Comparing the accuracy of different intraocular lens power calculation formulas using IOL-Master 700

doi:10.21203/rs.3.rs-2090195/v1

Download PDF

Research Article

Comparing the accuracy of different intraocular lens power calculation formulas using IOL-Master 700

https://doi.org/10.21203/rs.3.rs-2090195/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Purpose: The aim of this study was to compare the accuracy of different IOL power calculation formulae in the IOL-Master 700 and to establish the impact of ACD on different IOL power calculation formulae in intermediate eyes.

Methods: A total of 80 patients who underwent uncomplicated phacoemulsification surgery between July 2020 and June 2022 were analyzed retrospectively. Eight formulae (SRK/T, Hoffer Q, Haigis, Holladay 2, Kane, Barrett Universal II, T2, and the Ladas Super Formula) were compared.

Results: This study included 113 eyes. SRK/T achieved the lowest MAE (0.30D), MedAE (0.25D), SD of prediction error (0.33), and the greatest percentage of eyes within ±0.25D (50.4%), ±0.50D (83.2%), and ±1.00D (100%). According to the percentage of eyes within ±0.50D, Kane showed better results than Barret Universal II (82.3% vs 79.6%) while traditional formulae (Holladay 2, Hoffer Q, and Haigis) had poor refractive outcomes in comparison to the new generation formulae. For the shallow ACD group, the lowest MAE and MedAE were noted in the Barrett Universal II formula (MAE= 0.26D, MedAE= 0.20D) while the highest percentage of eyes within ±0.50D was observed in the Ladas Super Formula (87.1%). There were no statistically significant differences between the formulas in normal and deep ACD groups.

Conclusion: SRK/T was the most accurate formula for the overall study, followed by new-generation formulae. For normal and deep ACD groups all eight formulae had similar results while for the shallow ACD group Barrett Universal II and the Ladas Super Formula performed better.

intraocular lens

formula

power calculation

IOL-Master 700

Representing the most common and curable cause of visual impairment worldwide, cataract is treated surgically nowadays by removing the natural lens of the eye and replacing it with an intraocular lens (IOL) [1]. The selection of the proper IOL power is one of the most important planning steps for cataract surgery [2]. The refractive outcome has dramatically changed with the evolution of surgical techniques and is affected by the selection and accuracy of the IOL power formula and the biometer used for measurements [3]. IOL-Master 700 is an innovative optical biometer used to obtain data such as axial length (AL), anterior chamber depth (ACD), central corneal thickness (CCT), and lens thickness (LT) due to swept source optical coherence tomography (SS-OCT) incorporated in this device [4].

Over the past several years, a variety of formulae have been described, but there is no consensus on which of them are generally better. The new generation formulae (Kane, Barrett Universal II, T2, the Ladas Super Formula) use more than AL and ACD to predict the lens power, attempting to replace the traditional ones. These formulae are still being studied to provide good results for both clinicians and patients [5]. The Kane and Ladas Super Formula are based on artificial intelligence (AI) and Barrett Universal II is based on ray tracing, while the T2 formula is a modification of the original SRK/T [6]. The postoperative refractive prediction errors can be minimized by choosing the appropriate biometric variables and the proper IOL formula and being also aware of the differences between short and long eyes [7].

The aim of this study was to compare the accuracy of different IOL power calculation formulae (SRK/T, Hoffer Q, Haigis, Holladay 2, Kane, Barrett Universal II, T2, and the Ladas Super Formula) using IOL-Master 700 and to establish the impact of ACD on different IOL power calculation formulae in intermediate eyes.

We analyzed retrospectively all the biometric and refractive data obtained from 80 patients who underwent cataract surgery at Acibadem University Atakent Education and Research Hospital (Istanbul) between July 2020 and June 2022. This study was approved by Acibadem University Ethical Committee and adhered to the Declaration of Helsinki. We included in our study all patients who had phacoemulsification with implantation of Acrysof IQ SN60WF (Alcon Laboratories) performed by a single surgeon through a 2.4 mm main incision. All patients used Tobramycin 0.3% and Dexamethasone 0.1% eye drops with tapering for a month after the surgical procedure. Exclusion criteria were any ocular diseases that could affect the refractive status after surgery, previous corneal surgeries, intra- and postoperative complications, and incomplete documentation.

