Home About us Editorial board Ahead of print Current issue Search Archives Submit article Instructions Subscribe Contacts Login 
  • Users Online: 3819
  • Home
  • Print this page
  • Email this page

   Table of Contents      
Year : 2020  |  Volume : 68  |  Issue : 11  |  Page : 2458-2461

Do all children need a cycloplegic refraction? A comparison of Mohindra's versus cycloplegic refraction

1 Institute of Ophthalmology, Jawaharlal Nehru Medical College, AMU, Aligarh, Uttar Pradesh, India
2 Aditya Birla Sankara Nethralaya, Kolkata, West Bengal, India

Date of Submission03-Feb-2020
Date of Acceptance06-May-2020
Date of Web Publication21-Sep-2020

Correspondence Address:
Prof. Abadan K Amitava
Institute of Ophthalmology, JNMC, Aligarh Muslim University, Aligarh - 202 001, Uttar Pradesh
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/ijo.IJO_229_20

Rights and Permissions

Purpose: In 1–12 years old children, we assessed correlation, regression, and agreement between spherical equivalents (SE) obtained on Mohindra's near retinoscopy (MNR) and the post cycloplegic refraction (PCRef), performed 72 h after a cycloplegic refraction (CRef) using cyclopentolate 1% drops. Methods: In this prospective comparative study, Mohindra's near retinoscopy (MNR) was performed on 202 eyes of 101 children, from 50 cm with a streak retinoscope, in a dimly lit room, subtracting 1.25 from the trial lens used for neutralization, to obtain the final refraction. Subsequently we undertook CRef, half-hour after instilling 1% cyclopentolate, with a PCRef 72 h later. All refractive data were converted to SE for evaluation. We compared the SEs using correlation, linear regression, and agreement (Bland–Altman graphic analysis) and paired t-test. Significance was set at P ≤ 0.05. Results: The mean SE on MNR was 1.71 ± 2.49 D compared to 1.43 ± 2.42 D on PCRef. A significant correlation with r = 0.97 (r2= 0.94, P < 0.001) existed. Agreement analysis suggested that MNR overestimates hypermetropia and underestimates myopia each by 0.3 D than the standard procedure of CRef-PCRef. The regression analysis suggested that SE on PCRef is 95% of that on MNR, less 0.20. Conclusion: Our study suggests that MNR offers single point refraction very similar to CRef-PCRef, and may be considered as a viable option more often.

Keywords: Cycloplegic refraction, Mohindra's near retinoscopy, post cycloplegic refraction

How to cite this article:
Kauser F, Gupta Y, Amitava AK, Saxena J, Raza S A, Masood A, Alam MS. Do all children need a cycloplegic refraction? A comparison of Mohindra's versus cycloplegic refraction. Indian J Ophthalmol 2020;68:2458-61

How to cite this URL:
Kauser F, Gupta Y, Amitava AK, Saxena J, Raza S A, Masood A, Alam MS. Do all children need a cycloplegic refraction? A comparison of Mohindra's versus cycloplegic refraction. Indian J Ophthalmol [serial online] 2020 [cited 2020 Nov 27];68:2458-61. Available from: https://www.ijo.in/text.asp?2020/68/11/2458/295629

Refractive errors are one of the most common cause of avoidable and treatable blindness worldwide,[1],[2],[3],[4],[5] and in younger children if uncorrected, can lead to amblyopia.[1],[5],[6] In younger children, if detected and appropriately corrected, amblyopia can be either prevented or remedied with appropriate therapy. Thus, refraction forms an important part of the examination of the eyes in children. Usually this involves one of three approaches: (A) performing static retinoscopy, when a co-operative person views the visual acuity chart at optical infinity viz. 20 feet. (B) by retinoscoping while suspending the accommodation of the eye by the use of topical cycloplegic drugs, or (C) by doing near retinoscopy as described by Mohindra.[7],[8] All these methods basically involve holding the accommodation steady: an important criteria for obtaining the correct refractive state. The younger population have stronger accommodation, which gradually decreases with age.[9],[10] This accommodation is eliminated by using topical cycloplegic drugs, of varying potency, appropriate for the age. The commonly used drugs are 1% topical tropicamide for adults, and 1% atropine or 1% cyclopentolate for younger children and 0.5% of the latter in infants.[11],[12],[13],[14] Despite being effective, they have their downsides: in the case of atropine, this often induces a prolonged cycloplegia associated with photophobia; occasionally acute psychosis occurs, more commonly with cyclopentolate while hyperpyrexia and flushing along with local allergic reactions may happen too, along with the need of a repeat visit for a postcycloplegic test.[15],[16],[17],[18],[19]

