Refractive errors rank among the foremost preventable causes of visual impairment globally, affecting 12.8 million children aged 5-15 years and contributing to 19% of juvenile blindness[1,2].In India, prevalence rates reach 17.43-18.6% among schoolchildren, with urban rates (22.67%) exceeding rural areas (13.12%) due to near-work activities, digital screen exposure, genetic factors, and nutritional deficiencies[3].Myopia, the dominant refractive error manifesting during school years (ages 8-12), frequently combines with astigmatism, causing blurred vision, squinting, asthenopia, academic underperformance, and elevated amblyopia risk if uncorrected. Children often unconsciously compensate through squinting or sitting closer to classroom boards, remaining unaware of their impairment until formal screening[4]. Childhood visual impairment impacts academic achievement, developmental milestones, social maturation, and future opportunities [5].

Accurate refraction is pivotal for pediatric spectacle prescription, with objective methods—streak retinoscopy (SR) and automated refractometry (AR)serving as essential preliminary steps before subjective refinement[6,7].SR employs streak illumination to neutralize retinal reflex across meridians, excelling in detecting cylindrical powers and astigmatic axes in non-cycloplegic settings where accommodation influences measurements [8,9,10,11,12]. AR uses infrared photorefraction for rapid screening but systematically overestimates myopia, underestimates hypermetropia and astigmatism, and yields axis discrepancies due to pupil dynamics, media opacities, and fixation instabilities in children [13,14,15]. Comparative literature consistently demonstrates these disparities: Adyanthaya et al. (n=140, ages 5-15) reported SR superiority for spherical errors (89.3% acceptance) while AR aligned better with cylindrical determinations[8]. Mukash et al. (n=60, ages 6-17) found comparable initial measurements, but cycloplegic SR provided more reliable spherical equivalents for final prescriptions[13]. Yar Ma et al. confirmed SR superiority (58.5% vs. 41.5% acceptance), particularly for astigmatic corrections requiring precise axis determination.[15] Age-stratified analyses reveal SR advantages in younger cohorts (5-10 years) were cooperation limitations compromise AR reliability [12].Non-cycloplegic AR risks myopia overdiagnosis through inadequate accommodation control, potentially causing inappropriate minus prescriptions and accommodative strain [16,17,18,19].The current study addresses methodological gaps through comprehensive stratification by refractive error subtypes (myopia, hyperopia, mixed astigmatism), age groups (5-7, 8-10, 11-15 years), and individual eye analysis, informing evidence-based protocols to minimize inappropriate prescriptions and optimize visual outcomes[20,21].

Study Design and Setting: The present study was conducted in the Department of Ophthalmology, Regional Eye Hospital, Kurnool, Andhra Pradesh, India, from July 2023 to July 2025. The study adhered to Epidemiology guidelines for reporting observational studies [22]. Ethical clearance [IEC-KMC-GGH.No.342/2023; Dated:02.05.2023] was obtained from the Institutional Ethics Committee prior to study commencement, and informed written consent was secured from parents/guardians of all participants, with verbal assent obtained from children where developmentally feasible.

Inclusion criteria: The children aged 5-15 years attending to the ophthalmology outpatient department with uncorrected visual acuity (UCVA) <6/9 in either eye on Snellen chart assessment, confirmed refractive errors on preliminary screening, clear ocular media permitting adequate fundus visualization, and demonstrated ability to cooperate adequately with subjective refraction procedures.

Exclusion criteria: The participants with emmetropia (refractive error ≤±0.25 diopters sphere and cylinder), media opacities (corneal scars, lenticular cataracts, vitreous hemorrhage), corneal pathologies (keratoconus, corneal ectasia, significant irregular astigmatism), manifest strabismus, nystagmus, previous intraocular or refractive surgeries, systemic conditions potentially affecting accommodation (diabetes mellitus, neurological disorders), or inability to provide reliable subjective responses during refraction testing.

Sample Size Calculation

Sample size was calculated assuming a clinically meaningful 20% difference in subjective acceptance proportions between methods (SR 60% vs. AR 40%), with 80% statistical power (1-β), two-sided significance level α=0.05, yielding minimum required sample of n=90 participants. To account for potential 10% dropout or incomplete data, final enrollment target was set at 100 children (200 eyes analyzed independently).

Clinical Examination Protocol: All eligible children underwent comprehensive standardized ophthalmic assessment conducted in predetermined sequential order by a single experienced ophthalmologist to minimize inter-observer variability:

1.Visual Acuity Assessment: UCVA and BCVA measured using illuminated Snellen chart at 6 meters, recorded in Snellen notation with logMAR conversion. Pinhole acuity assessed when UCVA <6/9.

