Comparison between two devices in the detection of corneal thickness changes after cataract surgery

This study compared corneal thickness (CT) changes obtained with specular microscopy (SM) and a rotating Scheimpflug camera (RSC) after conventional phacoemulsification surgery (PS). One hundred sixty six eyes of 83 patients were analyzed before and one month after PS. One eye underwent PS, while the fellow phakic one was used as control. CT was measured with SM at the center of the cornea and with RSC at the pupil center, at the corneal apex and at the thinnest point. In the operated eye, SM showed a larger CT mean increase than those one detected at the three different measurements’ points evaluated by RSC. Inversely, in the fellow phakic eye, SM showed a greater CT mean decrease than those one registered by RSC at its three measurement’s points. Thus, one month after surgery, even if cornea appears clear at the slit-lamp, a significant thickness increase is still present. This is even more evident if the slight decrease of the fellow phakic eye is considered. The differences between the two devices are probably related to the different measured areas.

www.nature.com/scientificreports/ underwent refractive surgery were excluded. CT measured with both Nidek CEM-530 and Pentacam HR were compared before and one month after the surgical procedure, at the time of lens prescription. For each patient, one eye underwent surgery, while the fellow phakic one was used as control, to check if the difference of readings before and after was due to surgery and not related to the normal measurements' variation 19 . The present study adhers to the ethical principles of the Declaration of Helsinki. All the patients were carefully informed about the purpose of the study and written informed consent from each participant was acquired. Institutional Review Board (IRB) approval was also obtained from the ComEtico Campania Sud (CECS), Prot. No. 16544.

Instruments and methods.
Pentacam HR (Oculus, Wetzlar, Germany, version 1.19r11) utilizes a rotating Scheimpflug camera and a monochromatic slit light source (blue led at 475 nm), rotating together around the optical axes of the eye.
CEM-530 (Nidek Co, Japan) is an optical device that concurrently provides SM and pachymetric measurements.
CT was consecutively and randomly measured by the above-mentioned devices on the same day, between 10 and 12 a.m.
In the present study, measurements were automatically obtained with the two instruments. The patients were instructed to sit in front of the instrument, with their forehead and chin placed on the device supports, holding both eyes open and focusing on a blinking fixation target in the camera's center. On a computer screen, the operator visualized the image of the patient's eye and centered it by rotating the joystick of the device. When the image was correctly aligned, the patient was advised to remain still, keeping his eyes wide open, so the scan automatically began.
CT values obtained with CEM-530 from central cornea were compared with the three different CT points provided by Pentacam, named pupil center (PC), corneal apex (CA) and thinnest point (TP). Statistical analysis. All data were put in a Microsoft Excel spreadsheet and MedCalc 19.5.3 (Mariakerke, Belgium) was used for statistical evaluation. Kolmogorov-Smirnov test showed a normal distribution (p > 0.05) for all data. For this reason, Student paired T-test and Pearson correlation coefficient were performed 20,21 . Furthermore, reliability and agreement between the two devices were calculated, using the Bland-Altman plots, intraclass correlation coefficient (ICC), and 95% confidence interval of limits of agreement 22,23 . P values less than 0.05 were considered statistically significant. Maximizing the statistical power, sample size was determined. To perform the analysis, G*Power software (version 3.1.9.4) was utilized. Sample size was computed with two-tailed significance level (α = 0.05) and 90% power 24 . A difference between two dependent means (matched pairs) was calculated. Effect size was set as medium at around 0.361. Results showed: actual power = 0.902; non-centrality parameter δ = 3.289; critical t = 1.989; Df = 82; total sample size = 83.

Results
CT measurements with both CEM-530 and Pentacam HR and their respective ICC and Pearson correlation coefficients are summarized in Tables 1 and 2, showing that Nidek results are thicker than the Pentacam ones, both before and after surgery.
In addition, focusing on the operated eye and considering the differences between the preoperative and postoperative evaluations with the two devices, a statistically significant difference can be noticed only comparing CEM-530 with Pentacam HR TP (p < 0.05), whereas no statistically significant differences were found comparing CEM-530 with Pentacam HR PC and CA (p = 0.32 and p = 0.09, respectively). In other words, the two instruments in these points (PC and CA) show a similar mean CT difference, comparing the preoperative and postoperative evaluation.
Pearson correlation coefficients and limits of agreement between preoperative and postoperative data obtained by each device are summarized in Table 3. Bland-Altmann plots are depicted in Figs. 1, 2 and 3.