In 1977, Indra Mohindra popularized a near retinoscopy technique performed from 50 cm in an otherwise dark room, which dispensed with the use of any cycloplegics. The retinoscope filament is the only interesting thing that the child focuses on, thus stabilizing the accommodation to an extent that reliable measurements of retinoscopy are obtained.[20] Allowing for “dark myopia,” a subtraction of 1.25 D is made from the power of lens used for neutralization from a half meter.[7],[8],[21] Despite being quick, Mohindra's near retinoscopy (MNR) has the advantage of reducing the number of visits for the children/parents while avoiding systemic and neurological side effects of atropine and cyclopentolate and being feasible for a mass screening approach.[15],[16],[17],[18],[19],[22]

There are a just a handful of studies, which have compared the MNR with cycloplegic refraction, with most performing correlation analyses while only one looked at agreement.[7],[8],[21],[22],[23],[24] Although they have reported high and significant correlations of 85%–97%, this is hardly surprising since both methods are measuring the same refractive state, and thus are bound to correlate. We sought to answer the far more appropriate question: How well do the measurements of ocular refraction values agree between the two techniques, apart from looking at correlation as a prelude to regression, to explore the possibility of predicting the postcycloplegic refraction (PCRef) with MNR. We aimed to study agreement, correlation, and linear regression between Mohindra's near retinoscopic refraction and cycloplegic-post cycloplegic technique of refraction under cyclopentolate.

  Methods Top

After clearance from the institutional review board and obtaining informed consent from the parents, we recruited 101 cooperative children between 1 and 12 years of age, form the eye OPD. We excluded children with previous ocular surgeries, past or present ocular inflammation, poor media clarity, and neurological disease. After basic ophthalmic workup, each subject underwent objective refraction, which included Mohindra's near retinoscopy (MNR), followed by cycloplegic refraction with 1% topical cyclopentolate (CRef) on the same visit, with a PCRef at least 72 h later. The spherical and cylindrical values of neutralizing lenses were masked while performing refraction.

1. Mohindra's near retinoscopy technique

MNR included dry retinoscopy performed in a dimly lit room so that there should not be any accommodative target visible to the children. Self-illuminated Heine Beta 200 Streak Retinoscope was used to perform retinoscopy for each eye at 50 cm, by an experienced refractionist (>30 years experience). To ensure the correct distance, a nonstretchable cord of 50 cm was tied to the handle of streak retinoscope [Figure 1]. After obtaining the dioptric value of the neutralizing lenses, 1.25 D was subtracted to obtain the refraction. The average illumination of the room while performing MNR was 6–10 lux measured by digital Lux meter (Brand: ANMOL, Model: LX-1010B).[22],[25]
Figure 1: Mohindra's Near Retinoscopy being performed in a dim lit room

Click here to view

2. Procedure for cycloplegic and postcycloplegic refraction

Each subject underwent wet retinoscopy using topical cyclopentolate. Two drops of 1% cyclopentolate was instilled in the conjunctival cul-de-sac of each eye, 10 min apart. Punctal pressure was applied for two min each time the drops were instilled. Following 30 min of the last drop, retinoscopy was performed from 50 cm, the usual distance at which our senior refractionist performs retinoscopy. Adequate cycloplegia was confirmed by observing nonvarying retinoscopic reflexes, when the child attempted to look at different objects. Diopteric equivalents for distance (2.0 Diopters for 50 cm) and for cycloplegia (0.5 Diopters for cyclopentolate) were subtracted from the values of spherical neutralizing lenses to obtain ocular refraction.[14]

Children were asked to return after 72 h for CRef, when PCRef was performed with subjects encouraged to fixate at the vision charts at six meters.

For statistical analysis, we considered the average values of right and left eye for each measured entity.

Statistical analysis

We used JASP Version[26] and MedCalc Trial Version for statistical analysis.[27] Paired t-test was used to compare means of refractive data. We described our data using mean (SD). 95% confidence intervals (CI) are quoted wherever possible. P value of ≤0.05 was considered as statistically significant.