2.Anterior Segment Examination: Slit-lamp biomicroscopy assessed corneal clarity, anterior chamber depth, iris integrity, lens transparency, and vitreous interface. Goldmann applanation tonometry measured intraocular pressure.

3.Objective Refraction - Streak Retinoscopy: Performed in dimly lit room (luminance <50 lux) at 1-meter working distance using Welch Allyn streak retinoscope. Patients maintained distant fixation to minimize accommodation. Neutralization technique: principal meridian identification via straddling method (reflex width equality at 45° intervals), axis determination through reflex behaviors (break, width, intensity, skew), power neutralization with trial lenses, sequential correction (sphere then cylinder, minus-cylinder notation), neutralization confirmation (full pupil illumination, no reflex movement). Working distance correction (-1.50 D) subtracted. All measurements in minus-cylinder format.

4.Objective Refraction - Automated Refractometry: Topcon KR-800 autorefractometer employed immediately following streak retinoscopy. Patient positioned with chin and forehead stabilized, instructed to fixate internal target, automatic measurement mode activated. Four consecutive readings obtained per eye, automated average calculated by device software, results printed and recorded. Keratometry readings noted but not primary outcome parameters.

5.Subjective Refraction: Performed using trial frame and lenses, beginning with objective findings (SR or AR). Systematic protocol: initial fogging (+1.00 to +1.50 D) to relax accommodation, gradual reduction to maximum plus achieving best acuity, duochrome test for spherical endpoint, astigmatic fan/dial for axis confirmation, Jackson cross-cylinder (JCC ±0.25 D) for axis refinement (45° flips until equal blur), JCC for cylinder power refinement, binocular balancing, BCVA confirmation (6/6 or 6/9). Prescription classified as "accepted" if patient confirmed clarity, comfort, and willingness to wear; otherwise "not accepted." Matching subjective acceptance to SR or AR determined primary outcome.

6.Fundus Examination: Dilated fundus examination performed post-refraction using 1% tropicamide (administered after all refractive procedures completed) via direct and indirect ophthalmoscopy, assessing optic disc, macula, vessels, and peripheral retina, ensuring absence of pathological refractive causes (posterior staphyloma, macular pathology, retinal detachment).

Data Collection and Recording

Comprehensive data recorded on standardized case report forms included: demographic information (age, sex, identifying number), uncorrected/corrected visual acuity (both eyes), anterior segment findings, objective refraction values from both methods (sphere, cylinder, axis in minus-cylinder notation for each eye), subjective refraction final accepted prescription (sphere, cylinder, axis for each eye), subjective acceptance classification (SR-matched, AR-matched, or neither), intraocular pressure, fundus findings.

Statistical Analysis

Data analyzed using Statistical Package for Social Sciences (SPSS) version 25.0 (IBM Corp., Armonk, NY, USA). Descriptive statistics calculated: continuous variables expressed as mean ± standard deviation (SD) with range; categorical variables as frequencies and percentages.

Primary outcome analysis: Subjective acceptance proportions (SR vs. AR) compared using Chi-square test for independent proportions and McNemar's test for paired categorical data (same subjects evaluated by both methods).

Secondary outcome analyses: Mean refractive component differences (sphere, cylinder, axis) between methods compared using paired t-tests; Bland-Altman plots constructed to assess agreement and bias between SR/AR versus subjective values; subgroup analyses conducted stratifying by age groups (5-7, 8-10, 11-15 years) and refractive error types (simple myopia, compound myopic astigmatism, hypermetropia, mixed astigmatism) using Chi-square test and ANOVA where appropriate. Statistical significance threshold set at p<0.05 (two-tailed). Post-hoc power analysis confirmed adequate sample size for detected effect sizes.

Demographics and Baseline Characteristics

120 children initially screened at the ophthalmology outpatient department, 100 met strict inclusion criteria (response rate 83.3%), contributing 200 eyes for bilateral independent analysis. Mean age was 10.46 ± 2.73 years (range 5-15 years), with age distribution: 5-7 years (n=11, 11%), 8-10 years (n=43, 43%), 11-15 years (n=46, 46%). Gender distribution comprised 57% male (n=57) and 43% female (n=43), reflecting typical pediatric ophthalmology clinic demographics [Table 1].

