FAROOK, MOHAMED DipOptom; VENKATRAMANI, JAYANT MBBS; GAZZARD, GUS MA(Cantab), FRCOphth; CHENG, ANGELA BSc; TAN, DONALD MBBS, FRCOph; SAW, SEANG-MEI MBBS, PhD
The Retinomax (handheld autorefractor) and table-mounted autorefractors have become increasingly important components of refractive testing in optometric practice in recent times,1–8 especially in children. They have also gained favor as quick screening tools1,2,4–6,9,10 and have proven useful for efficient data collection in epidemiologic studies.
The Retinomax autorefractor is especially convenient because it is portable and has potential for use in large-scale studies involving refraction, especially those conducted outside the clinical or laboratory setting. Furthermore, some studies have revealed it to be as reliable as a table-mounted autorefractor and subjective refractive testing.2,3,6,9–12
The Nikon Retinomax (Tokyo, Japan) is a monocular autorefractor. It is portable, light, and quick to use. Although many studies have indeed found the Retinomax autorefractor to be as reliable as its table-mounted counterpart in children,2,3,9,11,12 there have been few studies to our knowledge evaluating the use of the Nikon Retinomax autorefractor in adults.3,9,11 Prior studies in adults either had a wide age range11 or compared two refractive methods,3,9,11 Retinomax with tabletop autorefraction or subjective refraction, on the same group of subjects.
The primary objective of this study is to therefore to evaluate the accuracy of the Retinomax autorefractor in testing refraction in comparison with the table-mounted autorefractor and subjective refraction testing in Singapore adults.
A study was carried out on 100 adults 21 years of age and above recruited after chart reviews from the Singapore National Eye Centre over a 2-month period from 2001 to 2002. One hundred consecutive subjects from refractive surgery clinics between the ages 21 to 40 years were examined. The mean age of the subjects was 29.5 years (standard deviation = 5.3). The study population consisted of 74% females and 26% males. The racial distribution of the population comprised of 80% Chinese, 11% were Asian Indians, and 9% were Malays. All subjects with prior ocular surgery, or with decreased vision resulting from any cause including amblyopia or ocular media opacity, were excluded. Only those subjects who had the best-corrected visual acuity of at least 20/20 were included in the study. Written informed consent was obtained after the nature of the study was explained to each subject. Approval for the study was granted by the Ethics Committee, Singapore Eye Research Institute, and the conduct of the study followed the tenets of the Declaration of Helsinki.
The subject’s eye testing occurred in three stages and all measurements were performed by a trained optometrist. First, measurements were taken with subjective refraction using trial lenses in a trial frame. Second, a handheld autorefractor, the Nikon Retinomax, was used as a measure of refraction. Third, the Topcon RM8000B—a table-mounted autorefractor—was used to obtain refractive readings. In all three stages, the right eye was always tested first, followed by the left eye.
For the first stage, subjective refraction testing using trial frame and trial lenses were carried out in a dim-lit room. Fogging lens of +3 D was used on the nontested eye to eliminate the possibility of accommodation because cycloplegia was not used. Subjects had to read the Snellen chart at 6 m. The final spherical power was defined as the highest plus value or the lowest minus that gave the best Snellen visual acuity. Jackson cross cylinder of ±0.50 D was used to determine the amount of astigmatism.
Handheld autorefraction was then carried out. The Nikon Retinomax is a portable handheld autorefractor and has a measurement range of –18 to +22 D for spheres and 8 D for cylinders. The autorefractor works in two modes. One is the normal mode, which uses the inbuilt fogging mechanism, and the other is the quick mode, in which the fogging mechanism is turned off. In this study, all subjects underwent handheld autorefraction using the normal mode. Measurement occurs once there is adequate alignment and the fixation target is fogged to reduce accommodation. The instrument takes a number of readings on each eye and prints a maximum of the last eight readings recorded. It then averages the readings and reproduces the representative reading as a separate line. While taking measurements with Nikon Retinomax, a consistency value is printed by the instrument. This is a measurement of the reliability of measurements and can assume values from zero to 10. In all measurements, the consistency value was at least 8, indicating high reliability. Readings with values <8 were repeated until values of 8 or more were obtained.