  Results Top

We included 101 children (202 eyes). The baseline demographic data, and refractive data are depicted in [Table 1]. The mean ± SD in logarithm of the minimum angle of resolution (logMAR) for average uncorrected visual acuity (UCVA) was 0.26 ± 0.20 and best corrected visual acuity (BCVA) was 0.1 ± 0.01. Three subjects were from preverbal age group (1.5 years), and therefore we were unable to obtain their visual acuity (VA) data. The mean ± SD spherical equivalent with MNR was 1.71 ± 2.49 D and with PCRef was 1.43 ± 2.42 D: implying that on average, it was more with MNR by a mere 0.28 D, although it was statistically significant (Paired t-test: P value <0.001; 95% CI of 0.16 to 0.40 D).
Table 1: Gender wise distribution of baseline demographic data

Click here to view

We split the children into two groups of 1-6 years and 7-12 years with 36 children in the younger group and 65 children in the older group. For children from younger group, the SE was more by 0.26 ± 0.75 D on MNR compared to PCRef (P value = 0.05; 95% CI-0.03 to 0.5). Likewise, for children from older group, the SE was more by 0.29 ± 0.49 D on MNR compared to PCRef (P value <0.001; 95% CI 0.18 43 to 0.42). On comparing the mean SE of the two groups the younger group measured more than the older group by a non-significant difference of 0.01 ± 0.8 D (Paired t-test: P value = 0.93; 95% CI of 0.26 to 0.28). Expectedly a high and significant positive correlation was demonstrated between the SE on MNR and PCRef (r = 0.97; 95% CI of 0.96–1.0 and P value <0.001).{Table 1}

Linear regression analysis (LRA) was done to predict spherical equivalent (SE) of PCRef on the basis of the values of the SE obtained by MNR technique. The analysis showed that 94% of the values of PCRef can be predicted on the basis of MNR (r2 change = 0.96). The regression formula thus obtained was: [Figure 2]a:
Figure 2: (a) Linear Regression analysis for average spherical equivalent values on MNR and PCRef. (b) Linear Regression analysis for average spherical equivalent values on MNR and PCRef excluding children with ≥ ±5.0 D of refractive error

Click here to view

SE of PCRef = -0.188 + (0.945 × SE by MNR)

Considering that most glass prescription are usually of a small power; we re-evaluated the 91 subjects with refractive error within ±4.0 D. The regression formula obtained for those 91 children was [Figure 2]b:

SE of PCRef = -0.05 + (0.82 × SE by MNR)

Bland–Altman plot [Figure 3] used for agreement analysis, showed MNR measures on average a mere 0.3 D more hypermetropia, which in terms of myopia means 0.3 D less myopia, than that measured by PCRef.
Figure 3: Bland–Altman plot for analyzing agreement between MNR and PCRef

Click here to view

  Discussion Top

We included 101 children of 1–12 years of age. We found a high and significant correlation between SE on MNR and PCRef: (r = 0.97, P < 0.001). Although, the mean difference between SE on MNR and PCRef was statistically significant, it was clinically unimportant: 0.28 ± 0.12 D; P < 0.001. Linear regression analysis (LRA) suggested that 94% of the variation of the SE on PCRef could be explained on the basis of the SE on MNR [Figure 2]a. Agreement analysis suggested that MNR overestimates hypermetropia and underestimates myopia each by on average 0.3 D (Bland–Altman plot- Fig. 3) than the standard procedure of CRef-PCRef.

Indra Mohindra in 1977 came out with her novel technique of near dry retinoscopy, which was later popularized as Mohindra's near retinoscopy. After exhaustive online and offline search, we found only the short abstracts of her first two studies (1977).[7],[8]

In 1979, Mohindra and her colleague Molinari, compared the technique of Mohindra's near retinoscopy, with that of cycloplegic refraction using a combination of 1% tropicamide and 10% phenylephrine eye drops: the two techniques were each performed by two optometrists with varying experience. They enrolled 31 primary grade children of 6–12 years of age with mean age = 6.26 years. The SE error by the two techniques performed by the experienced optometrist-I (extensive experience with near retinoscopy technique) and nonexperienced optometrist-II (no experience with near retinoscopy technique) demonstrated high correlation (r = 0.97; P value = 0.001); in addition, the SE by the two different techniques by the same optometrist also demonstrated high correlation. This supported Mohindra's hypothesis that her technique can stabilize the accommodation as adequately as 1% tropicamide and 10% phenylephrine does.[21] High correlation should not surprise anyone, since the two optometrists or the two techniques were essentially measuring the same parameter: higher values by one person or technique are likely to be measured higher by the other person or technique: we believe agreement should be the outcome of interest.