We used IOL-Master 700 (software version 1.88, Carl Zeiss Meditec AG) to obtain the preoperative biometric data. The SRK/T, Hoffer Q, Haigis, and Holladay 2 formulae were calculated automatically by the software. The Kane formula (https://www.iolformula.com), Barrett Universal II formula (https://calc.apacrs.org/barrett_universal2105/), T2 formula (http://www.richardsheard.net/T2Formula.aspx), and the Ladas Super Formula (https://www.iolcalc.com/home) were calculated manually using the online calculators. The implanted IOL power was noted. Subjective refraction and spherical equivalent (SE) were obtained 1 week and 1 month postoperatively.

We accepted the surgical success as the percentage of eyes within ± 0.50 diopters (D) of refractive error as suggested by Hoffer et al. [8]. The refractive prediction error (PE) was determined by calculating the difference between the first week and first month of postoperative SE and the predicted postoperative refraction. The mean PE, mean absolute error (MAE), median absolute error (MedAE), and the percentage of eyes that had a PE of ± 0.25D, ± 0.50D, ± 0.75D, and ± 1.00D were calculated for all 8 formulae. The subgroups were made according to AL: short eyes (≤ 22 mm), intermediate eyes (22–25 mm), and long eyes (≥25 mm), and to IOL power: low-power (≤ 18D), medium-power (18.5-24.0D) and high-power (≥24.5D). The group with normal axial length was analyzed according to ACD: shallow eyes (≤ 3 mm), normal eyes (3.01–3.49 mm), and deep eyes (≥3.5 mm).

All the data was tabulated in Excel (version 16.16.10, Microsoft Corp). SPSS software (version 26.0, SPSS Inc.) was used for statistical analysis. The Kolmogorov-Smirnov test was used to test normality. The Friedman test with Wilcoxon signed-rank post hoc analysis and Bonferroni correction were used for statistical comparison. Correlation analysis was performed to identify variables that significantly correlated with postoperative 1st -month SE. In case of a significant result, the regression analysis was conducted to predict postoperative 1st -month SE. A p-value less than 0.05 was considered statistically significant.

A total of 113 cataract extraction and IOL implantation cases were analyzed. The demographic and anatomic variables of all patients were shown in Table 1.1. The mean age was 67.12± 11.01, while the female/male ratio was 44/36. The mean AL was 23.69 ± 0.87 mm and the mean ACD was 3.21 ± 0.35 mm. The mean implanted IOL power was 21.10 ± 2.47D and the mean postoperative 1^st-month SE was -0.42 ± 0.30D. The subgroup according to AL (Table 1.2) revealed that the majority of eyes (88.5%) were intermediate eyes, while the subgroup according to IOL power (Table 1.3) showed that we frequently implanted medium power lenses (78.8%). The eyes with normal axial length were divided into three groups (Table 1.4): shallow eyes (31%), normal eyes (49%), and deep eyes (20%).

Table 1.1 The demographic and anatomic variables of all patients

Variables	Value
Patients/Eyes	80/113
Eyes: OS/ OD	51/62
Female/ Male	44/36
	Mean ± SD
Age (yrs)	67.12± 11.01
AL (mm)	23.69 ± 0.87
ACD (mm)	3.21 ± 0.35
LT (mm)	4.46 ± 0.41
WTW (mm)	11.93 ± 0.29
Implanted IOL power (D)	21.10 ± 2.47
Post-op first week SE (D)	-0.37 ± 0.30
Post-op first month SE (D)	-0.42 ± 0.30

OS, left eye; OD, right eye; AL, axial length; ACD, anterior chamber depth; LT, lens thickness; WTW, white-to-white; IOL, intraocular lens; SE, spherical equivalent; D, diopter; SD, standard deviation.

Table 1.2 Subgroup analysis according to AL

	Frequency	Percent (%)
Short Eye (≤ 22 mm)	2	1.8
Intermediate Eye (22-25 mm)	100	88.5
Long Eye (≥ 25 mm)	11	9.7
Total	113	100.0

AL, axial length.

