Comparison of photoscreeners and hand-held autorefractometer with cycloplegic autorefraction as a vision screening tool for children with newly diagnosed attention decit hyperactivity disorder

Purpose: To compare non-cycloplegic refraction measurements of two photoscreeners and the hand-held autorefractometer with cycloplegic measurements of the autorefractometer in patients with attention decit hyperactivity disorder (ADHD). Methods: This cross-sectional, comparative study consisted of 53 children who were newly diagnosed with ADHD. We compared spherical, cylindrical, cylindrical axis and spherical equivalent (SE) measurements in Plusoptix A12, Spot Vision Screener, and Retinomax K-plus Screen with Tonoref II. Reliability was analyzed by using the interclass correlation coecient (ICC) and Bland-Altman plot was used to evaluate the agreement between devices. Results: The mean age of children was 9.45 ± 1.68. All of the devices measured spherical power and SE signicantly more myopic than the Tonoref II. While The Spot Vision Screener, PlusoptiX A12, and Tonoref II provided similar cylindrical power, Retinomax K-plus Screen measured signicantly lower than the Tonoref II. The excellent reliability was detected in spherical power, cylindrical power, SE and J0 between Tonoref II and PlusoptiX A12 (ICC:0.930, 0.921, 0.927 and 0.920, respectively. All of the hand-held devices showed excellent reliability in terms of cylindrical power and J0 (ICC>0.90, for all) and good reliability for J45 (ICC:0.75-0.90 for all). Conclusion: Despite all devices having advantages or disadvantages, Plusoptix A12 showed excellent reliability for detecting refractive errors in children with ADHD.


Introduction
Attention de cit hyperactivity disorder (ADHD) is one of the most prevalent neurodevelopmental disorder among children and adolescents aged between 6-17 years. The prevalance of ADHD was reported as 2%-18% in different studies. The characteristics of ADHD are inattention, increased hyperactivity, impulsivity, and lack of controlling inappropriate behaviors [1]. To date, several studies have been conducted in children with ADHD in terms of ocular abnormalities such as refractive errors, ocular pathologies, and ocular side effects of agents, methylphenidate, and atomoxetine used for the treatment of ADHD [2,3].
It is known that refractive errors and amblyopia were higher in children with ADHD. Visual problems such as refractive errors may lead to the diminution of visual acuity and concentration and contribute symptoms of ADHD [3]. Thus, it is essential to have an eye examination as a part of complete physical and psychological evaluation in children with ADHD and rule out underlying ocular disorders that may affect children's attention [4]. However, children diagnosed with ADHD may have di culties and limitations in adapting to the measuring devices used for the ophthalmologic examination which starts with refraction assessment due to the symptoms of inattention, mobility, and impulsivity, and therefore inappropriate measurements may be obtained [5,6].
Accurate measurement of the refraction values is crucial in determining the refractive errors, appropriate treatment, and prevention of amblyopia. The cycloplegic refraction, which eliminates accommodation, is the gold standard for detecting refractive errors in children [7]. However, cycloplegic medications can cause unwanted side effects and may be di cult to administer to children each time. Moreover, it takes about 30-45 minutes to reach full cycloplegia, making the process time consuming [8]. Automated refractometers need cycloplegia in children, and table-mounted design can make it challenging to maintain proper position, and providing visual xation on a target for a su cient period may be arduous in uncooperative patients [7]. Although the main measurement methods for detecting refractive errors in children are cycloplegic assessments, both photoscreeners and hand-held autorefractometers are alternatives for vision screening in a short time for physically and mentally disabled, non-cooperative patients such as children with ADHD [7,6].
In this context, the aim of our study is to compare non-cycloplegic refraction measurements of two photoscreeners (Plusoptix A12 and Spot vision screener) and the hand-held autorefractometer (Retinomax K-plus Screen) with cycloplegic measurements of the autorefractometer (Tonoref II).

