Precise ocular biometry is crucial to the advancement of ophthalmology, necessitating ongoing updates to biometric instruments. Given the rapid development of new biometric technologies, it is essential to assess whether their measurements can be integrated into clinical practice and to validate their accuracy and precision. Reliability in clinical practice can only be assured through rigorous validation and comparison of data from emerging biometric devices, thereby enabling more precise diagnoses and enhanced treatment efficacy for patients [22]. The objective of this study is to assess the repeatability and reproducibility of the Colombo IOL instrument based on SD-OCT principle and their agreement with the IOLMaster 700 based on SS-OCT principle in normal subjects.

Our data demonstrated the high repeatability of the Colombo IOL in healthy young eyes. The Sw value of the AL in this study was 0.01 mm, and the ICC value was 1.000, showing excellent repeatability. In addition, AL also showed excellent reproducibility with an ICC value of 1.000. These results are similar to those previously reported in normal subjects with optical biometers based on a similar technology. Sikorski et al. used the Revo-NX (Optopol, Poland), based on the principle of SD-OCT, to measure AL in IOL eyes with repeatability and reproducibility ICC values of 1.000 [23]. Ni et al. using the AOCT-1000M (Aoying, China) based on the same SD-OCT principle as the Colombo IOL obtained an ICC value of 1.000 for AL [24]. Moreover, Domínguez-Vicent et al. analyzed a fully automated SS-OCT biometer, Eyestar 900 (Haag Streit AG, Koeniz, Switzerland), and obtained a low Sw value of 0.008 mm for AL [25].

SD-OCT has high resolution and faster image acquisition capability. The ICC, Sw, TRT, and CoV of CCT measured by Rao et al. using RTVue were 0.990, 2.2 μm, 4.2 μm, and 0.4%, respectively [26]. Mansoori et al. used RTVue to measure CCT in normal subjects with an ICC of 0.994 [27]. The ICC values of CCT measured by Ni et al. using AOCT-1000 M and RTVue were 0.998 and 0.994, respectively [24]. Hong et al. used RTVue to measure a TRT value of 4.7 μm for CCT in normal subjects [28]. In our study, we observed better repeatability and reproducibility of CCT measured by the Colombo IOL in contrast to previous studies, by low Sw (first observer: 1.63 μm, second observer: 1.18 μm), TRT (first observer: 4.52 μm, second observer: 3.26 μm) and CoV (first measurer: 0.31%, second measurer: 0.22%). Moreover, the ICCs for CCT (first observer: 0.997, second observer: 0.998) were almost close to 1.000. The CCT measurements also showed excellent reproducibility with low Sw of 0.90 μm, TRT of 2.49 μm, and CoV of 0.17%.

Previous studies have shown that biometers based on FD-OCT have good repeatability in measuring AQD, ACD, and LT. Shetty et al. used IOLMaster 700 and Anterion to measure the repeatability of ACD in cataract patients with ICC of 0.9972 and 0.9999, respectively [29]. Fişuş et al. used the IOLMaster 700 and Anterion to measure the ACD of cataract patients with Sw < 0.135 mm and CoV < 0.373 mm. Sw of AQD was < 0.133 mm and CoV was < 0.366 mm [30]. Montés-Micó et al. used instruments such as the IOLMaster 700 and Lenstar LS 900 to measure LT values of normal human eyes with good repeatability and reproducibility [31]. Venkataraman et al. used the MS-39 (CSO, Italy) based on SD-OCT in combination with Placido disc principle and Eyestar 900 (Haag Streit AG, Koeniz, Italy) based on SS-OCT principle. The CoV values of ACD in healthy subjects measured were all < 1.2% [32]. In our study, the repeatability Sw and TRT values of AQD, ACD and LT were all < 0.04 mm and < 0.12 mm, respectively, the ICCs were all > 0.980. The reproducibility Sw, TRT, ICCs of AQD and ACD were 0.02 mm, 0.06 mm and 0.995, respectively. The reproducibility of LT was good with Sw < 0.03 mm, CoV < 0.72% and ICC > 0.994. This demonstrates the excellent repeatability and reproducibility of AQD, ACD, and LT measurements with the Colombo IOL.

The Colombo IOL, utilizing radial scanning technology of SD-OCT, provided reliable corneal curvature data. This study found that the CoV for repeatability and reproducibility of Km measurements using the Colombo IOL was below 0.25%, and the ICC exceeded 0.994, demonstrating excellent repeatability and reproducibility. It would be beneficial to include comparisons with other device reports to highlight this performance. In our study, the AST parameters J0 and J45 measured by the Colombo IOL were found to be reliable. These results surpass other devices, such as the Cassini Color LED Corneal Analyzer (TRT: J0 = 0.42 D, J45 = 0.25 D), Humphrey Atlas 9000 based on Placido disc (TRT: J0 = 0.25 D, J45 = 0.39 D), and more so than the IOLMaster 700 based on SS-OCT (TRT: J0 = 0.33 D, J45 = 0.35 D) [33, 34].

In healthy eyes, the present study found similar measurement results for AL using the Colombo IOL and IOLMaster 700. The MD was − 0.03 ± 0.03 mm, and the absolute value of the 95% LoA was 0.08 mm. Although the difference was statistically significant, it is negligible in clinical practice and can be considered high agreement [35].

