/journal.pone.0078003 [CrossRef]
  8. Rotsos T, Grigoriou D, Kokkolaki A, Manios N. A comparison of manifest refractions, cycloplegic refractions and retinoscopy on the RMA-3000 autorefractometer in children aged 3 to 15 years. Clin Ophthalmol. 2009;3:429–431. doi:10.2147/OPTH.S5145 [CrossRef]
  9. Simons K. Preschool vision screening: rationale, methodology and outcome. Surv Ophthalmol. 1996;41:3–30. doi:10.1016/S0039-6257(97)81990-X [CrossRef]
  10. Wasserman RC, Croft CA, Brotherton SE. Preschool vision screening in pediatric practice: a study from the Pediatric Research in Office Settings (PROS) network. American Academy of Pediatrics. 1992;89:834–838.
  11. American Academy of Pediatrics, Committee on Practice and Ambulatory Medicine, Section on Ophthalmology. Eye examination and vision screening infants, children, and young adults. Pediatrics. 1996;98:153–157.

Summary of Sample Characteristics

CharacteristicNValue
Age (years)36
  Mean ± SD8.1 ± 2.8
  Median (Q1, Q3)7.6 (6.0, 9.9)
  Range4.2 to 16.4
Sex36
  Female16 (44%)
  Male20 (56%)
Race36
  Black5 (14%)
  White31 (86%)
Ethnicity31
  Hispanic2 (6%)
  Non-Hispanic29 (94%)
Amblyopia36
  No19 (53%)
  Yes17 (47%)
Subset36
  Myopic22 (61%)
  Hyperopic14 (39%)

Summaries of Wet vs Dry Measurements

ParameterWetDry


No.Mean ± SDNo.Mean ± SD
Myopia group
  Sphere
    OD22−0.55 ± 2.4022−1.80 ± 2.59
    OS22−0.69 ± 2.5222−1.66 ± 2.25
  Cylinder
    OD220.89 ± 1.37220.98 ± 1.33
    OS221.05 ± 1.20220.95 ± 1.14
  Axis
    OD22105.55 ± 47.712290.36 ± 40.37
    OS22110.09 ± 40.5422105.95 ± 48.03
  Spherical equivalent
    OD22−0.10 ± 2.0522−1.31 ± 2.35
    OS22−0.17 ± 2.2622−1.18 ± 2.04
Hyperopia group
  Sphere
    OD142.75 ± 2.50142.48 ± 2.56
    OS142.70 ± 2.61142.32 ± 2.66
  Cylinder
    OD140.79 ± 0.57140.80 ± 0.72
    OS140.73 ± 0.71140.89 ± 0.87
  Axis
    OD1496.29 ± 39.8914111.21 ± 42.44
    OS14101.36 ± 43.461496.36 ± 47.07
  Spherical equivalent
    OD143.14 ± 2.62142.88 ± 2.81
    OS143.06 ± 2.79142.77 ± 2.94
Myopia without amblyopia group
  Sphere
    OD12−0.96 ± 2.8612−1.60 ± 2.71
    OS12−1.00 ± 2.5112−1.67 ± 2.50
  Cylinder
    OD121.08 ± 1.80121.06 ± 1.72
    OS121.04 ± 1.50120.92 ± 1.39
  Axis
    OD12105.92 ± 46.331282.75 ± 48.92
    OS12103.17 ± 43.541299.67 ± 52.69
  Spherical equivalent
    OD12−0.42 ± 2.3612−1.07 ± 2.25
    OS12−0.48 ± 2.2212−1.21 ± 2.25
Hyperopia without amblyopia group
  Sphere
    OD71.36 ± 0.7071.46 ± 1.78
    OS71.36 ± 1.5171.43 ± 2.41
  Cylinder
    OD70.61 ± 0.4870.54 ± 0.59
    OS70.46 ± 0.4770.43 ± 0.45
  Axis
    OD790.86 ± 54.047112.29 ± 50.06
    OS7121.43 ± 54.947121.57 ± 54.96
  Spherical equivalent
    OD71.66 ± 0.8871.73 ± 1.86
    OS71.59 ± 1.5271.64 ± 2.46