Study population and study design
We evaluated 61 children at the rst line; however, ve were excluded due to lack of at least one measurement of three devices or having out-of-range measurements, three of total children excluded due to not giving permission for cycloplegia. Finally, this cross-sectional, comparative study consisted of 53 children aged 6-12 years who were newly diagnosed with ADHD in our university Child and Adolescent Psychiatry Department between 01.03.2020 and 01.01.2021. The Local Ethics Committee approved the study design and procedures, which were conducted in accordance with the principles outlined in the Declaration of Helsinki. Written informed consent was obtained from all participants and parents of children.
Attention de cit hyperactivity disorder was diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-V) criteria. Also, all children were evaluated by a child psychiatrist with a semi-structured interview to con rm the diagnosis of ADHD and to determine whether the child had any other psychiatric disorder. After the diagnosis of ADHD, these children were referred to the ophthalmology department for an ophthalmological examination. Children with pure ADHD and drug-naive, and without any ocular pathologies except refractive errors were included in the study. Patients were excluded from the study if they had mental retardation, speci c learning disabilities, and autism spectrum disorders accompanying ADHD that could adversely affect visual examinations. Additionally, children with ptosis, strabismus, corneal and lens opacities, retinal abnormalities, and refractive errors beyond limits measurements according to manufacturers' recommendations were excluded. All of the patients underwent a complete ophthalmic examination, including pupil reactions, Hirschberg's test, cover-uncover, alternate cover tests, slit-lamp biomicroscopy, and dilated fundus examination.

Measurements
The refraction measurements of 106 eyes of 53 children were obtained in the same semi-lit room and the following order, and an average of three consecutive measurements for each device was recorded. Firstly, without cycloplegia, PlusoptiX A12 (Plusoptix GmbH, Nuremberg, Germany), Spot Vision Screener (Welch Allyn, Skaneateles Falls, NY), and Retinomax K-plus Screen (Righton, Tokyo, Japan) were performed respectively. Then, cycloplegia was induced by using cyclopentolate 1% (Sikloplejin; Abdi İbrahim, Istanbul, Turkey) eye drops two times with an interval of 5 minutes. After 45 minutes, pupillary light reactions were checked, and complete cycloplegia was accomplished. The cycloplegic measurements were performed by Tonoref II (Nidek Co. Ltd, Gamagori, Japan). We also evaluated the amblyogenic risk factors according to the American Association for Pediatric Ophthalmology and Strabismus (AAPOS) referral criteria for ages > 48 months, published in 2013 [9].
PlusoptiX A12 (Plusoptix GmbH, Nuremberg, Germany) PlusoptiX A12 is a hand-held photorefractor, particularly for infants, children, and uncooperative patients, to measure refractive errors without cycloplegia. The device provides measurements of both eyes simultaneously within one second from a one-meter distance. The device uses sound and lights for xation target and measures pupil size, interpupillary distance, and ocular alignment. Besides, Plusoptix offers a referral recommendation for ophthalmologic examination based on pre-determined referral criteria [8] . The device has a measurement range from -7.0 to 5.0 D for both spherical and cylindrical measurements.
Spot Vision Screener (Welch Allyn, Skaneateles Falls, NY) The Spot Vision Screener is a hand-held photorefractor that provides measurements of both eyes simultaneously within two seconds from a distance of 1 meter without cycloplegia. The Spot Vision Screener uses random LED visual patterns and audible sound for xation target and provides a report including pupillary diameter, eye alignment, and estimated refraction errors. The Spot Vision Screener also gives a warning for referral a complete eye examination according to the preprogrammed referral criteria [6] . The device can make spherical measurements at a range of -7.50 to 7.50 D and cylindrical measurements at a range of -3.00 and 3.00 D.
Retinomax K-plus Screen (Righton, Tokyo, Japan) The Retinomax K-plus Screen is a portable, hand-held autorefraction keratometer with a 3.5-inch monitor.
The device provides a monocular measurement of refractive errors from a distance of 5 cm. Retinomax is both an autorefractor and a keratometer and uses a fogging mechanism to control accommodation [10].
The device also provides a quick mode in which measurements are taken in one second, and a melody is played to catch children's attention. The constantly changing color screen inside and outside of the device involves children in the examination throughout the process. The device can make spherical measurements at a range of -20.0 D to 23.0 D and cylindrical measurements at a range of 0 D to ±12 D Tonoref II (Nidek Co. Ltd, Gamagori, Japan) Tonoref II is an automatic non-contact tonometer and autorefractometer. The device is mounted on the table, and the patients' forehead must be placed on the forehead of the device to take measurements.
The measurement range of the device is from -30.00 to +25.00 D for spherical and from 0 D to ±12 D for cylindrical measurements.