Borghi and Rouse, in 1985, compared Mohindra's near retinoscopy with cycloplegic retinoscopy using 1% cyclopentolate eye drop in 22 children of 3.6–10 years old. 85% correlation was found between the two techniques. Results suggested that this noncycloplegic technique measured more hypermetropia by 0.5 to 0.75 D, (Student's t-test, P < 0.05 (sic)).[23]

Chen et al., in 2011, performed a comparative study between cycloplegic and noncycloplegic refraction in Chinese neonates. Cycloplegic retinoscopy, with the combination of 0.5% cyclopentolate and 0.25% phenylephrine eye drops, was done on 81 neonates randomly selected from 185 neonates who had undergone noncycloplegic refraction (nC group). On an average the cycloplegic measurement was more by a mean of +3.0 D: +3.58 D Vs +0.50 D. This study concluded that the Mohindra's technique of non cycloplegic refraction was not sufficiently uncovering the hypermetropia, and thus should be considered inappropriate to assess refractive error in neonates.[24] The authors also elaborate that the accommodation in Mohindra's near retinoscopy may vary on the time duration when it is performed, suggesting that in the initial period children do accommodate more, but over two min tend to relax their accommodation. This may explain some of the variation in the different studies. We do not as a routine see or refract such small children, since all the subjects in this study were less than a week old. Therefore, although this study does suggest higher differences, this is quite not representative of the population that most of us are screening in our hospitals in India.

In this era, where most ophthalmologists strongly recommend cycloplegic refraction to unmask hypermetropia completely, Mohindra's retinoscopy technique has not gained much popularity. Hence, studies using noncycloplegic techniques have been far and few.

Recently, Natarajan et al. (2016) did a prospective observational study, very similar to our study, including 500 eyes of 250 patients from 30 to 60 years of age. They have found 99.1% correlation between Mohindra's retinoscopy and static refraction and 98.3% correlation between monocular and binocular subjective refraction. Agreement analysis using Bland–Altman plot showed that on an average Mohindra's near retinoscopy measured 0.39 (1.96) D more hypermetropia in comparison to the static refraction. They proposed Mohindra's near retinoscopy as a quick and effective measure of objective refraction in mass screening scenario taking consideration of time, space, and money.[22]


In our study the retinoscopy by the two methods was performed by the same refractionist, and therefore there could be a chance of incorporation of examiner bias. Although we masked the data of the MNR from the refractionist, it would have been better to have a different refractionist perform the two techniques. Arguably, on the contrary, that may have introduced interobserver differences. It might have been informative to compare the cylindrical errors, but we avoided the tedious vector analysis that would have been necessary.

  Conclusion Top

We are of the opinion that MNR measurement agrees well with the practice of cycloplegic post cycloplegic refraction. For the mathematically inclined, we have developed a regression equation to predict the standard PCRef spherical equivalents on the basis of that measured by MNR. This is especially true for the children from toddler to early school years. Especially for those within ±4.0 D ametropia [Figure 2]b, this could simply be reduced to: SE on PCRef is 80% of the MNR, rounding off 0.82 to 0.80 D and dropping off the 0.05 subtraction, since it tends to zero, being 1/20th of a diopter.

With the inherent advantage of being quick, drugless, one-visit procedure, and measuring up well with the standard technique, the MNR may be resorted to more often in a clinical setting for routine assessment of children of ages 1–12 years old. In future studies, it would be important to compare the MNR with cycloplegic refraction in those with anisometropia, strabismus, and amblyopes and assess how well it performs, for a more universal adoption. For the moment then, it offers a viable option in routine screening of children.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

