Refractive Precision of Ray Tracing IOL Calculations Based on OCT Data versus Traditional IOL Calculation Formulas Based on Reflectometry in Patients with a History of Laser Vision Correction for Myopia

Figures & data

Table 1 Demographics

Eyes		37
Subjects		20
Sex, F		30%
		Mean ± SD	Range
Age (years)		56,9 ± 4.9	49 to 66
K (D)^a		40.5 ± 1.9	36.5 to 44.4
Corneal astigmatism^a		0.86 ± 0.40	0.20 to 1.74
ACD (mm)^a		3.35 ± 0.33	2.8 to 4.0
AL (mm)^a		25.3 ±1.2	22.6 to 28.06
Previous LVC (D)	SE (n=31)^b	−3,7 ± 2.6	−10 to −1.6
	Cyl (n=18)^b	−1.6 ±1.8	−5,3 to 0
IOL power implanted		20.3 ± 2.3	15 to 24.5
Toric IOLs (IOL cyl 1.0/1.5)		65% (35%/30%)

Notes: ^aLenstar measurement; ^bsubjects with reliable information on the previous LVC treatment.

Abbreviations: F, female; D, diopters; SD, standard deviation; RLE, refractive lens exchange; SE, spherical equivalent refraction; Cyl, cylinder refraction; LVC, laser vision correction; IOL, intraocular lens; IOL cyl, IOL cylindrical power.

Table 2 Refractive Results

	Mean ± SD	Range	p^a
Months postop	2.8 ± 0.8	1.5 to 4.1
UDVA (logMAR)	0.00 ± 0.09	−0.17 to 0.22	0.8
CDVA (logMAR)	−0.06 ± 0.07	−0.18 to 0.07	<0.01*
SE (D)	0.05 ± 0.31	−0.88 to 0.75	0.23
Cyl (D)	−0.26 ± 0.30	−1.00 to 0	<0.01*
SE ≤ ±0.25	68%
Cyl ≤ 0.5	86%

Notes: ^aWilcoxon sign-rank test difference from zero; *statistically significant.

Abbreviations: SD, standard deviation; UDVA, uncorrected distance visual acuity, CDVA, corrected distance visual acuity; SE, spherical equivalent refraction; Cyl, cylinder refraction.

Table 3 Arithmetic RPE, Absolute RPE and Median Absolute Error

	Eyes	Arithmetic RPE		Absolute RPE
Calculation/formula	n	Mean ± SD	Min to max (range)	Mean ± SD (range)	MedAE** (range)
Anterion-OKULIX	25	−0.13 ± 0.30	−0.7 to 0.41 (1.11)	0.26 ± 0.19 (0.64)	0.21 (0.57)
Casia-OKULIX	37	0.29 ± 0.36*	−0.44 to1.31 (1.75)	0.35 ± 0.30 (1.31)	0.23 (1.00)
Barret TK NH	37	−0.32 ± 0.27*	−0.87 to 0.40 (1.27)	0.35 ± 0.24 (0.87)	0.16 (0.72)
Haigis-L	37	−0.40 ± 0.34*	−1.08 to 0.55 (1.63)	0.45 ± 0.26* (1.05)	0.18 (0.95)

Notes: *Mixed models estimates statistically significantly different from Anterion-OKULIX (Holm-Bonferroni adjusted p-values); **arithmetic mean error reduced to zero.

Abbreviations: MedAE, median absolute error; Barret TK NH, Barret true K no history.

Figure 1 Boxplot of (A) arithmetic and (B) absolute prediction error.

Notes: *Adjusted p ≤0.05; ***adjusted p <0.001 (mixed models estimates different from Anterion OKULIX).

Abbreviation: RPE, refractive prediction error.

Figure 2 Percentages of eyes within certain range of RPE.

Notes: *Logistic mixed models estimates statistically significantly different from Anterion-OKULIX (Holm-Bonferroni adjusted p-values).

Abbreviations: Abs, absolute; RPE, refractive prediction error; p, adjusted p-value.

Figure 3 Bland–Altman plot with 95% limits of agreement between two repeated measurements/calculations of RPE with (A) Anterion-OKULIX, and (B) Casia-OKULIX.

Figure 4 Correlation between RPE and aqueous depth prediction error for the OCT devices.

Notes: *Slope statistically significant; **intercept and slope statistically significant.

Abbreviations: RPE, refractive prediction error (D); AqdPE, aqueous depth prediction error (mm); Calc, IOL-calculation.