Model-Based Estimates and Comparisons by Whether Myopic or Hyperopic

SubsetMeasurementAdjusted Wet Mean (95% CI)aAdjusted Dry Mean (95% CI)aAdjusted Difference: Wet–Dry (95% CI)aP
MyopicAxis103.9 (87.5 to 120.2)94.2 (77.9 to 110.5)9.7 (−8.2 to 27.5).28
MyopicCylinder1.7 (1.0 to 2.4)1.7 (1 to 2.4)0.0 (−0.1 to 0.1).99
MyopicSpherical equivalent−0.4 (−1.8 to 1)−1.5 (−2.9 to −0.1)1.1 (0.7 to 1.6)< .0001
MyopicSphere−1.3 (−2.8 to 0.2)−2.4 (−3.9 to −0.9)1.1 (0.7 to 1.6)< .0001
HyperopicAxis92.8 (67.3 to 118.4)97.8 (72.2 to 123.3)−5.0 (−20.1 to 10.2).51
HyperopicCylinder0.8 (0.3 to 1.3)0.9 (0.4 to 1.4)−0.1 (−0.2 to 0.1).26
HyperopicSpherical equivalent2.4 (0.2 to 4.6)2.2 (0.0 to 4.4)0.3 (−0.1 to 0.7).16
HyperopicSphere2.0 (0.0 to 4.1)1.7 (−0.3 to 3.7)0.3 (−0.1 to 0.7).11

Model-Based Estimates and Comparisons by Whether Myopic/Hyperopic and Whether Patient Has Amblyopia

SubsetAmblyopiaMeasurementWet Mean (95% CI)aDry Mean (95% CI)aDifference: Wet–Dry (95% CI)aP
MyopicNoAxis99 (74.8 to 123.1)85.6 (61.5 to 109.8)13.3 (−15 to 41.7).34
MyopicNoCylinder1.9 (0.8 to 3)1.8 (0.8 to 2.9)0.1 (0 to 0.2).090
MyopicNoSpherical equivalent−0.7 (−2.8 to 1.4)−1.4 (−3.5 to 0.7)0.7 (0.5 to 0.9)< .0001
MyopicNoSphere−1.7 (−4 to 0.7)−2.3 (−4.6 to 0)0.7 (0.5 to 0.8)< .0001
MyopicYesAxis111.3 (82.5 to 140.2)106.1 (77.2 to 134.9)5.3 (−10.4 to 20.9).49
MyopicYesCylinder1.3 (0.7 to 1.8)1.4 (0.8 to 1.9)−0.1 (−0.3 to 0.1).29
MyopicYesSpherical equivalent0.5 (−1.5 to 2.5)−1.1 (−3.1 to 0.9)1.6 (0.6 to 2.6).0024
MyopicYesSphere−0.1 (−2.3 to 2)−1.8 (−4 to 0.4)1.7 (0.7 to 2.6).0022
HyperopicNoAxis96.5 (34.4 to 158.5)107.2 (45.2 to 169.3)−10.8 (−38.8 to 17.2).42
HyperopicNoCylinder0.7 (0.1 to 1.4)0.7 (0 to 1.3)0.1 (−0.1 to 0.2).51
HyperopicNoSpherical equivalent1.3 (−1.3 to 3.8)1.3 (−1.2 to 3.9)−0.1 (−0.8 to 0.6).85
HyperopicNoSphere0.9 (−1.4 to 3.2)1 (−1.3 to 3.3)−0.1 (−0.8 to 0.6).78
HyperopicYesAxis89.6 (65.3 to 113.9)88.8 (64.5 to 113.1)0.9 (−15 to 16.7).91
HyperopicYesCylinder0.8 (−0.3 to 1.9)1.1 (0 to 2.2)−0.2 (−0.5 to 0).084
HyperopicYesSpherical equivalent3.5 (−0.3 to 7.4)2.9 (−0.9 to 6.8)0.6 (0.2 to 1).0035
HyperopicYesSphere3.1 (−0.3 to 6.6)2.4 (−1.1 to 5.8)0.7 (0.3 to 1.1).0016
Authors

From the Department of Ophthalmology, Jones Eye Institute (SG, PHP, RSL), Pediatric Ophthalmology, Department of Ophthalmology (SG, PHP, RSL), and the Section of Biostatistics, Department of Pediatrics (MR, JMG), Arkansas Children's Hospital, University of Arkansas for Medical Sciences, Little Rock, Arkansas.

The authors have no financial or proprietary interest in the materials presented herein.

Correspondence: Sunali Goyal, MD, Jones Eye Institute, University of Arkansas for Medical Sciences, 4301 W. Markham Street, #523, Little Rock, AR 72205. E-mail: sgoyal@uams.edu

Received: May 30, 2017
Accepted: September 28, 2017
Posted Online: June 19, 2018

10.3928/01913913-20180410-02

Sign up to receive

Journal E-contents