Finally, the Topcon RM8000B table-mounted autorefractor was used to obtain refractive readings. The measurements were started once the subject’s pupil was aligned and focused on the built-in viewing monitor of the instrument. Eight consecutive readings were taken. To ensure reliability, the measurement values had to be free of bracketed values—an in-built mechanism indicative of low reliability—and in-between readings obtained should not be more than ±0.75 D in sphere and cylinder, respectively.
The calibration of the table-mounted and the Retinomax autorefractors was checked at the start of the study and the beginning of each day, before testing on the subjects, using model eyes. The same autorefractors were used for all subjects throughout the study. Following a standard protocol, ophthalmologic examinations, including cover tests and slit lamp examinations, were performed to exclude amblyopia, strabismus, ocular media opacity, or any other causes of decreased vision besides refractive error.
Refractive error was expressed as spherical equivalent (SE = sphere + half negative cylinder). SE obtained for both the right and left eyes were similar (Spearman correlation coefficient for Retinomax, table-mounted, and subjective refractions for the right and left eyes were 0.95, 0.96, and 0.95, respectively, all p < 0.001). Therefore, only data obtained from the right eye were used. The astigmatism analysis used the vector system whereby refractive error is expressed using the three vectors: M, J0, and J45. These vectors have a real physical meaning: M is simply the spherical equivalent, which is commonly used by researchers. J0 and J45 describe astigmatism. J0 describes the difference in refractive error between vertical and horizontal meridians, being positive for with-the-rule astigmatism and negative for against-the-rule astigmatism. The final component, J45, is usually small and describes the oblique component of astigmatism: being positive if negative cylinder axis is closer to 45° and negative if the axis is closer to 135°.
The data were analyzed using the commercially available software SPSS version 12.0 (Chicago, IL).
The following three comparisons were made:
1. Retinomax was compared with the table-mounted autorefractor;
2. Retinomax was compared with subjective refraction; and
3. The table-mounted autorefractor was compared with subjective refraction.
The Bland-Altman plots of the differences of the two readings against the mean of the readings were constructed for each of the three comparisons. The mean differences and their confidence intervals are calculated. The 95% limits of agreement (LoA) for individual differences are presented. The 95% LoA were derived by the mean difference ± 1.96 × standard deviation (SD) in which SD is the standard deviation of the differences between the two readings. For all analyses, a p value of < 0.05 was considered statistically significant.
The mean spherical equivalent measured by the Retinomax was −4.69 D (SD = 3.38). This was more minus compared with the table-mounted (mean = −4.05 D, SD = 3.33) and the subjective refraction (mean = −3.90 D, SD = 3.18). The cylinder power was larger for the Retinomax autorefractor (mean = −0.83 D, SD = 0.80); the cylinder power for the table-mounted autorefractor was −0.70 D (SD = 0.70) and for subjective refraction, it was −0.54 D (SD = 0.64). The Fourier-transformed vectors (J0 and J45) were also compared and significant differences for comparisons between subjective refraction and Retinomax autorefraction, and table-mounted autorefraction and Retinomax autorefraction were found (Table 1).
