Investigating Factors Associated With Vault Prediction Errors In The Nk Icl Sizing Formula

The NK formula with anterior segment OCT (CASIA2; TOMEY) is widely used for ICL size selection. In this study, we investigated the factors that affect the prediction error when using the NK formula.

This study included 272 eyes of 178 patients (mean age 31.6 ± 6.3) who underwent ICL implantation at Nagoya Eye Clinic between July 2022 and March 2024, with at least 3 months of follow-up. ICL sizes were calculated using the NK formula. 

Patients were classified into three groups based on the vault prediction error (VPE) calculated by subtracting the 3-month postoperative vault from the predicted vault the overestimation (OE) group (VPE ≧ 200 μm), the underestimation (UE) group (VPE ≦ -200 μm), and the accurate estimation (AE) group (-200 μm < VPE < 200μm).We analyzed age, refractive error, axial length, ICL size and power, fixation direction, and anterior segment OCT parameters. Comparisons were made between UE vs. AE and OE vs. AE, with logistic regression used to identify factors associated with vault prediction errors.

Subjective spherical power was significantly lower in the UE group (-7.14 ± 2.52 D) than in the AE group (-6.31 ± 2.58 D) (p < 0.05). Anterior chamber depth(ACD) was also significantly lower in the UE group (3.21 ± 0.27 mm) than in the AE group (3.31 ± 0.28 mm) (p < 0.05). Crystalline lens rise(CLR) was significantly higher in the UE group (0.11 ± 0.26 μm) than in the AE group (0.02 ± 0.18 μm) (p < 0.01). Logistic regression analysis revealed that larger angle opening distance(AOD) (odds ratio: 11.2) and CLR (50.6) were associated with an overestimation of vault (p < 0.05). Conversely, no parameters were identified that explained an underestimation of vault.

In conclusion, AOD and CLR were identified as parameters associated with overestimated vaults when using the NK formula. Further optimization of the NK formula can be expected by re-evaluating CLR already included in the NK formula and by investigating the relationship between AOD and vault. In contrast, we couldn’t find any parameters that explain underestimated vaults. Unknown factors such as anatomical structures of the ciliary sulcus or iris cysts may be involved.