Detection, assessment, and management of refractive errors in children are of great importance, as this ability is a fundamental clinical skill among optometrists.1 Being aware of the refractive status of the child timely and its appropriate correction are important to reduce the incidence of the amblyopia, the costs of visual rehabilitation and ultimately reduce the health and social costs of the community.2 Accordingly, children in the early years of life need comprehensive eye examinations due to the probable presence of amblyopia, strabismus, significant refractive error or other factors affecting the reduction of visual acuity.3

In most cases, evaluation of refractive error in infants and young children through routine adult procedures, such as standard retinoscopy and autorefraction, is associated with problems due to the excessive activity of the accommodative system, lack of long-term fixation at a particular point, and the failure to cooperate for various reasons,4,5 so applying these methods may not lead to optimal results. Accordingly, different researchers have proposed methods for better and more reliable results, such as, cycloplegic refraction in high hyperopia with greater accommodative efforts,6 and Mohindra’s method in children who look at the light of a retinoscope during refraction,7 each in turn eliminating some of the problems associated with the evaluation of children’s refraction, but the problem does not end with this kind of non-cooperation. Some young children, mentally retarded children, and infants may become anxious and frightened by approaching and placing the trial lens on the spectacle plane and may begin to develop inadequacy, crying, and inaction.8,9 Better cooperation between the examiner and child in the retinoscopic process will be formd due to the emotional relationship between parents and children. Infants and young children may feel less anxious when they are physically close to their parents and this brings the most comfort to children. In these cases, if a trial lens is placed farther away (the Tele-Lens) from the spectacle plane, the optical reflex can be seen through retinoscope as a reflection from the retina, and eventually, the child’s refractive status is evaluated without disturbing the child. Can the results of this Tele-Lens retinoscopy be trusted and applied?

This study aimed to introduce a new retinoscopy method called the “Mirza” tele-lens retinoscopy for evaluation of refractive error in uncooperative infants and children and to compare it with standard retinoscopy.

The introduction of the “Mirza” tele-lens retinoscopy as the topic of this study requires firstly to determine the correct location of the trial lens for retinoscopy. As the lens moves away from the patient’s faceplate, the problem of lens disturbance for the child is reduced and eliminated. The clinical experience of the authors has shown that an approximate distance of 20 cm far from the faceplate is sufficient enough to eliminate the reactive behaviors of child, which are relative to the object located near the face. On the other hand, given that the usual working distance in the standard retinoscopy procedure is often 66.6 cm10,11 if the child’s spectacle plane departs from as much as 1/3 of this distance, a point will be created that is 22.2 cm from the child’s spectacle plane and 44.4 cm away from the peephole of the retinoscope (Fig. 1).

Figure 1.

The “Mirza” tele-lens and standard retinoscopy. In A, the child’s reaction to approaching the trial lens in standard retinoscopy procedure was showed. In B, the trial lens magnification from its position farther from the spectacle plate was illustrated. In C, the “Mirza” tele-lens retinoscopy was performed by placing the trial lens at 22.2 cm far from the child’s spectacle plane.

(0.11MB).

This point has two features that help to select the optimal distance for refraction. First, this point is sufficiently far from the patient’s spectacle plane to eliminate the child’s reactive behaviors to the trial lens. Secondly, given that the lens is about 1/3 the distance between the child’s eye and the peephole of the retinoscope, evaluation of a child’s refractive error is easily obtainable by a skilled optometrist during retinoscopy despite optical changes in retinoscopic reflex.

Optics of retinoscopy with a distant lens should now be assessed. It is worth mentioning that for some reasons, choosing a distance of 1/3 would be better:

First of all, it would be easier to consider this distance than other distances because it divides the usual distance for standard retinoscopy (66.6 cm) into three parts of equal size.

Secondly, based on the author’s experiences in using this method, it has been observed that by reducing the distance and getting closer to the patient, retinoscopy will be difficult due to the lack of necessary cooperation. Also, moving away from a distance of 1/3 will make it more difficult to see the retinoscopic reflex and increase optical errors.

The study protocol was approved by the institutional review board of the Iran University of Medical Sciences and the Steering Committees of the School of Rehabilitation Sciences and conformed with the tenets of the Declaration of Helsinki. The informed consent was obtained for experimentation with human subjects from All subject’s parents.