The mean SE difference between the measurements taken with the table-mounted autorefractor and those taken using the Retinomax were significantly different: 0.63 D (95% confidence interval [CI], 0.51–0.75). The Retinomax consistently recorded more minus values than its table-mounted counterpart. The mean SE difference between table-mounted autorefraction and Retinomax autorefraction for younger adults (mean = 0.57 D) was lower than that for older adults (mean = 0.73 D). The Retinomax also consistently recorded more minus SE values than subjective refraction, with a mean difference of 0.78 D (95% CI, 0.64–0.93). The difference in measurement between subjective refraction and handheld autorefraction was larger in older adults (mean = 0.89 D) than in younger adults (mean = 0.72 D). The Bland-Altman plots comparing the subjective refraction and table-mounted autorefractor with the Retinomax autorefractor are shown in Figures 1 and 2. The mean difference for the Retinomax autorefractor and subjective refraction was 0.78 (95% LoA, −0.57–1.83), which was slightly larger when compared with the Retinomax autorefractor and the table-mounted autorefractor (mean = 0.63; 95% LoA, −0.57–.83). The table-mounted autorefractor and subjective refraction were the most comparable, with the refraction measurements differing by only 0.15 D (Fig. 3). The 95% limits of agreement for the comparisons of table-mounted autorefraction and Retinomax, and of subjective refraction and Retinomax, are wider compared with the comparisons of subjective refraction with table-mounted autorefraction.
The refraction measurements taken using the Nikon Retinomax autorefractor were found to be more minus compared with subjective refraction and table-mounted autorefraction using the Topcon RM8000B, and there were significant differences in the J0 and J45 comparisons of subjective versus the Retinomax, and table-mounted versus the Retinomax.
Our study results are consistent with data from previous studies, which have found moderate similarities between handheld autorefractors and their table-mounted counterparts, although the values from the handheld autorefractor in our study were more minus.2,3,9–11 This is similar to the findings by Liang et al who found that noncycloplegic autorefraction gave a more myopic refraction than cycloplegic autorefraction in children.12 They felt that accommodation most likely explained the more minus result, which may be the case in our study also. Other possibilities are that the difference could be attributed to recruitment strategies, age of participants, and the range of refractive errors in the studies.
Most prior studies were conducted in children. Recent studies by Buchner on 216 children aged between 3.5 and 4.5 years found that noncycloplegic handheld autorefraction was comparable to tabletop autorefraction for cylinder power and axis with limitations in accuracy for the spherical equivalent.2 Wesemann et al in Germany examined 100 young adult eyes and 50 eyes of children aged 2 to 10 years and revealed that the handheld autorefractor (Retinomax) was fairly comparable to subjective refraction results.3,9 Similarly, Kallay et al found no significant differences in sphere, cylinder, or axis for the Nikon Retinomax and the table-mounted autorefractors (Topcon RM-A 6000) in 132 children and young adults.10 Cordonnier found no significant bias in spherical equivalent after examining 276 subjects (age range, 2–57 years) with Retinomax and tabletop autorefractors.11
Other studies, however, have found good correlations between values recorded by the handheld autorefractor and the gold standard subjective refraction measurements.3,6 A study by Harvey et al of 36 Native American preschoolers revealed values of the Nikon Retinomax to be reproducible and reliable in young children under cycloplegia.6 Wesemann et al also found the handheld autorefractor—Retinomax K-Plus—to be a valuable tool in adult refractive testing, having good correlation with values obtained by subjective refractive testing.3 Interestingly, however, the measurements obtained by handheld autorefraction in the study by Wesemann were found to be consistently more plus than those obtained by subjective refraction in contrast to the results from our study.
There are several positive aspects of our study setup. One observer performed all the refraction measurements and thus interobserver bias was negated. Our study population consisted only of adults, giving way to better analysis of the use of Retinomax among adults. Retinomax has already been extensively studied in younger population groups. This study also did not use cycloplegia in measurement of refractive values. Cycloplegia was deliberately not used in this study to assess the potential use of Retinomax as a quick, convenient tool for future large-scale epidemiologic studies and as a possible quick screening or diagnostic tool for measuring refraction.