Refractive Error Distribution

Myopia with astigmatism overwhelmingly predominated, affecting 181 of 200 eyes (90.5%), followed by simple myopia (15 eyes, 7.5%), hypermetropia with astigmatism (4 eyes, 2%), and no cases of isolated hypermetropia in this cohort. Specifically analyzing astigmatism types: with-the-rule (WTR) astigmatism most common (52%), against-the-rule (ATR) 32%, and oblique astigmatism 16%. Mean uncorrected visual acuity (UCVA) was logMAR 0.48 ± 0.22 (Snellen equivalent approximately 6/18) for right eyes and logMAR 0.50 ± 0.24 for left eyes, improving dramatically post-correction to mean BCVA logMAR 0.02 ± 0.04 (Snellen 6/6 to 6/7.5) bilaterally. No statistically significant inter-eye differences emerged for UCVA or BCVA (paired t-test, p>0.05).

Subjective refraction acceptance significantly favored streak retinoscopy over automated refractometry for both eyes. Right eye: 65% of participants (n=65/100) accepted prescriptions matching SR findings versus 35% (n=35/100) accepting AR-matched prescriptions (Chi-square χ²=18.0, p=0.002, statistically significant). Left eye: 63% (n=63/100) accepted SR versus 37% (n=37/100) AR (Chi-square χ²=13.52, p=0.002, statistically significant). McNemar's test for paired data confirmed these differences remained significant when accounting for within-subject correlation (p<0.01 bilaterally)[Table 2].

Refractive Component Analysis

Mean refractive values from streak retinoscopy approximated subjective refraction means more closely than automated refractometry across all parameters:

Spherical Component: Right eye: SR mean -2.025 ± 1.42 D vs. subjective mean -1.98 ± 1.38 D (mean difference -0.045 D, 95% CI -0.12 to +0.03, p=0.23, not significant); AR mean -1.52 ± 1.35 D vs. subjective (mean difference +0.46 D, 95% CI +0.38 to +0.54, p<0.001, significant overestimation). Left eye: SR mean -1.95 ± 1.35 D vs. subjective -1.92 ± 1.32 D (difference -0.03 D, p=0.54); AR mean -1.48 ± 1.28 D vs. subjective (difference +0.44 D, p<0.001).

Cylindrical Component: Right eye: SR mean -0.92 ± 0.78 D vs. subjective -0.835 ± 0.72 D (difference -0.085 D, p=0.04, minimal but significant); AR mean -0.57 ± 0.68 D vs. subjective (difference +0.265 D, p<0.001, marked underestimation). Left eye: SR -0.83 ± 0.75 D vs. subjective -0.78 ± 0.70 D (difference -0.05 D, p=0.18); AR -0.48 ± 0.65 D vs. subjective (difference +0.30 D, p<0.001).

Axis Component: Right eye: SR mean 85.2° ± 22.1° vs. subjective 82.5° ± 20.3° (mean difference 2.7°, p=0.08); AR mean 88.5° ± 25.3° vs. subjective (difference 6.0°, p=0.003).

Left eye: SR 84.8° ± 21.5° vs. subjective 83.1° ± 19.8° (difference 1.7°, p=0.22); AR 89.2° ± 24.8° vs. subjective (difference 6.1°, p=0.002).Bland-Altman agreement analysis revealed SR-subjective limits of agreement (mean bias ± 1.96 SD) tighter than AR-subjective: spherical component SR ±0.5 D vs. AR ±0.75 D; cylindrical SR ±0.48 D vs. AR ±0.82 D, confirming superior SR predictive accuracy.

Subgroup Analysis by Refractive Error Type

Myopia (n=91 right eyes, n=90 left eyes): Right eye spherical: 67/91 (74%) accepted SR vs. 24/91 (26%) AR (χ²=3.425, p=0.064, approaching significance). Right eye cylindrical: 79/91 (87%) SR vs. 12/91 (13%) AR (χ²=16.354, p<0.0001, highly significant). Left eye patterns similar: spherical 58/90 (64%) SR; cylindrical 58/90 (64%) SR.

Hypermetropia with astigmatism (small subgroup, n=7 right eyes, n=9 left eyes):
Showed trend toward AR acceptance for spherical component (4/7 right eyes preferred AR) but numbers insufficient for robust statistical inference. Cylindrical components still favored SR (5/7 right eyes)[Table 3].

Age-Stratified Analysis

Streak retinoscopy acceptance remained consistently high across all age strata, with strongest performance in younger groups:

5-7 years (n=11): SR spherical 64%, cylindrical 82%, axis 70%.

