Introduction
Auto-refractometers have an essential place in ophthalmology consultation to detect refractive errors. Various devices, such as conventional table-mounted auto-refractometers or portable devices, are commonly used. Although table-top auto-refractometers quickly measure refractive errors and provide valid results, they are bulky, cannot be mobilized, and are unsuitable for bedridden patients or patients with reduced mobility, nor young children [1, 2]. On the other hand, portable auto-refractometers are small, practical, and suitable for newborns, infants, bedridden patients, or those with reduced mobility, limiting their ability to sit on the examination chair.
In this study, we compared the measurements obtained using a table-mounted auto-refractometer (Topcon RM-800; Topcon Medical Systems, Inc. Japan) and a portable auto-refractometer (Nidek HandyRef-K; Nidek Co., Ltd. Japan). The main objective of the study is to verify if the two devices give comparable results and could, therefore, be used interchangeably. The secondary objectives are to determine the variations in the measurements of the astigmatism axis after tilting the head upwards with the portable auto-refractometer (patient sitting position, operator standing) and to specify the variability of inter-operator measurements with the portable measuring device.
Material and methods
Study design
This is a prospective study involving 100 eyes of 100 patients who consulted for corrective lens prescriptions. The study was conducted in July 2022 at the 5th Military Medical-Surgical Center of Errachidia.
Patients
The inclusion criteria were as follows: all patients aged over 3 years who consulted for decreased visual acuity or for a prescription or change of glasses.
The exclusion criteria were:
patients younger than 3 years old or over 3 years old for whom refraction measurement by table-mounted auto-refractometer is not feasible (non-cooperative subjects);
patients with a refractive error that is not measurable with one or both devices (myopia or cylinder beyond the measurement ranges of the machine);
any ocular pathology at the level of the anterior or posterior segment that interferes with auto-refractometer measurements (corneal opacity, cataract, vitreous disorder).
As refractive errors in both eyes are strongly linked, and for statistical analysis to be conducted on independent variables, only refractions of one eye of each patient were used for statistical analysis. The included eye was chosen randomly, unless one of them did not meet the inclusion criteria.
Measurements
Measurements were taken using two devices:
- • the manual auto-refractometer from Nidek (HandyRef-K; Nidek Co., Ltd., Japan) consists of two parts: the main body and the portable measurement device, which is wirelessly connected to the main body. It has an automatic fogging mechanism to minimize accommodation. It automatically records 10 eye measurements and provides the best result. Its measurement range is from -20.00 to +20.00 Diopters (D) for spheres and from 0 to 12 D for cylinders;
- • the table-mounted auto-refractometer from Topcon (Topcon RM-800; Topcon Medical Systems, Inc., Japan). It has a measurement range from –25.00 to +22 D and from 0 to 10 D for cylinders. It uses a color slide with an automatic fogging technique;
- • each participant underwent three refraction measurements performed by the ophthalmologist: one measurement using the table-mounted auto-refractometer from Topcon (Topcon RM-800; Topcon Medical Systems, Inc., Japan), then one measurement with the manual auto-refractometer from Nidek (HandyRef-K; Nidek Co., Ltd., Japan) in a simple seated position (patient and operator seated), and finally one measurement with the head tilted upward at about 45 degrees (patient seated, operator standing). In 20 of our participants, a fourth measurement was made using the manual auto-refractometer from Nidek performed by an assistant with no training in ophthalmology or knowledge of refraction (multifunctional nurse);
- • for children under 12 years old, measurements were taken under systematic cycloplegia using (Cyclopentolate 10mg/ml); beyond 12 years old, cycloplegia was performed as needed.
Statistical analysis
Statistical analysis was performed using SPSS software version 20. The Kolmogorov-Smirnov test was conducted to assess the normality of the distribution of variables. Student’s t-test and Wilcoxon test were used to test differences in variables. A p-value of < 0.05 was considered statistically significant. Concordance between measurements obtained by the two devices was evaluated using the Bland-Altman plot [3, 4] for each component of the power vector: spherical equivalent (SE), Jackson vertical cylinder (J0), and oblique Jackson cylinder (J45). These measurements were obtained from the sphere (S), cylinder power (C), and its axis (A) according to the formulas: SE = S + C/2; J0 = –C/2Cos(2A); J45 = –C/2 sin(2A). The limits of the 95% confidence interval were calculated using the mean ± 1.96 standard deviation (SD).