The dry and cycloplegic results of the two methods were compared by paired t-test. Also, the intraclass correlation coefficient (ICC) between the net results of the dry and cycloplegic “Mirza” tele-lens and standard retinoscopy was determined. To assess the agreement between the two methods of retinoscopy, the Bland–Altman plots were used. Differences between measurements were plotted against their mean and the 95% limits of agreement were determined as the mean difference ±1.96 *standard deviation of the differences. Besides, a two-way repeated-measures ANOVA with factors of retinoscopy method (standard vs. “Mirza” tele-lens) and use of cycloplegic drop (dry vs wet) was performed using the refractive error values. A single expert optometrist was considered to perform the standard and “Mirza” tele-lens retinoscopy to avoid the inter examiner’s errors.

The refractive errors of 78 eyes from 39 cooperative children with a mean age of 9.69 ± 1.91 years (range from 7 to 12 years) and mean dry refractive errors of +1.39 ± 1.43 D (range from −0.50 D to +5.75 D) were evaluated. Also, the refractive findings of 60 eyes of 31 uncooperative infants with mean age 21.85 ± 8.79 months (range from 8 to 37 months) and mean refraction of +2.23 ± 1.00 D (range from +0.50 D to +4.50 D) were evaluated. It should be noted that these children were related to the kindergarten that is next to the School of Rehabilitation Sciences of Iran University and many of whose parents were employees of the school and were available.

The results of the optical calculation

In this part of the present study, by applying the physical laws of lenses and by producing refractive errors of −1.50 D to +10.00 D, the amount of correction lens power was determined at a trial lens distance of 22 cm from spectacle plane and the results were recorded in Table 1.

Table 1.

Equivalent refractive errors to the values obtained by the “Mirza” retinoscopy method.

Trial lens power at neutrality	Spectacle plane refractive error(net values)
0	−1.50
0.12	−1.45
0.25	−1.38
0.37	−1.33
0.5	−1.25
0.62	−1.20
0.75	−1.12
0.87	−1.06
1	−0.97
1.12	−0.91
1.25	−0.82
1.37	−0.74
1.5	−0.65
1.62	−0.55
1.75	−0.45
1.87	−0.35
2	−0.24
2.12	−0.13
2.25	0.00
2.37	0.12
2.5	0.26
2.62	0.40
2.75	0.56
2.87	0.72
3	0.89
3.12	1.07
3.25	1.28
3.5	1.72
3.62	1.96
3.75	2.25
3.87	2.53
4	2.86
4.12	3.20
4.25	3.59
4.37	4.00
4.5	4.50
4.62	5.01
4.75	5.63
4.87	6.27
5	7.07
5.12	7.93
5.25	9.00
5.37	10.18

Based on these results, with the change in hypermetropia from 0.00 to 10.00 D, the tele-lens power was changed from +2.25 D to +5.35 D. In other words, by changing 10 D in standard net retinoscopy values, the tele-lens power results change by 3.15 D.

Possible errors due to changes of ±3 cm in the placement of the trial lens in Fig. 2 with the range of refractive error of −1.50 D to +10.00 D are shown. As it turns out, the error first increases to a certain extent, then the changes progress to a decrease in the error, and again from +4.50 D onwards, an error-increasing trend occurs. the error rate of the 3 cm displacement of the lens in the −1.50 D of the tele-lens power is zero. This error gradually increases from zero in the range of −1.50 D to +4.50 D, and in +3.50 D (equivalent to +1.73 D hyperopic net refraction), the error rate reaches a maximum of 0.21 D and then the error rate decreases. At +4.50 D, this error reaches almost zero.

Figure 2.

The effect of working distance changes from 22.2 to 3 cm back and forth on the “Mirza” tele-lens retinoscopy.

(0.09MB).

The error rate is again increased from the point of the +4.50 D by increasing the hypermetropic refractive error and in high hyperopic refractive errors, such that in +7.00 D reaches to about 0.50 D error, in +8.00 D to 0.75 D error and in the +9.00 D and +10.00 D reaches to 1.00 D error.

The retinoscopy results of the cooperative children

The mean and standard deviation of the net result of the dry and cycloplegic standard retinoscopy were +1.39 ± 1.43 D (range from −0.50 D to +5.75 D) and +2.37 ± 1.44 D (range from 0.00 D to +7.00 D) according to subtracting 1.50 D equivalent to working distance lens, respectively.

Also, the results of dry and cycloplegic “Mirza” tele-lens retinoscopy were +1.36 ± 1.39 D (range from −0.25 D to +5.95 D) and +2.41 ± 1.37 D (range from +0.50 D to +7.00 D) according to the optical “Mirza” tele-lens retinoscopy conversion Table 1, respectively.