Table 1.3 Subgroup analysis according to IOL Power

	Frequency	Percent
Low Power (≤ 18 D)	16	14.2
Medium Power (18.50-24.00 D)	89	78.8
High Power (≥ 24.50 D)	8	7.1
Total	113	100.0

Table 1.4 ACD Groups of Eyes with Intermediate Length (N= 100)

ACD Groups	Frequency	Percent (%)
≤3 mm	31	31
3.01- 3.49 mm	49	49
≥3.5 mm	20	20
Total	100	100

ACD, anterior chamber depth

The minimum, maximum, and mean values of implanted IOL power and recommended IOL power to achieve emmetropia of each formula were indicated in Table 2. The range of the implanted IOL power was between 15D and 27D. The mean IOL power of the Kane formula (21.02 ± 2.45D) was closer to the mean implanted IOL power (21.10 ± 2.47D) when compared to the other formulas. The Barret Universal II and the Ladas Super formulas had the lowest SD value (2.45), while the Hoffer Q formula had the highest SD value (2.72).

Cumulative percentages within ±0.25D, ±0.50D, ±0.75D, and ±1.00D in addition to the mean PE, MAE, and MedAE for each formula were reported in Table 3 and the binary comparison of AEs was presented in Table 4. When MAEs were compared, SRK/T and Barrett Universal II formulas had the lowest values (0.30D and 0.31D, respectively) while Holladay 2 formula had the greatest value (0.41D). There were significant differences between SRK/T and Holladay 2, Hoffer Q, Haigis, and Ladas Super Formula (p<0.05), but not between SRK/T and the new generation formulae (Kane, p=0.440; Barret Universal II, p=0.882; T2, p= 0.063). For Barret Universal II the difference was statistically significant compared to Holladay 2, Hoffer Q, Haigis, Kane, and Ladas Super Formula (p<0.05), but not in comparison to SRK/T (p=0.882) and T2 (p=0.289). SRK/T, Kane, T2, and Ladas Super Formula had the lowest SD of PE (0.33 and 0.35, respectively) while Hoffer Q had the highest SD of PE (0.40). According to the MedAE, the lowest and the greatest values were achieved by SRK/T (0.25D), Kane (0.27D), Barret Universal II (0.27D), and Hoffer Q (0.39D), respectively.

Table 2. Descriptive statistics of implanted IOL and recommended IOL diopter to achieve emmetropia of each formula for all patients (N= 113)

	Minimum	Maximum	Mean ± SD
Implanted IOL power (D)	15.00	27.00	21.10 ± 2.47
Recommended IOL power to achieve emmetropia (D)
SRK/T	14.74	26.72	20.75 ± 2.46
Holladay 2	14.55	27.18	20.97 ± 2.64
Hoffer Q	14.39	27.44	20.88 ± 2.72
Haigis	14.73	27.09	20.96 ± 2.61
Kane	15.00	27.00	21.02 ± 2.45
Barrett Universal II	14.79	26.68	20.73 ± 2.45
T2	14.8	26.8	20.79 ± 2.53
Ladas Super Formula	14.76	26.86	20.81 ± 2.50

D, diopter; IOL, intraocular lens; SRK/T, Sanders-Retzlaff-Kraff/theoretical; SD, standard deviation

Table 3. Prediction Errors of each formula for all patients (N= 113)

Formula	SRK/T	Holladay 2	Hoffer Q	Haigis	Kane	Barrett Universal II	T2	Ladas Super Formula
Mean Prediction Error (D)± SD	-0.19± 0.33	-0.35± 0.38	-0.27± 0.40	-0.30± 0.39	-0.20± 0.35	-0.16± 0.37	-0.19± 0.35	-0.24± 0.35
Mean Absolute Error (D)± SD	0.30± 0.23	0.41± 0.30	0.40± 0.27	0.40± 0.29	0.32± 0.25	0.31± 0.26	0.33± 0.24	0.34± 0.25
Median Absolute Error (D)	0.25	0.37	0.39	0.35	0.27	0.27	0.29	0.29
Prediction error (% of eyes)
±0.25 D	50.4	34.5	33.6	35.4	46.0	49.6	45.1	44.2
±0.50 D	83.2	64.6	68.1	68.1	82.3	79.6	81.4	77.0
±0.75 D	93.8	87.6	88.5	86.7	93.9	93.8	94.2	92.9
±1.00 D	100	93.8	98.2	98.2	98.2	98.2	99.1	98.2