Results
Measurements were taken from 124 subjects, but 24 of them met the exclusion criteria. Thus, the total sample size in our study was 100 eyes of 100 subjects. The average age was 31 years, with extremes of 3 and 71 years. 56 patients were female (56%), and 52% were right eyes. The sphere of the studied population ranged from (–18.5D to +8D), and the cylinder ranged from (0 to –5.50 D), with a spherical equivalent between (–19.75D and +5.5D). The results obtained by both devices in the seated position are shown in Table 1.
|
Table 1. Comparison of mean values of measurements obtained by both devices |
|||
|
|
Topcon (n = 100) |
Nidek (seated position) (n = 100) |
p |
|
Sphere* |
–0.89 ± 4.39 |
–0.93 ± 4.38 |
0.09 |
|
Cylinder* |
–1.07 ± 0.85 |
–1.11 ± 0.86 |
0.18 |
|
Axis** |
102 ± 62 |
106 ± 62 |
< 0.001 |
|
Spherical equivalent * |
–1.42 ± 4.62 |
–1.47 ± 4.59 |
0.15 |
|
J0* |
–0.07 ± 0.54 |
–0.01 ± 0.47 |
0.24 |
|
J45* |
–0.05 ± 0.41 |
–0.03 ± 0.52 |
0.85 |
Our results indicate no significant differences between the two devices regarding sphere power (p = 0.09) and cylinder (p = 0.18). However, the HandyRef-K tends to overestimate the cylinder’s axis by an average of 4° (p < 0.001). However, directly comparing axes in degrees does not provide an interpretable appreciation of the statistical difference. On the one hand, a 4° difference in the axis of astigmatism, even if very statistically significant (p < 0.001), may not be noticeable by the patient regarding optical correction. On the other hand, it would be confusing to compare, for example, 178° and 1°; these two axes are very close but seem very different as numbers. Therefore, statistical analysis cannot be performed separately on raw cylinder powers and axes; it requires conversion into a power vector. According to Fourier, an arbitrary sphero-cylindrical power can be expressed by a spherical power (the spherical equivalent) and two crossed cylinders (Jackson), one on the axis of 0 degrees (J0) and the other at the axis of 45 degrees (J45). The power of these three components can be interpreted as the coordinates (x, y, z) of a vector representation of the power profile. This vector representation allows for a more reliable numerical and graphical analysis. According to our results, there is no difference between the two devices in terms of spherical equivalent (p = 0.15) and the J0 (p = 0.24) and J45 components (p = 0.85).
Table 2 compares measurements obtained by the same operator (ophthalmologist) using the HandyRef-K in the seated position with the head straight and with the head tilted upward at about 45° (patient seated, physician standing). It is noted that there is no difference between the sphere and cylinder power, spherical equivalent, J0, and J45.
|
Table 2. Comparison of mean values of measurements obtained by the HandyRef-K between the straight head position and the head tilted upward position |
|||
|
|
HandyRef-K (Straight head) (n = 100) |
HandyRef-K (Titled head) (n = 100) |
p |
|
Sphere* |
–0.96 ± 4.20 |
–0.99 ± 4.19 |
0.38 |
|
Cylinder* |
–1.11 ± 0.86 |
–1.13 ± 0.89 |
0.35 |
|
Axis** |
106 ± 62 |
108 ± 63 |
< 0.01 |
|
Spherical equivalent* |
–1.48 ± 4.59 |
–1.54 ± 4.58 |
0.24 |
|
J0 |
0.01 ± 0.47 |
–0.04 ± 0.55 |
0.38 |
|
J45 |
–0.03 ± 0.52 |
0.01 ± 0.46 |
0.43 |
The use of the portable refractometer by the assistant yielded similar results to those obtained by the physician (Tab. 3).
|
Table 3. Comparison of mean values of measurements obtained by the ophthalmologist and the assistant using the HandyRef-K |
|||
|
|
Ophthalmologist (n = 20) |
Assistant (n = 20) |
p |
|
Sphere* |
–0.32 ± 3.64 |
–0.31 ± 3.53 |
0.85 |
|
Cylinder* |
–1.21 ± 0.79 |
–1.22 ± 0.82 |
0.82 |
|
Axis** |
111 ± 61 |
114 ± 62 |
0.22 |
|
Spherical equivalent* |
–0.89 ± 3.81 |
–0.86 ± 3.70 |
0.74 |
|
J0* |
0.08 ± 0.49 |
0.08 ± 0.63 |
0.39 |
|
J45* |
0.03 ± 0.54 |
–0.03 ± 0.40 |
0.68 |
The agreement of measurements between the two devices is represented by Bland-Altman plots concerning the spherical equivalent (Fig. 1), the J0 vector (Fig. 2), and the J45 vector (Fig. 3). It is noted that there is a strong correlation of all these parameters, as the majority (95%) of measurements fall within the 2 limits of agreement. Therefore, the two devices can be used interchangeably.