Discussion
Nowdays various devices are available for CT evaluation, therefore agreement studies are essential. Several authors evaluated corneal thickness changes in eyes subjected to cataract surgery, but most of them evaluated a small number of patients. Moreover, as far as we can tell, a comparison between these two devices before and after PS has never been performed.
In a study carried out with optical low coherence reflectometry and ultrasonic pachymetry on 30 patients, Ventura et al. 25 showed a considerable corneal thickness increase the day after surgery, that returned to the preoperative values in a period ranging from three months to one year.
In another study conducted with an optical low-coherence reflectometer slit lamp pachymeter on 13 patients, Salvi et al. 26 found a CT return to the preoperative value from the first week.
Kandarakis et al. 27 showed the same trend in 41 patients, but with values still high at seven days compared to the preoperative ones.
Utilizing Orbscan II topographer on 124 patients, De Juan 28 detected a significant decrease in CT that lasted up to the second week after cataract surgery, the period in which it tends to stabilize. On the other hand, the same authors noticed less significant CT decreases from the second to the fourth week when it usually tends towards the starting values. Caglar et al. 29 and Calabuig-Goena et al. 30 seem to have confirmed, in more recent times, these results in their studies.
Dzhaber et al. 31 , comparing CT measurements, provided by a non-contact SM, in 67 patients who underwent both conventional phacoemulsification surgery and femtosecond laser-assisted cataract surgery, found a non significant CT increase in both groups one month after surgery.
Concerning the comparison between instruments in normal patients that did not undergo cataract surgery, only two studies, utilizing the same devices used in the current research, were carried out by Karaca et al. 32  Karaca et al. 32 assessed CT values with Oculus Pentacam, Nidek CEM-530, and Konan Medical CellChek XL. They concluded that there is a good correlation between these three devices, with significant differences  www.nature.com/scientificreports/ between CEM-530 and Pentacam (p < 0.01). In particular, they reported CEM-530 measurements to be thinner than those obtained with Pentacam. On the contrary, De Bernardo et al. 17 found a good correlation between CEM-530 and Pentacam HR, with CEM-530 producing a thicker corneal measurement than Pentacam HR.
Compared to the aforesaid studies, in the present study, a group of operated eyes matched to the fellow phakic ones was evaluated. The same protocol was also used in the study of Salvi et al. 26 , but unfortunately, only a limited number of patients was evaluated.
Furthermore, to our knowledge, the present study compared for the first time CEM-530 and Pentacam HR CT values in patients before and after PS. www.nature.com/scientificreports/ In the operated eye, from these results, it is possible to highlight an increase in CT after one month. This is more evident with CEM-530 (p < 0.05), but it is also shown by Pentacam HR at PC (p < 0.05) and CA (p < 0.05), whereas no significant difference was found at TP (p = 0.15).
In our opinion, this statistically significant difference could be related to the unstable TP location, while PC and CA are stable.
In fact, the corneal thinnest point is not the same during repeated measurements over time but tends to change along with corneal remodeling following PS. www.nature.com/scientificreports/ This hypothesis could be strengthened when the two devices' differences between preoperative and postoperative period are considered, as a statistically significant difference, in this case, can be noticed just at the TP.
On the other hand, in the fellow phakic eye, no significant differences exist between the first and second Pentacam HR examination time, while CEM-530 points out a decrease in CT measurement (p < 0.05). This is probably related to the less accuracy of CEM-530 compared to Pentacam HR. Some of the reasons could be the lack of exam quality check provided by the instrument and the different areas evaluated by the two devices.  www.nature.com/scientificreports/ Moreover, both in operated and fellow phakic eye, CEM-530 supplies thicker CT values than Pentacam HR. These results, contrarily to Karaca et al. 32 , agree with the results provided by De Bernardo et al. 17 . Probably, this could be explained by the different devices utilized in these studies: Pentacam HR in the present and in the previous study by De Bernardo et al., and Pentacam in the Karaca et al. study 32 , though it could appear less plausible.
Another explanation could be related to the different points in comparison with the ones analyzed by Karaca et al. 32 , as well as Ucakhan et al. 33 . In fact, contrary to these studies, they did not report which RSC corneal parameters were compared with those provided by SM.
We acknowledge that this study has some limitations, among them, the potentially different area measured by the two devices, the short-term follow-up period and the lack of the endothelial cell count, as an increase in CT could be related to a decrease in cell count. To better evaluate the long term PS effects on CT, further studies with close-up measurements over time and a longer follow up should be performed.
In conclusion, although a great correlation between CEM-530 and Pentacam HR was found, these two devices should not be interchangeably utilized, as shown by Bland-Altmann plots.
In addition, one month after surgery, even if cornea appears clear at the slit-lamp, a significant corneal thickness increase is still present. This is even more evident if the slight decrease of the fellow phakic eye is considered. The differences obtained with the two devices are probably related to the different area measured by the two devices.

Data availability
The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.