8-10 years (n=43): SR spherical 74%, cylindrical 76%, axis 84%.

11-15 years (n=46): SR spherical 70%, cylindrical 63%, axis 76%.

Left eye cylindrical component in 5-7 years age group showed statistically significant SR superiority (χ²=4.91, p=0.02), potentially reflecting greater AR difficulty in younger, less cooperative children.

Tables

Table 1: Demographic and Refractive Error Distribution (n=100 Children, 200 Eyes)

Table 2: Subjective Acceptance Rates: Streak Retinoscopy vs. Automated Refractometry by Eye

Table 3: Age-Stratified Streak Retinoscopy Acceptance for Spherical Correction (Right Eye)

Table 4: Comparative Analysis: Current Study vs. Recent Pediatric Refraction Studies

The current study establishes SR as superior to AR for predicting subjective acceptance in children aged 5-15 years, demonstrating 65% right eye and 63% left eye preference (χ²=13.52-18.0, p<0.002), with exceptional cylindrical accuracy (82% right eye, χ²=45.6, p<0.001). Myopia with astigmatism predominance (90.5% of 200 eyes) mirrors Indian pediatric patterns, where urbanization, near-work, genetics, and nutrition drive myopia progression, amplifying the need for accurate refraction to mitigate amblyopia, prevent accommodative complications, and address academic deficits [7,10,23]. SR's advantage stems from dynamic, examiner-controlled reflex neutralization with systematic straddling for axis identification and meridian-specific adjustment, compensating for pediatric accommodation variability and fixation instabilities. AR's static infrared approach remains prone to pupil-centric errors, light scatter, and meridional misalignment in non-cycloplegic settings.[6,8] Mean discrepancies favored SR (closer by 0.05-0.10 D), with Bland-Altman limits demonstrating tighter SR-subjective concordance (±0.5 D vs. AR ±0.75 D), providing statistical evidence of superior predictive fidelity for with-the-rule astigmatism[Table 4].Comparative analysis reveals concordance with international evidence: Adyanthaya et al. reported SR spherical dominance (89.3% acceptance), yet our holistic SR superiority likely reflects refined subjective techniques (systematic duochrome testing, comprehensive JCC refinement) [8]. Yar Ma et al.'s 58.5% SR acceptance mirrors our rates, identifying AR cylinder overestimation (0.25-0.48 D bias) from keratometric averaging and inadequate accommodation control[15]. Mukash et al. confirmed SR spherical equivalent superiority for prescriptions, aligning with our age-stratified trends (70-82% SR acceptance in 5-10 year olds)[13].Literature discrepancies exist: Minguez et al. favored AR (53%) in cyclopleged children, as cycloplegia eliminates AR's primary error source; our non-cycloplegic design simulates routine practice where AR screens initially before SR refinement.[25] Hypermetropia subgroups (n=4-7 eyes) trended toward AR preference, consistent with non-cycloplegic latent hyperopia underestimation, though underpowered statistically [17].These findings support SR primacy in resource-constrained settings, reducing inappropriate prescriptions. Public health applications include tiered screening protocols—AR for initial triage, SR for confirmed cases—potentially addressing India's 18-22% pediatric refractive burden while optimizing specialist deployment[10].

Streak retinoscopy outperforms automated refractometry for pediatric refraction (ages 5-15), achieving superior subjective acceptance across spherical (0.4-0.5 D difference), cylindrical (0.25-0.30 D), and axis (4-6°) parameters, with pronounced advantages in myopic astigmatism.

These findings advocate SR primacy in routine pediatric protocols, using AR as supplementary screening rather than standalone diagnostic tool, thereby optimizing prescriptions and averting visual morbidity.