Discussion
There are few publications that have studied the validity of portable autorefractors compared to table-mounted autorefractors, and their results are not unanimous. Our study has shown that the two devices studied can be used interchangeably. Sayed et al. [2] also found a very good correlation between measurements obtained by the portable Nidek auto-refractometer (Nidek ARK-30) and the table-mounted Huvitz HRK 7000A auto-refractometer concerning all parameters of subjective refraction and could thus be used interchangeably.
However, Seymen et al. found similar results between the HandyRef-K and the table-mounted auto-refractometer (Unicos, URK-800) in terms of cylindrical power and its axis, but not in spherical equivalent [7]. Similarly, Mirzajani et al. [8] found the difference in sphere and spherical equivalent to be significant but not clinically significant. The authors attribute this difference to the large sample size studied. Arici et al. [9] and Prabakaran et al. [10] also reported a significant difference in spherical and cylindrical error between portable autorefractors and table-mounted ones. This difference could be due to using a different portable device, the Retinomax. Wesemann et al. compared the Retinomax from Nikon to a conventional auto-refractometer, finding comparable results in adults and children after cycloplegia. However, results differed in children without cycloplegia [11].
Cyclotorsions occur when the eye rotates around the optical axis. They depend on head orientation. Static type occurs when the patient goes from the upright or seated position to the lying position. According to some studies, this transition induces cyclotorsion up to ±9° without being statistically significant [2, 12, 13]. In our study, we performed measurements at 45 degrees because, at 60 degrees and above, the “lying position” mode integrated into the HandyRef-K is automatically triggered and allows for astigmatism compensation. Thus, we found that the cyclotorsion between the seated position and the head tilted upward at about 45° was not significant, with a maximum amplitude of 7°. This change in axis could also be partly related to lateral micro-tilts of the head that cannot be visually controlled, unlike the table-mounted refractometer where the patient’s chin and forehead are placed on a support, allowing head position control.
The portable Nidek auto-refractometer starts measurement automatically when alignment is correct; therefore, it does not require an experienced operator. Thus, it offers the advantage of being usable by a healthcare professional without any training in ophthalmology. It could be an interesting option in screening campaigns for refractive errors, such as in school environments. In our study, measurements obtained by the ophthalmologist and by a multifunctional nurse using the HandyRef-K were consistent.
Our study has some limitations. Firstly, it is a single-center study with a limited number of cases. Also, the studied population is heterogeneous, including adults and children with some measurements performed with cycloplegia and others without cycloplegia. Another limitation is that the study compared two types of commercially available autorefractors; therefore, the results cannot be extrapolated to other brands of autorefractors or other measurement technologies.
Conclusion
Given the good agreement between the results of both devices, the portable Nidek HandyRef-K auto-refractometer can be used interchangeably with table-mounted autorefractors to measure refractive errors, especially in screening programs, epidemiological studies, for individuals with reduced mobility or when rapid measurements are required, particularly in children or non-cooperative individuals [8]. Similarly, the portable auto-refractometer can be used by non-ophthalmology professionals, which can be advantageous in mass screenings and school settings.
Data availability
The data used to support the findings of this study are available from the corresponding author upon request.
Ethics statement
This research was conducted in accordance with the ethical principles. Every effort was made to ensure the confidentiality and privacy of the participants. Any identifying information has been anonymized.
Additionally, conflicts of interest that could potentially influence the research findings have been disclosed. No conflicts of interest exist that could be perceived as prejudicing the impartiality of this research.
Should there be any questions regarding the ethical conduct of this research, please contact the corresponding author (T.A.)
Author contributions
T.A.: design of the study, realization, data analysis, writing; M.Y.: supervision and correction; O.A.: supervision and correction.
Conflict of interest
The Authors declare that they have no conflict of interest.
Funding
None declared.