D, diopter; SRK/T, Sanders-Retzlaff-Kraff/theoretical; SD, standard deviation

Table 4. Binary comparison of AE values of formulas for all patients (N=113)

	SRK/T	Holladay 2	Hoffer Q	Haigis	Kane	Barrett Universal II	T2	Ladas Super Formula
SRK/T		0.000*	<0.001*	<0.001*	0.440	0.882	0.063	0.019*
Holladay 2	<0.001*		0.754	0.841	<0.001*	<0.001*	<0.001*	<0.001*
Hoffer Q	<0.001*	0.754		0.952	<0.001*	<0.001*	<0.001*	<0.001*
Haigis	<0.001*	0.841	0.952		<0.001*	<0.001*	<0.001*	0.001*
Kane	0.440	<0.001*	0.001*	<0.001*		0.048*	0.633	0.240
Barrett Universal II	0.882	<0.001*	<0.001*	<0.001*	0.048*		0.289	0.003*
T2	0.063	<0.001*	<0.001*	<0.001*	0.633	0.289		0.518
Ladas Super Formula	0.019*	<0.001*	<0.001*	0.001*	0.240	0.003*	0.518

SRK/T, Sanders-Retzlaff-Kraff/theoretical.

Examining the percentages of eyes within ±0.50D prediction error (Figure 1) the values are between 64.6% and 83.2%. SRK/T obtained the highest percentage of ±0.50 D (83.2%), followed by Kane (82.3%) and T2 (81.4%) while Holladay 2 reported the lowest percentage (64.6%), followed by Hoffer and Haigis (68.1%). SRK/T was also the formula that achieved the largest percentage of eyes within ±0.25D (50.4%) and ±1.00D (100%). From the new generation formulae, Barret Universal II yielded the largest percentage of eyes within ±0.25D (49.6%), while T2 obtained the largest percentage of eyes within ±1.00D (99.1%). Subgroup analysis could not be performed due to the low number of short/long eyes and low/high IOL power. There were no significant correlations between each formula's recommended IOL power for emmetropia, postoperative 1^st-month SE, and age.

Descriptive characteristics within the 3 ACD groups were presented in Table 5. There was a significant difference between the 2 groups (shallow and normal eyes) for LT (p=0.001, p=0.033). The greatest mean LT was observed in shallow eyes (4.69 mm) while the lowest mean LT was noticed in deep eyes (4.14 mm). Even if the axial length was between 22 and 25 mm, the groups with the highest mean AL, listed in order, were: deep eyes (23.95 mm), normal eyes (23.62 mm), and shallow eyes (23.21 mm). Refractive prediction errors, MAE, MedAE, and the percentages of eyes within ±0.25D, ±0.50D, ±0.75D, and ±1.00D for each ACD subgroup were included in Table 6. The lowest MAE and MedAE were noted in Barret Universal II (0.26D, and 0.20D, respectively) for shallow eyes, and SRK/T for normal eyes (0.30D, and 0.21D, respectively), and deep eyes (0.29D, and 0.22D, respectively) while the highest MAE and MedAE were observed in traditional formulae for all the ACD subgroups. Cumulative percentages of AE groups of each three ACD subgroups are also shown in Figure 2, Figure 3 and Figure 4. The percentage of eyes within ±0.50D was higher for the Ladas Super Formula for shallow eyes (87.1%), SRK/T and T2 Formula for normal eyes (83.7%), SRK/T and Ladas Super Formula for deep eyes (85%). There was a statistically significant difference between formulas just in the first group represented by shallow eyes (p<0.05). A positive correlation between the prediction error and ACD was found for the following formulae: SRK/T (p=0.31), Hoffer Q (p=0.001), and T2 (p=0.014).