The mean difference between the dry and cycloplegic refraction of the two methods with their p-values are presented in Table 2.

Table 2.

Paired comparison of the standard and “Mirza” retinoscopy.

		Paired Differences
		Mean	Std. Deviation	P-value
Pair 1	Dry Standard–Dry Tele	0.03	0.23	.253
Pair 2	Cyclo Standard-Cyclo Tele	0.04	0.29	.252

As can be seen in Table 2, the comparison of net finding of dry standard retinoscopy (+1.39 D; 95% CI: +1.06 D to +1.71 D) with the dry “Mirza” tele-lens retinoscopy (+1.36 D; 95% CI: +1.04 D to +1.67 D) was not statistically significant (paired-sample t = 1.15; p = .253). Also, the comparison of mean cycloplegic value of the standard retinoscopy (+2.37 D; 95% CI: +2.04 D to +2.69 D) with the amount of the cycloplegic “Mirza” tele-lens retinoscopy (+2.41 D; 95% CI: +2.10 D to +2.72 D), was not statistically significant (paired-sample t = −1.15; p = .252). Besides, Two-way repeated measures ANOVA revealed no significant retinoscopy method × use of drops interaction (P = 0.103) in comparing two methods of standard and “Mirza” tele-lens retinoscopy.

Also, the ICC between the findings of dry and cyclo mode of the “Mirza” tele-lens and standard retinoscopy showed a similar correlation for both methods, i.e. 0.993 (95% CI: 0.989, 0.996) for dry results and 0.989 (95% CI: 0.983, 0.993) for cyclo results. The calculated 95% limits of agreement for dry and cyclo findings are shown in Fig. 3A and B. The agreement and the obtained mean differences were not statistically different (p > 0.05).

Figure 3.

Bland–Altman plots showing the agreement of the dry (A) and cycloplegic (B) results taken with the standard and “Mirza” tele-lens retinoscopic methods. The solid line indicates the mean difference, and the dotted lines indicate 95% limits of agreement.

(0.11MB).

This study was conducted to introduce a new method of retinoscopy called the “Mirza tele-lens retinoscopy”. Based on the findings of this study and the high ICC between the net results of this new method and standard retinoscopy in cooperative older children, it can be said that this method is an alternative to standard retinoscopy with acceptable precision and accuracy and can be used in some young mentally retarded children and infants which are anxious and frightened by approaching and placing the trial lens on the spectacle plane. Using this new method, the refractive status of the eye can be assessed with minimal discomfort for the tested person. Since the evaluation of refractive error with standard retinoscopy method in non-cooperative children does not lead to accurate results and viewing the initial retinoscopic reflex may be with some difficulties and in the most case actually fails,9 this new method takes into account a solution such as increasing the distance of the correction lens from the child and the least contact with the child that is more optimal for assessment of refractive error.

It is clear that the “Mirza” tele-lens retinoscopy performance, like the standard retinoscopy method, needs to acquire skills with continuous practice to be done correctly, which means that at first, one may have difficulty viewing and assessing the retinoscopic reflex. Therefore, the “Mirza” tele-lens retinoscopy with practice and proficiency can be performed almost as quickly as the standard retinoscopy by a skilled optometrist in some cases that do not have the mentioned necessary cooperation for the relevant procedure. In the early days of beginning the “Mirza” tele-lens retinoscopic procedure, to facilitate the maintenance of the correct working distance, the tie a string to the head of the scope with knots at a distance of 66.6 and 22.2 cm from the spectacle plane was adjusted. After performing this procedure for a long time and increasing the skill and gaining experience by an optometrist in performing the “Mirza” tele-lens retinoscopic method, no string was used to maintain the working distance, although the authors of the article initially advised using this string to maintain the correct working distance. Besides, the lighting conditions of the room were changed to dim light in order to perform the “Mirza tele-lens retinoscopy” as required for the standard retinoscopy.