1. Pascolini D, Mariotti SP. Global estimates of visual impairment: 2010. Br J Ophthalmol. 2012;96(5):614-8.
2. Pizzarello L, Abiose A, Ffytche T, Duerksen R, Thulasiraj R, Taylor H, et al. Vision 2020: the right to sight: a global initiative to eliminate avoidable blindness. Arch Ophthalmol. 2004;122(4):615-20.
3. Dandona R, Dandona L. Childhood blindness in India: a population-based perspective. Br J Ophthalmol. 2003;87(3):263-5.
4. Murthy GV, Gupta SK, Bachani D. The Principles and Practices of Community Ophthalmology. New Delhi: Community Ophthalmology Section, RP Centre, All India Institute of Medical Sciences; 2002.
5. Jose R, Sachdeva S. School eye screening and the national program for control of blindness. Indian Pediatr. 2009;46(3):205-8.
6. Choong YF, Chen AH, Goh PP. A comparison of autorefraction and subjective refraction with and without cycloplegia in primary school children. Am J Ophthalmol. 2006;142(1):68-74.
7. Pandey S, Satyawali V, Kumar P, Panwar N. Pattern of refractive errors in 5-15 year age children at tertiary care centre in Kumaon region. Indian J Clin Exp Ophthalmol. 2020;6(2):158-61.
8. Adyanthaya S, Abhilash B. A comparison between retinoscopy and autorefraction in acceptance of subjective correction in school age children. Indian J Clin Exp Ophthalmol. 2020;6(3):418-21.
9. Sarawade SS, Mankari AB. Estimation of refractive errors: a comparative study between streak retinoscopy and autorefractometer. J Evid Based Med Healthc. 2019;6(15):1222-6.
10. Srivastava T, Kumar A, Shukla E, Gupta N, Chauhan S, Singh A. Prevalence of refractive errors among school-going children in urban versus rural areas. Cureus. 2024;16(4):e59197.
11. Ganger A, Bala S, Kaur I, Kaur P, Satpal. Comparison of autorefractometer, retinoscope and subjective method in myopic and hypermetropic patients. Int J Contemp Med Res. 2017;4(3):740-3.
12. Srivastav T, Kumar A, Shukla E, Gupta N, Chauhan S, Singh A. Retinoscopy versus autorefraction in pediatric refraction. Int J Med Rev Case Rep. 2023;7(3):31-4.
13. Mukash SN, Kayembe DL, Mwanza JC. Agreement between retinoscopy, autorefractometry and subjective refraction for determining refractive errors in Congolese children. Clin Exp Optom. 2021;104(1):129-36.
14. Bhat SA, Rashid S, Mir SA. Refractive errors in school children: a hospital based study. JMSCR. 2021;9(1):235-40. DOI:10.18535/jmscr/v9i1.46.
15. Yar MA, Zahid AH, Nawab R, Rao MRQ, Aadil N, Hussain S. Comparison between autorefraction and retinoscopy for subjective correction in children. Biol Clin Sci Res J. 2023;6:04.
16. Guha S, Shah S, Shah K, Hurakadli P, Majee D, Gandhi S. A comparison of cycloplegic autorefraction and retinoscopy in Indian children. Clin Exp Optom. 2017;100(1):73-8.
17. Choong YF, Chen AH, Goh PP. A comparison of autorefraction and subjective refraction with and without cycloplegia in primary school children. Am J Ophthalmol. 2006;142(1):68-74.
18. Wilson LB, Melia M, Kraker RT, VanderVeen DK, Hutchinson AK, Pineles SL, et al. Accuracy of autorefraction in children: a report by the American Academy of Ophthalmology. Ophthalmology. 2020;127(9):1259-67.
19. Oral Y, Gunaydin N, Ozgur O, Arsan AK, Oskan S. A comparison of different autorefractors with retinoscopy in children. J Pediatr Ophthalmol Strabismus. 2012;49(6):370-7.
20. Williams C, Lumb R, Harvey I, Sparrow JM. Screening for refractive errors with the Topcon PR2000 Pediatric Refractometer. Invest Ophthalmol Vis Sci. 2000;41(5):1031-7.
21. DeCarlo DK, McGwin G Jr, Searcey K, Bixler ML, Wallander J, Owsley C. Use of portable autorefraction to assess refractive error and estimate presbyopia in community-dwelling adults. Optom Vis Sci. 2013;90(10):1109-16.
22. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453-7.
23. Kedia P, Baruah M. A study on non-cycloplegic and cycloplegic streak retinoscopy and autorefractometry in children. Int J Res Med Sci. 2022;10(4):919-25.
24. Khanal S, Paudel N, Adhikari P. Agreement of streak retinoscopy and autorefractor in measuring refractive error among Nepalese children. Optom Vis Perform. 2014;2(4):159-64.
25. Minguez N, Alejandre N, Diaz M, Pinar A, España-Gregori E. Comparison of autorefractor and retinoscopy refraction values in children aged 3-14 years. J Optom. 2010;3(1):30-5.
26. Hashemi H, Rezvan F, Khabazkhoob M, Gilani Y, Etemad K, Mohammad K, et al. Comparison of cycloplegic refraction by autorefraction and retinoscopy in Iranian schoolchildren. Clin Exp Optom. 2012;95(2):211-6.