Table 5. Descriptive Data within the ACD Subgroups

	ACD (mm)	Lens Thickness (mm)	AL (mm)
Group 1: ACD ≤3 mm (N= 31)
Mean ± SD	2.81± 0.19	4.69± 0.35	23.21± 0.67
Range	2.16- 3.00	4.00- 5.38	22.16- 24.34
Group 2: ACD 3.01- 3.49 mm (N= 49)
Mean ± SD	3.24±0.15	4.39± 0.37	23.62± 0.66
Range	3.02- 3.49	3.64- 4.91	22.03- 24.99
Group 3: ACD ≥3.5 mm (N= 20)
Mean ± SD	3.70± 0.18	4.14± 0.29	23.95± 0.63
Range	23.01- 24.98	3.58– 4.79	23.01- 24.98

Table 6. Prediction Errors of each formula for ACD groups

Formula	SRK/T	Holladay 2	Hoffer Q	Haigis	Kane	Barrett Universal II	T2	Ladas Super Formula
Group 1 (ACD of ≤3 mm, N= 31)
Mean Prediction Error (D)± SD	-0.31± 0.27	-0.48± 0.33	-0.51± 0.25	-0.41± 0.29	-0.23± 0.26	-0.20± 0.33	-0.32± 0.25	-0.27± 0.24
Mean Absolute Error (D)± SD	0.35± 0.21	0.51± 0.28	0.51± 0.25	0.43± 0.26	0.27± 0.21	0.26± 0.28	0.34± 0.21	0.30± 0.20
Median Absolute Error (D)	0.30	0.48	0.51	0.43	0.24	0.20	0.28	0.27
Prediction error (% of eyes)
±0.25 D	45.2	19.4	12.9	29.0	54.8	64.5	41.9	45.2
±0.50 D	80.6	51.6	48.4	71.0	83.9	83.9	80.6	87.1
±0.75 D	93.5	80.6	83.9	83.9	96.8	96.8	96.8	96.8
±1.00 D	100	93.5	100	100	100	96.8	100	100
Group 2 (ACD 3.01- 3.49 mm, N= 49)
Mean Prediction Error (D)± SD	-0.17± 0.35	-0.32± 0.37	-0.22± 0.37	-0.26± 0.38	-0.17± 0.39	-0.14± 0.40	-0.18± 0.37	-0.34± 0.37
Mean Absolute Error (D)± SD	0.30± 0.25	0.37± 0.32	0.34± 0.26	0.36± 0.30	0.34± 0.26	0.34± 0.26	0.31± 0.26	0.34± 0.26
Median Absolute Error (D)	0.21	0.29	0.29	0.28	0.26	0.28	0.26	0.29
Prediction error (% of eyes)
±0.25 D	51	42.9	46.9	44.9	49	46.9	49	46.9
±0.50 D	83.7	69.4	73.5	69.4	79.6	79.6	83.7	73.5
±0.75 D	91.8	89.8	91.8	85.7	89.8	89.8	91.8	91.8
±1.00 D	100	93.9	98	98	98	98	98	98
Group 3 (ACD of ≥3.5 mm, N= 20)
Mean Prediction Error (D)± SD	-0.16± 0.35	-0.36± 0.39	-0.18± 0.47	-0.35± 0.48	-0.29± 0.42	-0.23± 0.35	-0.10± 0.36	-0.34± 0.40
Mean Absolute Error (D)± SD	0.29± 0.24	0.43± 0.30	0.41± 0.28	0.48± 0.34	0.40± 0.30	0.35± 0.23	0.31± 0.20	0.44± 0.28
Median Absolute Error (D)	0.22	0.39	0.41	0.38	0.41	0.36	0.31	0.40
Prediction error (% of eyes)
±0.25 D	55	35	25	25	30	40	45	35
±0.50 D	85	60	80	60	75	70	80	85
±0.75 D	95	90	90	90	90	95	100	95
±1.00 D	100	100	95	95	95	100	100	100