It is clear that most infants are hyperopic and the average cycloplegic refractive error is approximately +2.00 D with a standard deviation of approximately 2.00D.12 On the other hand, the refractive error value is equal to +4.50 D when the neutral point was observed with the trial lens of +4.50 D in the “Mirza” tele-lens retinoscopy, therefore this lens as a starting point in this retinoscopy procedure can be used with sufficient accuracy and speed to detect refractive errors and in turn, to manage the visual functions of infants with visual disorders such as strabismus, accommodative anomaly, and amblyopia because the hypermetropic values of more than 4.00 D in early life in infants can mostly cause such problems.13–15

According to the results of the present study, when the hyperopic changes are 4.50 D from 0.00 D to +4.50 D, the “Mirza” tele-lens retinoscopic changes occur within a range of 2.25 D from +2.25 to +4.50 D. When the net standard retinoscopic changes are 5.50 D from +4.50 D to +10 D, the equivalent spectacle plane hyperopia in the “Mirza” tele-lens retinoscopy will be only 0.85 D from +4.50 to +5.35 D, so the results are very close to each other in this method and the different amounts of hypermetropia in theses ranges are a bit difficult to distinguish, although the brightness and motion of the retinoscopic reflex will be very favorable due to the magnification caused by the trial lens at a distance of 22.2 cm of the spectacle plane. Therefore, applying this new method to evaluate the refraction of infants requires a great deal of precision, as in higher refractive error may require the use of trial lenses with steps of ±0.12 D for neutralization procedure. Although the “Mirza” tele-lens retinoscopy method can be a good alternative to the standard retinoscopy in uncooperative infants, given the low range changes of corrective lens and the high potential for a significant error to reach the neutralization endpoint in very high hyperopic values, it is recommended that firstly, at higher hypermetropic values, more precision should be used to detect the ‘with’ or ‘against’ movement of the retinoscopic reflex, secondly, it is important to note that the accuracy of this method is reduced in very high values of hypermetropia. For example, according to Table 1, it is clear that by changing the amount of hypermetropia from 9.00 D to 10.00 D, the results of the “Mirza” tele-lens retinoscopy change from +5.25 D to +5.35 D, which means that for 1.00 D change in this hyperopic range, only 0.10 of a change in the “Mirza” tele-lens retinoscopy is seen. Another important point that needs to be explained is to evaluate the myopic infants because the minification of the negative lens increases with increasing the distance from the spectacle plane, which in turn makes the evaluation of the refractive error difficult and inaccurate, especially in amounts more than 3.00 D of myopia. In this study, patients with the refraction of more than 1.50 D myopic were also present that their number was very limited. On the other hand, the myopia of 1.50 D could be neutralized without any lenses at a working distance of 66.6 cm. And also myopia of 2.00 D could be neutralized without any lenses at a working distance of 50 cm and myopia of 5 D at 20 cm, so we do not need to use “Mirza” tele-lens retinoscopy at myopia more than 1.50 D.

It is worth mentioning that due to other reasons such as the probability of increasing the surface scattering and optical aberration of the trial lens in high myopic refractive errors, and the need for trial lenses outside the available lenses in the clinic’s trial lens set, the values above 1.50 D myopia have not been reported. In fact, the “Mirza” tele-lens retinoscopy is not practical in myopic cases above 1.50 D myopia, and in these cases, the radical retinoscopy can be considered a more appropriate alternative because by reducing the distance, a more visible reflex can be seen, so retinoscopy will be easier and can be performed without any trial lens.

As mentioned previously, in this new method, the trial lens is placed at a distance of 22.2 cm instead of on the spectacle plane and near the eyes. Therefore, a novice refractionist may not observe the exact distance. If this distance is incorrectly considered, for example, 3 cm forward (19.2 cm) or backward (25.2 cm), changes will occur based on the optics of the method.

The error rate of the 3 cm displacement of the lens increases from zero to 0.21 D in the refractive error ranges of −1.50 D to +3.50 D, and then gradually decreases up to the refraction of +4.50 D that reaching almost zero in this point. After this point of +4.50 D, the error gradually increases and this error rate reaches 1.00 D in high hyperopic refractive errors about 10.00 D.

Due to the most children with binocular anomaly such as accommodative esotropia that often have the average cycloplegic refractive error of +4.75 D with ranges between +1.50 and +7.00 D,16 the probability of error by changing the distance will reach its lowest level in these ranges (0.12 D to 0.50 D). Therefore, the maximum error of the 0.50 D is completely negligible and could not cause any problems in the course of clinical practice comparing to the facilities it provides for us.

In fact, this method is recommended when the standard method is not applicable, and therefore, methods such as Hirschberg test for measuring the angle of strabismus that is used in the ophthalmic clinic however is not very accurate, it is very helpful when the practitioner is forced to use.