In the measurement of CCT, the MD between them was − 13.34 ± 4.21 μm, and the maximum absolute value of 95% LoA was 21.58 μm. Ruan et al. used the CASIA 2 and IOLMaster 700 to measure the 95% LoA of CCT in cataract patients, ranging from − 30.06 to 0.43 μm [36]. The differences between the two instruments are clinically acceptable, and these differences are smaller in our study. This means that even if there are small differences in the measurement of CCT between the two instruments, these differences are acceptable for clinical diagnosis and treatment due to their low degree of correlation with intraocular pressure.

Lender et al. used the Eyestar 900 and IOLMaster 700 to measure the ACD measurements of patients before cataract surgery, and the difference was not statistically significant [37]. In our results, ACD values measured by the Colombo IOL and IOLMaster 700 were similar, with 95% LoA ranging from − 0.01 to 0.14 mm, and 95% LoA of AQD values ranging from 0.01 to 0.15 mm. Since each 0.10 mm deviation of the ACD only leads to an IOL refractive error of about 0.14 D [35], the narrow 95% LoA shows a high degree of agreement between the two instruments. On the other hand, the difference in LT was 0.02 ± 0.04 mm, with a 95% LoA of − 0.05 to 0.10 mm, indicating good agreement.

Consistency of Km measurement: Eibschitz-Tsimhoni et al. found that when K changes by 1.0 D, the IOL power calculation changes by 0.8–1.3 D [38]. Jasvinder et al. stated that the difference between 1.0 D and 0.5 D in K translates into a IOL power difference of around 1.0 D and 0.5 D [39]. To ensure proper visual acuity following IOL implantation, we set a clinical difference threshold at 0.5 D. Güçlü et al. used Pentacam and SS-OCT equipment to measure the corneal curvature of healthy subjects and keratoconus patients, and found that the measured values of the two were very close, and the difference was not statistically significant [40]. In our study, the 95% LoA of Km ranged from − 0.13 to 0.40 D and the absolute value of the 95% LoA (0.4 D) would still be below the threshold for clinical difference. Clearly, the two devices are clinically interchangeable when measuring Km.

Özyol et al. compared the Pentacam and IOLMaster 700 while measuring the 95% LoA of J0 in normal population from − 0.10 to 0.24 D, and the 95% LoA of J45 from − 0.31 to 0.27 D; they noted that the two devices had good consistency and could be used interchangeably [41]. The results of AST parameters measured by the Colombo IOL and IOLMaster 700 were found to be similar to those obtained previously. The mean J0 difference was 0.12 ± 0.11 D, and the maximum absolute value of 95% LoA was 0.34 D. The MD of J45 was 0.09 ± 0.08 D, and the maximum absolute value of 95% LoA was 0.25 D. These data showed the good agreement between these two biometers in measuring corneal AST.

Yang et al. used the IOLMaster 500, IOLMaster 700, and Argos devices to measure WTW in patients prior to cataract surgery [42]. The 95% LoA of WTW measured by the IOLMaster 500 and IOLMaster 700 was − 0.761 to 0.432 mm, and the 95% LoA of WTW measured by the IOLMaster 500 and Argos was − 1.641 to 0.631 mm. The 95% LoA measured by the IOL Master 700 and Argos ranged from − 1.458 to 0.748 mm. The IOLMaster devices measure the diameter of the corneal contour based on the camera image, while the Argos makes the measurement by identifying the junction between the cornea and the iris from OCT images. This different measurement method may have contributed to the differences in measurement results. Yang et al. also noted that when using the Holladay formula for calculating IOL power with WTW, the results of the two SS-OCT devices would be different [43]. In our study, the MD in WTW was 0.68 ± 0.40 mm, with a wide 95% LoA ranging from − 0.11 to 1.47 mm. Therefore, compared with the IOLMaster 700, the WTW measurements of the Colombo IOL tend to be larger because the Colombo IOL recognizes the junction between the cornea and the iris from OCT images, and these two instruments need to be used with caution when measuring WTW. When performing pre-cataract measurements and calculations, careful consideration should be given to the equipment used and the formula employed to ensure the accuracy of the diopter calculation.

Pupil size plays a crucial role in improving visual performance of biological and environmental factors, and age, illumination, and refractive error are important factors that affect pupil size [44]. In this study, the MD in PD was 0.67 ± 0.84 mm with a 95% LoA ranging from − 0.97 to 2.31 mm. The Colombo IOL tends to have larger PD measurements compared to the IOLMaster 700, and the two instruments need to be carefully interchanged in terms of PD measurements.

The detection of proportional bias for AL, CCT, and WTW suggests that as the magnitude of these parameters increases or decreases, the difference between the two devices also changes systematically. This highlights the need for caution when using these devices interchangeably for these particular measurements, especially in cases where extreme values of AL, CCT, or WTW are observed.

The limitations of this study include the absence of patients with ocular diseases such as cataract, glaucoma, and corneal diseases such as keratoconus and only enrolled young healthy myopic populations. Therefore, we need to further explore the repeatability, reproducibility, and consistency of the instrument measurements in disease states. Future studies should aim to broaden these findings to provide a more comprehensive comparison and evaluation of the performance differences among various instruments and their accuracy in measuring ocular biological parameters.