D, diopter; SRK/T, Sanders-Retzlaff-Kraff/theoretical; SD, standard deviation

SRK/T was the most accurate formula in our cohort represented almost entirely by intermediate eyes (88.5%), achieving the lowest MAE (0.30D), MedAE (0.25D), SD of prediction error (0.33), and the greatest percentage of eyes within ± 0.25D (50.4%), ± 0.50D (83.2%), and ± 1.00D (100%). SRK/T is a third-generation IOL formula [9] that uses linear regression combined with a theoretical eye model [10] that ensures a successful refractive outcome among the traditional formulas which presented the lowest percentages of eyes within ± 0.50D (Holladay 2, 64.6%; Hoffer Q, 68.1%; Haigis, 68.1%), but also the greatest MAE (Holladay 2, 0.41D; Hoffer Q, 0.40D; Haigis, 0.40D) and MedAE (Hoffer Q, 0.39D; Holladay 2, 0.37D; Haigis, 0.35D). Different studies conducted by Hoffer [11], Aristodemou et al. [12], and Olsen et al. [13] reported the superiority of the SRK/T formula in medium-long and long eyes, respectively.

The new generation formulae showed similar results regarding MAE (Barret Universal II, 0.31D; Kane, 0.32D; T2, 0.33D; Ladas Super Formula, 0.34D), SD of PE (0.35D for Kane, T2, Ladas Super Formula; 0.37D for Barret Universal II), and MedAE (0.27 for Kane and Barret Universal II; 0.29 for T2 and Ladas Super Formula). According to the percentage of eyes within ± 0.50D, Kane was the most accurate formula (82.3%), followed by the T2 formula (81.4%), the Barrett Universal II formula (79.6%), and the Ladas Super Formula (77%). All four formulas varied from 79.6–82.3% of eyes within ± 0.50D, which is known to be a good refractive outcome that reduces the necessity for spectacles [14, 15]. As a particularity Kane is the only formula that uses gender as a constant to predict the IOL position while Barret Universal II uses white-to-white (WTW) measurement and both are using LT as an optional variable [6].

Excellent results have been reported by multicentric studies, especially for Kane and Barret Universal II formulae. Kane et al. [16] used the same lens as in our study (Acrysof IQ SN60WF), but a previous-generation biometer (IOLMaster 500) and demonstrated that Barret Universal II was the most accurate formula (72.3% of eyes within ± 0.50D). Melles et al. [17] also showed the superiority of the Barret Universal II formula (80.8% of eyes within ± 0.50D) by implanting 2 different lenses (one of them was used in our study) and using a different biometer device (Lenstar 900). Kane formula was the most accurate in studies of Darcy et al. [18] and Savini et al. [19], reporting a percentage of eyes within ± 0.50D of 72% and 90%, respectively. Different biometer devices were used but Savini et al. performed the study using the same lens as in our study. The most recent studies used IOLMaster 500 [20] and IOLMaster 700 [21, 22] but not the same IOL as in our study, and reported comparable results for Kane and Barret Universal II formulae regarding the percentage of eyes within ± 0.50D (Pereira et al. Kane, 77.7%; Barret Universal II, 77.4%, Ryu et al. Barret Universal II, 88.5%; Kane, 85.9%, Chang et al. Kane, 70.9%; Barret Universal II, 64.6%). According to these results, a newer IOL formula should achieve more than 80% to be taken under consideration for replacing the Barret Universal II formula [23].

The T2 formula and the Ladas Super Formula are based on traditional formulae which they improved to increase the refractive outcome. Both formulae had not been studied as much as Barret Universal II and Kane formulae. The T2 Formula is a modified version of the SRK/T formula intending to correct the corneal height prediction [6], the “cusp phenomenon” mentioned by Haigis et al. [24], Kane et al. [16] and Taroni et al. [25] used the same lens as in our study, but different optical biometers and reported a percentage of eyes within ± 0.50D of 70% and 85.71%, respectively, for the T2 formula. The Ladas Super Formula combines traditional formulae (Holladay 1, Holladay 2, Haigis, SRK/T) and uses AI such as the Kane formula. This formula was calculated only in one study conducted by Pereira et al. [20]. The percentage of eyes within ± 0.50D was 75.9% for the Ladas Super Formula and 76.4% for the T2 formula, showing comparable outcomes with our study even if they used different IOLs and optical biometers.