Pascolini D, Mariotti SP. Global estimates of visual impairment: 2010. Br J Ophthalmol 2012;96:614-8.  Back to cited text no. 1
Resnikoff S, Pascolini D, Mariotti SP, Pokharel GP. Global magnitude of visual impairment caused by uncorrected refractive errors in 2004. Bull World Health Organ 2008;86:63-70.  Back to cited text no. 2
Padhye AS, Khandekar R, Dharmadhikari S, Dole K, Gogate P, Deshpande M. Prevalence of uncorrected refractive error and other eye problems among urban and rural school children. Middle East Afr J Ophthalmol 2009;16:69-74.  Back to cited text no. 3
[PUBMED]  [Full text]  
Dandona R, Dandona L. Refractive error blindness. Bull World Health Organ Int J Public Health 2001;793:237-43.  Back to cited text no. 4
VISION 2020 e-resource [Internet]. Available from: http://v2020eresource.org/home/newsle tter/SM103. [Last cited on 2017 Aug 28].  Back to cited text no. 5
Al-Tamimi ER, Shakeel A, Yassin SA, Ali SI, Khan UA. A clinic-based study of refractive errors, strabismus, and amblyopia in pediatric age-group. J Fam Community Med 2015;22:158-62.  Back to cited text no. 6
[PUBMED]  [Full text]  
Mohindra I. A non-cycloplegic refraction technique for infants and young children. J Am Optom Assoc 1977;48:518-23.  Back to cited text no. 7
Mohindra I. Comparison of 'near retinoscopy' and subjective refraction in adults. Am J Optom Physiol Opt 1977;54:319-22.  Back to cited text no. 8
Sergienko NM, Nikonenko DP. Measurement of amplitude of accommodation in young persons. Clin Exp Optom 2015;98:359-61.  Back to cited text no. 9
Duane A. Studies in monocular and binocular accommodation, with their clinical application. Trans Am Ophthalmol Soc 1922;20:132-57.  Back to cited text no. 10
Manny RE, Hussein M, Scheiman M, Kurtz D, Niemann K, Zinzer K. Tropicamide (1%): An effective cycloplegic agent for myopic children. Invest Ophthalmol Vis Sci 2001;42:1728-35.  Back to cited text no. 11
Hofmeister EM, Kaupp SE, Schallhorn SC. Comparison of tropicamide and cyclopentolate for cycloplegic refractions in myopic adult refractive surgery patients. J Cataract Refract Surg 2005;31:694-700.  Back to cited text no. 12
van Minderhout HM, Joosse MV, Grootendorst DC, Schalij-Delfos NE. A randomized clinical trial using atropine, cyclopentolate, and tropicamide to compare refractive outcome in hypermetropic children with a dark iris; skin pigmentation and crying as significant factors for hypermetropic outcome. Strabismus 2019;27:127-38.  Back to cited text no. 13
Duke-Elder S, Abrams D. Duke-Elders Practice Refraction. Tenth ed. Elsevier; 1993.p. 168-70.  Back to cited text no. 14
Wygnanski-Jaffe T, Nucci P, Goldchmit M, Mezer E. Epileptic seizures induced by cycloplegic eye drops. Cutan Ocul Toxicol 2014;33:103-8.  Back to cited text no. 15
Princelle A, Hue V, Pruvost I, Potey C, Dubos F, Martinot A. [Systemic adverse effects of topical ocular instillation of atropine in two children]. Arch Pediatr Organe Off Soc Francaise Pediatr 2013;20:391-4.  Back to cited text no. 16
Fan DS, Rao SK, Ng JS, Yu CB, Lam DS. Comparative study on the safety and efficacy of different cycloplegic agents in children with darkly pigmented irides. Clin Exp Ophthalmol 2004;32:462-7.  Back to cited text no. 17
Mirshahi A, Kohnen T. Acute psychotic reaction caused by topical cyclopentolate use for cycloplegic refraction before refractive surgery: Case report and review of the literature. J Cataract Refract Surg 2003;29:1026-30.  Back to cited text no. 18
Pooniya V, Pandey N. Systemic toxicity of topical cyclopentolate eyedrops in a child. Eye 2012;26:1391-2.  Back to cited text no. 19
Owens DA, Mohindra I, Held R. The effectiveness of a retinoscope beam as an accommodative stimulus. Invest Ophthalmol Vis Sci 1980;19:942-9.  Back to cited text no. 20
Mohindra I, Molinari JF. Near retinoscopy and cycloplegic retinoscopy in early primary grade schoolchildren. Am J Optom Physiol Opt 1979;56:34-8.  Back to cited text no. 21
Natarajan M, Hemlatha B, Santosh P, Sowbhagya HN. A novel method to screen and correct refractive errors in a camp scenario. Eur J Biomed Pharm Sci 2016;3:525-37.  Back to cited text no. 22
Borghi RA, Rouse MW. Comparison of refraction obtained by 'near retinoscopy' and retinoscopy under cycloplegia. Am J Optom Physiol Opt 1985;62:169-72.  Back to cited text no. 23
Chen J, Chen Z, Lin S, Zhang J, Wang Q, Zhong H, et al. Correlation analysis for school-age children's height and refractive errors. Adv Clin Exp Med Off Organ Wroclaw Med Univ 2018;27:1125-30.  Back to cited text no. 24
India Import data of Anmol [Internet]. Available from: https://www.eximpulse.com/import-produ ct-Anmol.htm. [Last cited on 2020 Jan 18].  Back to cited text no. 25
Download JASP [Internet]. JASP - Free and User-Friendly Statistical Software. Available from: https://jasp-stats.org/download/. [Last cited on 2020 Jan 18].  Back to cited text no. 26
Schoonjans F. MedCalc statistical software [Internet]. MedCalc. Available from: https://www.medcalc.org/. [Last cited on 2020 Jan 18].  Back to cited text no. 27


  [Figure 1], [Figure 2], [Figure 3]

  [Table 1]


    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Access Statistics
    Email Alert *
    Add to My List *
* Registration required (free)  

  In this article
Article Figures
Article Tables

 Article Access Statistics
    PDF Downloaded263    
    Comments [Add]    

Recommend this journal