Study outcomes
Spherical, cylindrical, cylindrical axis, and spherical equivalent (SE) measurements acquired from four devices were evaluated. While calculating SE with [sphere+(cylinder power/2) ] formula, the cylindrical and axis components were transformed into vectorial representations due to the problems related to the analysis of the astigmatic component in the traditional form with the following formulas [11]:

Statistical Analysis
The normality of the data was evaluated with the Kolmogorov-Smirnov test. Age and refractive measurements were presented as mean ± standard deviation. Categorical variables were stated as number (n) and percentage (%). The difference between measurements was assessed by the one-sample t-test. About 95% con dence interval of the difference between the compared methods was also calculated. Bland-Altman plot was used to evaluate the agreement between devices, and 95% limits of agreement (LoA) were determined for spherical equivalent, J0, and J45 values. Reliability was analyzed by using the interclass correlation coe cient (ICC). Based on the 95% con dence interval, reliability was divided into four groups as follows: values less than 0.5 as low, between 0.5-0.75 as moderate, 0.75-0.9 as good, and higher than 0.9 as excellent reliability. Data obtained in the study were analyzed using Statistical Package for the Social Science software version 20 (IBM, SPSS 20 for Windows. Armonk, NY, USA). A value of p<0.05 was considered statistically signi cant.

Results
106 eyes of 53 children who completed all of the measurements were taken into the analysis. The mean age of children was 9.45 ± 1.68 years. Eleven (21%) of the children were female, and 42 (79%) of those were male. We evaluated the measurements in Table 1 and Table 2, which were designed to de ne and compare the spherical power, cylindrical power, axis, spherical equivalent, J0 and J45 values between devices. Also, Bland-Altman plots for comparing devices in terms of spherical equivalent, J0, and J45 values are shown in gure 1. Also, the ARF rate in the study group was detected to be 24.5%.

