Comparison of the Accuracy of Four Pentacam Corneal Astigmatism Values in Non-Toric Pseudophakic Eyes

Tianyu Zheng,Jie Xu,Yi Lu
DOI: https://doi.org/10.1007/s00417-019-04585-8
2020-01-01
Graefe s Archive for Clinical and Experimental Ophthalmology
Abstract:Purpose To compare the accuracy of different corneal astigmatism values measured by Scheimpflug keratometry (Pentacam), including Simulated Keratometry (SimK) and three total corneal astigmatism values, equivalent K reading (EKR), true net power (TNP), and total corneal refractive power (TCRP). Methods We enrolled 168 eyes of 168 patients with non-toric IOL implantation. Pentacam examination and subjective refraction were performed 3 months after surgery. The agreement, arithmetic difference, and vector difference between refractive astigmatism (RA) and different corneal astigmatism values were compared. Results Differences in astigmatism magnitude were significant between SimK and RA in the against-the-rule (ATR) and with-the-rule (WTR) groups but not in total corneal measurements. The meridians of SimK and RA differed significantly in the oblique astigmatism group. The correlations between total corneal astigmatism values and RA were stronger than that between SimK and RA in the total, WTR, and oblique astigmatism groups in Pearson's correlation test. Bland-Altman plots revealed more data points exceeding the limits of agreement (LoA) in SimK measurement in total and WTR subjects. In the ATR group, fewer data points exceeded LoA in EKR. The mean difference vector between SimK and RA was larger than that of other measurements in each astigmatism group. The arithmetic mean of difference vector was significantly smaller in EKR in the total, WTR, and oblique groups. Conclusions Among different Pentacam readings, corneal astigmatism measurements considering anterior and posterior corneal surfaces were more representative of total ocular astigmatism than SimK, and EKR showed markedly better performance in astigmatism estimation.
What problem does this paper attempt to address?