Another way to reduce errors in high refractive hyperopic values is to instantly and for a moment placing the lens on the spectacle plane and neutralizing the retinoscopic reflex in the standard method. Although the child does not allow the lens to be close to his eyes, a skilled refractionist can capture this moment.

Despite all these ways of error reduction, the authors accept that this method requires the acquisition of skills to keep the working distance lens in its correct position to increase retinoscopic accuracy and reduce the error rate.

As a report of new technique that explores the possibility of doing the refraction by this technique and not validating the technique, it may be overlooked.

To determine the validity of the new technique of retinoscopy, the non cycloplegic and cycloplegic “Mirza” tele-lens retinoscopy findings were compared with the gold standard retinoscopy findings by using two methods of the intraclass correlation coefficient (ICC), and Bland–Altman plots. As the results of the present study showed, there were high agreement values (ICC higher than 0.98) between results provided from standard and the “Mirza” tele-lens retinoscopy in both situations of dry and cycloplegic refraction. These values between the results of these two retinoscopic methods indicate a strong relationship between them, which means that the results of the “Mirza” tele-lens retinoscopy are sufficiently accurate and valid, so this method of retinoscopy is useful in evaluating the refractive errors especially in poorly or un-cooperated children instead of the standard dry and cycloplegic retinoscopy method. Two-way repeated measure ANOVA showed a lack of interaction between the two factors, so the “Mirza” tele-lens retinoscopy results in both cyclo and dry mode were consistent with the standard retinoscopic results. In addition, the Bland–Altman analysis for the agreement of dry and cyclo findings of the “Mirza” tele-lens retinoscopy with standard retinoscopy showed no significant differences in the 95% limits of agreement with shorter range for dry than the cyclo results of the “Mirza” tele-lens retinoscopy (Table 2 and Fig. 3), suggesting that the data obtained from these measurements from the two different methods of retinoscopy can be interchanged. In fact, this way can be used to examine validity by assessing the level of agreement by visualizing the data with Bland–Altman plots. According to the mentioned results, a good argument between the data of the two retinoscopic methods indicates the optimal validation of the “Mirza” tele-lens retinoscopy method.

It is essential to pay attention to some tips:

First, the distance of 22.2 cm (1/3 of the standard retinoscopic distance) was considered in the present study to make the optical calculations as easily as possible.

Second, careful observance of the distance of 22.2 to perform in the present study is necessary because the change of this distance can cause mistakes in the evaluation of the high refractive errors and optical calculations.

Third, the optometrists and ophthalmologists who work in the field of pediatric visual assessment can consider distances other than 22.2 cm, in fact, any distance that leads to the child’s cooperation in the retinoscopy procedure can be considered for this purpose.

Fourth, the conversion optical calculations should be made out according to any distance provided for lens placement (more or less than 22.2 cm) based on the items mentioned in the present study.

The notable limitations of the present study are:

• 1

  To perform this new method for assessing the refraction in children with non-cooperation, it is necessary to acquire the incumbent skills, so the examiner must be an expert. there are challenges at beginning the standard retinoscopy, that over time by repeating the procedure and gaining skills, the procedure will be easier to perform and the refraction will be evaluated more accurately.

• 2

  It is necessary to observe the distance of the trial lens in order to reduce the measurement error during the retinoscopy process. However, the errors made by changing the distance of 3 cm to the desired position of the trial lens (22.2 cm) are discussed in the article.

• 3

  In high amounts of hyperopia, it is very important to use the trial lens with power steps of 0.12 D to neutralize the refractive errors, and in high myopia, as mentioned, it is better to use alternative methods such as radical retinoscopy, and the tele-lens method will not be accurate enough in such these cases, and better not to be used.

It should be noted that, when the standard retinoscopy procedure does not apply to uncooperative infants and children, some articles recommend the use of sedative agents and ocular examination under more invasive conditions. In this situation, if the distance of the trial lens to the spectacle plane increases which, in turn, most often causes the child to cooperate in the retinoscopy process, then the “Mirza” tele-lens retinoscopic method becomes applicable, meaning this new method may prevent the need to examine under anesthesia because this situation can be associated with risks.17,18

Given that most infants and children are hypermetropic and that 4.50 D of hyperopia in the “Mirza” tele-lens retinoscopic method is equivalent to this amount and starting point of the most of the accommodative problems and anomalies will begin from these values so this method can be used in the evaluation of refraction of infants and children who do not have the necessary cooperation for the standard retinoscopic procedure.

None declared.

The authors have no conflicts of interest to declare.