Many limitations are also inherent in the study. The same observer was not masked to subjective refraction readings while documenting the handheld or table-mounted autorefractor measurements, allowing for possible observer bias. Because no cycloplegia was used, it could have led to pseudoaccommodation despite the use of fogging while taking measurements. This could in turn have been a cause for the consistently more minus values being recorded using the Nikon Retinomax, as compared with subjective and table-mounted refraction measurements. Wesemann et al found that cycloplegia is necessary for precise refractive measurement in children.9 The study noted that minus overcorrection occurred by up to –2.0 D when handheld autorefraction was used without administration of cycloplegics. Another study carried out in children also cautioned the possibility of overdiagnosis of myopia by handheld autorefraction if used alone without cycloplegia.13
This study suggests that the handheld Nikon Retinomax cannot serve as a good alternative tool for use in large-scale refraction studies among adults and cannot remove the need for less portable table-mounted or more time-consuming subjective refractory testing. However, it can serve as a good screening tool for populations, but it must be noted that although there is moderate correlation of values with subjective refraction, in our study, myopia could be overdiagnosed by the sole use of handheld autorefraction without cycloplegia, dependent on what definitions of myopia were used. We should thus be cautious before using the handheld Nikon Retinomax in research studies because the Retinomax autorefractor readings gave readings that were more minus compared with both autorefraction and subjective refraction. More large-scale studies would need to be conducted to enable accurate calibration of the handheld autorefractor for use without cycloplegia in adults and children.
Singapore National Eye Centre
11 Third Hospital Avenue
Republic of Singapore
1. Steele G, Ireland D, Block S. Cycloplegic autorefraction results in pre-school children using the Nikon Retinomax Plus and the Welch Allyn SureSight. Optom Vis Sci 2003;80:573–7.
2. Buchner TF, Schnorbus U, Grenzebach UH, Stupp T, Busse H. [Examination of preschool children for refractive errors. First experience using a handheld autorefractor.] Ophthalmologe 2003;100:971–8.
3. Wesemann W, Dick B. [Experiences with the hand-held autorefractometer ‘Retinomax’ in adults and children.] Klin Monatsbl Augenheilkd 1997;211:387–94.
4. Cordonnier M, Kallay O. Non-cycloplegic screening for refractive errors in children with the hand-held autorefractor Retinomax: final results and comparison with non-cycloplegic photoscreening. Strabismus 2001;9:59–70.
5. el-Defrawy S, Clarke WN, Belec F, Pham B. Evaluation of a hand-held autorefractor in children younger than 6. J Pediatr Ophthalmol Strabismus 1998;35:107–9.
6. Harvey EM, Miller JM, Wagner LK, Dobson V. Reproducibility and accuracy of measurements with a hand held autorefractor in children. Br J Ophthalmol 1997;81:941–8.
7. Kinge B, Midelfart A, Jacobsen G. Clinical evaluation of the Allergan Humphrey 500 autorefractor and the Nidek AR-1000 autorefractor. Br J Ophthalmol 1996;80:35–9.
8. McCaghrey GE, Matthews FE. Clinical evaluation of a range of autorefractors. Ophthal Physiol Opt 1993;13:129–37.
9. Wesemann W, Dick B. Accuracy and accommodation capability of a handheld autorefractor. J Cataract Refract Surg 2000;26:62–70.
10. Kallay OP, Cordonnier MJ, Dramaix MM. Cycloplegic refractive errors in children: comparison of a standard and a hand-held refractor. Strabismus 1998;6:3–7.
11. Cordonnier M, Dramaix M, Kallay O, de Bideran M. How accurate is the hand-held refractor Retinomax(R) in measuring cycloplegic refraction: a further evaluation. Strabismus 1998;6:133–42.
12. Liang CL, Hung KS, Park N, Chan P, Juo SH. Comparison of measurements of refractive errors between the hand-held Retinomax and on-table autorefractors in cyclopleged and noncyclopleged children. Am J Ophthalmol 2003;136:1120–8.
13. Iuorno JD, Grant WD, Noel LP. Clinical comparison of the Welch Allyn SureSight handheld autorefractor versus cycloplegic autorefraction and retinoscopic refraction. J AAPOS 2004;8:123–7.