The new dynamic real-time visualization optical biometer ZW-30, using SS-OCT technology to obtain eye measurements, may be a powerful tool for clinical application. The aim of this study was first to evaluate the repeatability and reproducibility of this biometer, then to evaluate its agreement with another commonly adopted device utilizing the same technology, the IOLMaster 700. Based on the results, we report two main findings: (1) the new biometer exhibited outstanding intraobserver repeatability and interobserver reproducibility (ICCs for almost all parameters were higher than 0.900); (2) all anterior parameters and AL measurement data were interchangeable between the new biometer and IOLMaster 700.

In our study, the AL values measured by the new SS-OCT based biometer showed the best repeatability and reproducibility among the available parameters with an ICC of 1.000. Although there was a statistically significant difference in AL measurement between the new device and the IOLMaster 700, the mean difference was rather small (0.00 ± 0.01 mm) and the maximum absolute 95% LoA was only 0.02 mm. This finding was in accordance with the study by Panthier et al. [10], who reported that the mean difference between the IOLMaster 700 and ANTERION (Heidelberg Engineering GmbH, Heidelberg, Germany) was 0.01 mm. In another study, Liao et al. [11] studied 103 healthy eyes and found that the mean difference between the IOLMaster 700 and OA-2000 was 0.00 ± 0.02 mm and the maximum absolute 95% LoA was 0.03 mm. Since a measurement error of 1 mm of AL value induces 2.5 D to 3.0 D deviation in IOL power calculation [10, 12], a 0.02 mm AL difference would reflect a 0.025 D to 0.030 D refractive error, which is difficult to distinguish for the human eye. Thus, the measuring of AL can be performed on the ZW-30 during clinical use. In addition, the scanning speed of ZW-30 is ten times faster than traditional scanning optical biological measurement instruments, requiring shorter patient cooperation time, and real-time data collection during the measurement process, resulting in better data correlation. The number of scanning lines and scanning range used for each measurement of multi-directional radiation scanning has doubled, resulting in a larger amount of data collection. For data with significant errors caused by blinking, eye rotation, and etc., the average value can be deleted to further ensure the accuracy of the data and analysis. Further, ZW-30 provides a real-time dynamic view of the entire eye from the cornea to the retina to determine whether the axial measurement is from the anterior surface of the cornea to the fovea of the retina, and thus reduces the risk of refractive error caused by incorrect measurement due to undetected poor fixation. Therefore, its utility in clinical practice is justified.

For CCT, our results found no significant difference between the ZW-30 and IOLMaster 700. The maximum absolute 95% LoA shown in the Bland–Altman plot was 8.08 µm. A study comparing the IOLMaster 700 with the Anterion obtained a maximum absolute 95% LoA of 19.05 µm [10]. Montes-Mico et al. [13], Liao et al. [11], and Cheng et al. [14], compared the IOLMaster 700 with the OA-2000 and found a maximum absolute 95% LoA of 19.72 µm, 24.67 µm, and 24.40 µm, respectively. Our result was much smaller than those reported in the above-mentioned studies. Considering the narrow 95% LoA and no significant mean difference values, we conclude that the ZW-30 and IOLMaster 700 can be used interchangeably for CCT measurements.

With regard to AQD, ACD and LT, the ZW-30 and IOLMaster 700 displayed excellent agreement, as the 95% LoA ranged, from − 0.07 mm to 0.04 mm, − 0.07 mm to 0.04 mm, and − 0.07 mm to 0.08 mm, respectively. ACD and LT are important parameters for calculating IOL power, especially with last generation formulas [12]. Besides, LT has been shown to play a role in ICL sizing [15]. A 1 mm error in ACD and LT measurement may lead to an approximately 1.0 D to 1.5 D difference of IOL power [16, 17], demonstrating that the differences revealed in the current study would not have any clinically detectable effect. The results are in good agreement with those reported by Omoto et al. [18], Liao et al. [11], and Dong et al. [19].

Despite the statistically significant difference (P < 0.001), good agreement between the ZW-30 and IOLMaster 700 was observed for Km (95% LoA: − 0.16 D to 0.30 D). Hua et al. [20] proposed that a 1.00 D measurement difference in Km would result in a 1.40 D difference of IOL power. Based on this, it could be inferred that a difference of 0.30 D in keratometric power would lead to an IOL power difference of approximately 0.42 D, which lies within the usual 0.50 D step increments of IOLs. Similar to our result, a previous study evaluated the Km values obtained by the IOLMaster 700 and OA-2000 and reported a mean difference of 0.00 ± 0.09 D with narrow LoA range [11]. Moreover, our team had compared the SS-OCT based device with the Scheimpflug based optical biometer (Pentacam AXL, OCULUS) and found comparable outcomes between both devices (95% LoA: − 0.48 D to 0.09 D) [21]. However, Tañá-Rivero et al. [22] analyzed the interchangeability between the IOLMaster 700 and the Pentacam AXL and demonstrated that the LoA range was wide and may have a significant impact, especially when selecting the Toric IOL power. Thus, the agreement between the ZW-30 and devices based on other corneal topography measurement principles in keratometric value measurement still needs to be further studied.

The mean difference values of AST, J0, and J45 measured in our study between the ZW-30 and IOLMaster 700 were 0.00 ± 0.15 D, 0.00 ± 0.08 D, and − 0.05 ± 0.09 D, among which the difference in J45 was statistically significant (P < 0.001). Nevertheless, the LoAs range were all narrow, with the widest being 0.29 D, marginally above the 0.25 D clinical limit. The double-angle plot also showed that the distribution of corneal AST measured by these two devices was similar, suggesting that the differences between these two devices can be considered clinically negligible.

Due to the popularity of phakic IOL implantation surgery, accurate measurements of CD have attracted the attention of surgeons [23]. In addition, an increasing number of new IOL formulas (such as Barrett Universal II and Holladay 2 formulas) also consider this parameter as one of the predicting variables [24]. Dong et al. [19] demonstrated a 0.24 ± 0.30 mm significant difference and a wide 95% LoA range from − 0.83 mm to 0.35 mm between the IOLMaster 700 and ANTERION. In another study, Shetty et al. [25] studied 127 eyes and found the maximum absolute 95% LoA to be 0.76 mm, indicating that the potential differences in CD value measurement should be non-negligible in clinical practice. Contrary to these studies, the current study concluded that the ZW-30 and IOLMaster 700 had high agreement in the measurement of the CD distances, with a narrow 95% LoA (− 0.39 mm to 0.10 mm). Therefore, we can conclude that the CD data are interchangeable and can be used for clinical practice.

There are several limitations in our study. The first drawback is that only normal unoperated myopic eyes were included. Hence, the conclusions could not be extended to those who had other types of refractive errors or had other ocular diseases history (such as keratoconus and cataract). Besides, the mean AL value in this study was 25.76 ± 1.10 mm (range: 23.31 mm to 29.16 mm). No short eyes (AL < 22.0 mm) and only three long eyes (AL > 28.0 mm) were included in the whole dataset, which warrant further studies. Finally, we only compared the new instrument with the same measurement technology based device (IOLMaster 700). More efforts will be made to compare it with other devices based on different technologies to better evaluate its precision.