Discussion
Previous studies showed that children with ADHD had a higher prevalence of refractive errors such as astigmatism, hypermetropia and amblyopia, Hence, vision screening is crucial for the early detection of refractive errors and prevent amblyopia in these children [3,12,5]. In this context, this is the rst study that compares the noncycloplegic refraction measurement values obtained by Spot Vision Screener, PlusoptiX A12, and Retinomax K-plus Screen, and cycloplegic values by Tonoref II in children with ADHD. Although noncycloplegic measurements of hand-held devices underestimate or overestimate the values in terms of spherical and cylindrical powers, axis, SE, J0, and J45 compared to cycloplegic measurements of autorefractometer, PlusoptiX A12 had an excellent interclass correlation with Tonoref II regarding the spherical and cylindrical powers, SE, and J45.
Studies investigating the visual function and ocular abnormalities in children with ADHD reported a higher frequency of reduced visual acuity, subnormal stereoacuity, refractive errors, convergence abnormalities, small optic discs, heterophoria, and cognitive visual problems [12]. Ocular abnormalities in ADHD were explained by various mechanisms consisting of neurological dysfunctions in cortical regions, inhibition impairment, decrease in the optic nerve's axonal volume, and balance control failure [3].
In children with ADHD, amblyopia, hypermetropia, astigmatism, and heterotropia were found 1.89, 1.82, 1.73, and 2.01 times more likely than without ones in a population based study [3]. Refractive errors may contribute to a decrease in concentration, leading to more evident symptoms of inattention and hyperactivity. Astigmatism and hyperopia affect both near and distance vision, unlike myopia, which affects only distance vision. Accordingly, astigmatism and hyperopia can affect an individual's attentional abilities more than myopia. In particular, uncorrected hyperopia and astigmatism, and amblyopia were found to be associated with reading impairment and thus academic performance [13,14]. Therefore, detecting refractive errors is also important in the differential diagnosis of learning disabilities accompanying ADHD [15].
Amblyopia is a common, preventable cause of visual impairment in children and adults with a prevalence of 1.6-3.6% and one of the associated risk factors is uncorrected refractive errors. Higher prevalences of amblyopia in ADHD were mentioned in previous studies above. Furthermore, in a recent study, it was reported that the risk of developing ADHD is higher (1.81 fold) in children with amblyopia. The relationship between ADHD and amblyopia was speculated on sharing similar predisposing factors such as developmental delay, mental retardation, and respiratory allergic diseases [16]. Although the ARF rate may differ according to study groups, in our study group, the ARF rate at 24.5% was higher than the ARF rate which is 3.6% in the general population [17].
Photoscreeners and hand-held autorefractometers have been used for screening refractive errors, and several studies were conducted to compare devices in healthy school-aged children and also some speci c patient populations, including autism and intellectual disability [18,19,7]. However, there was no study comparing non-cycloplegic measurements of photoscreeners and hand-held autorefractometers with cycloplegic measurements in children with ADHD in the literature.
Teberik et al. evaluated the comparison of non-cycloplegic measurements of three different photoscreeners consist of Plusoptix A12, Spot Vision Screener, and Retinomax K-plus 3 with cycloplegic autorefractometer measurements in healthy school-aged children. They reported that Plusoptix A12 had compatible results with cycloplegic autorefractometer in all measurements, but the integration level was moderate. [7]. The accuracy of Plusoptix A12 was reported better in the myopic, astigmatic, and anisometropic eyes than hyperopic eyes due to underestimating hyperopic refractive errors [17]. . Similarly, in our study, the Retinomax K-plus Screen had excellent reliability (ICC greater than 0.90) in terms of cylindrical power and J0. Akil et compared the cycloplegic and non-cycloplegic refractive measurements of Retinomax K-plus 3 and tabletop autorefractokeratometer with cycloplegic retinoscopy. They found a good agreement between Retinomax and autorefractokeratometer in J0 and J45, unaffected by cycloplegia [10].
The Spot Vision Screener has demonstrated good sensitivity and moderate speci city in detecting ARFs in children with developmental disabilities [6]. Teberik et al. reported moderate integration levels between non-cycloplegic measurements of Spot and cycloplegic measurements of autorefractometer in terms of spherical, cylindrical, and SE values [7]. In another study, Spot was found with high speci city to detect refractive errors; however, the sensitivity was found low for hyperopia [24]. In our study, we found excellent reliability in terms of cylindrical power and J0. However, there was a 0.33 D myopic shift of SE compared to cycloplegic measurements. Similarly, Qian et al. indicated that Spot had a myopic shift of 0.17 D compared to cycloplegic retinoscopy. They also reported that J0 and J45 values had medium correlations with cycloplegic retinoscopy. Nevertheless, the Spot was found to strongly agree with cycloplegic retinoscopy for the refractive error and strabismus evaluation [25].
The current study is limited by small sample size, lack of cycloplegic measurements of hand-held autorefractometer, and investigation of both eyes of children. Also, not comparing devices in terms of detecting ARFs may be another limitation of the study.
In conclusion, novel vision screening programs are widely used for detecting visual problems in children.
In our study, all of the hand-held devices showed excellent reliability in terms of cylindrical power and J0 and good reliability for J45. This may guide for prescribing cylindrical power and axis; however, cycloplegic measurements may be necessary for prescribing accurate spherical values. Even though all devices had advantages or disadvantages, Plusoptix A12 showed excellent reliability for detecting refractive errors in children with ADHD. The elimination of the cycloplegia requirement, remote measurement, and short examination time in these children may increase patient compliance.

Declarations
Con ict of interest / Competing interest: All of the authors declare they have no con ict of interest