The position of the IOL was predicted by many formulas using ACD measurements [26] and for that reason the subgroup analysis was made according to ACD. There were statistically significant differences between formulas just in the shallow ACD group, so in the other two groups, all the formulas showed similar results. The lowest MAE and MedAE were noted in the Barrett Universal II formula (MAE = 0.26D, MedAE = 0.20D) while the highest percentage of eyes within ± 0.50D was observed in the Ladas Super Formula (87.1%), both of them using preoperative ACD values as a predictor in calculating the IOL power. Gökce et al. [26] reported that the Barrett Universal II formula had the best results for shallow eyes (MAE = 0.24D, MedAE = 0.18D, 90.2% of eyes within ± 0.50D), confirming that formulae including ACD perform better in eyes with ACD less than 3 mm. The shallow eyes had the greatest mean LT value (4.69 mm) as shown in our study too. As a result, it may be concluded that increased lens thickness and innate anatomical factors can both contribute to a shallow anterior chamber [26].

The impact of AI was observed in all medical fields but especially in ophthalmology. Despite screening and diagnostic studies, remarkable progress in developing new formulas for calculating the IOL power has been made [27–29]. The main reason for introducing AI in different formulas was to achieve a better refractive outcome more than the 74% of eyes within ± 0.50D reported by the European Registry of Quality Outcomes for Cataract and Refractive Surgery (EURQUE) [30]. In our study, we evaluated two formulae using AI and the percentage of eyes within ± 0.50D was better for the Kane formula (82.3%) than the Ladas Super Formula (77%). There was no statistically significant difference between Kane and the Ladas Super Formula (p = 0.478) in our study. Similar results were obtained in two more studies conducted by Taroni et al. [25] and Carmona-Gonzalez et al. [31] showing there was no statistically significant difference between formulae using AI.

Our study had some limitations due to the low number of patients compared to the previous multicentric studies. We minimized this disadvantage by using the same IOL model, the same biometer device, and a standardized surgical technique performed by a single surgeon who followed the same pre and postoperative protocol for all patients. Secondly, the subgroup analysis was affected by the lack of extreme biometric values, so we could not analyze the accuracy of each formula for the subgroups made according to AL and IOL power. Furthermore, we only used Acrysof SN60WF IOL and its power was between 6.0 and 30.0D so extreme length eyes were not included in this study. Since there were not enough short and long eyes, we focused on the intermediate eyes group which we divided based on the ACD and tried to demonstrate the importance of this variable in predicting the appropriate IOL formula. Another limitation of our study was related to the inclusion of both eyes for some patients, even if the recommendation of Hoffer et al. [32] was to include only one eye for each patient to avoid ambiguous elements. Finally, there are still new formulase(Pearl-DGS, Hill-RBF 2.0, EVO, Olsen, etc) that we did not include in our study.

SRK/T was the formula that provided the most accurate results for the overall study, followed by new-generation formulae (Kane, Barrett Universal II, and T2). For normal and deep ACD groups all 8 formulas had similar results while for the shallow ACD group Barrett Universal II and the Ladas Super Formula performed better, showing that preoperative ACD has an important role in predicting the IOL power.

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Competing Interests

The authors have no relevant financial or non-financial interests to disclose.

Thompson J, Lakhani N. Cataracts. Prim Care. 2015;42(3):409-423.
strong> Lee AC, Qazi MA, Pepose JS. Biometry and intraocular lens power calculation. Curr Opin Ophthalmol. 2008;19(1):13-17.
Kane JX, Chang DF. Intraocular Lens Power Formulas, Biometry, and Intraoperative Aberrometry: A Review. Ophthalmology. 2021;128(11):e94-e114.
Shi Q, Wang GY, Cheng YH, Pei C. Comparison of IOL-Master 700 and IOL-Master 500 biometers in ocular biological parameters of adolescents. Int J Ophthalmol. 2021;14(7):1013-1017.
Luo Y, Li H, Gao L, et al. Comparing the accuracy of new intraocular lens power calculation formulae in short eyes after cataract surgery: a systematic review and meta-analysis. Int Ophthalmol. 2022;42(6):1939-1956.
Savini G, Taroni L, Hoffer KJ. Recent developments in intraocular lens power calculation methods-update 2020. Ann Transl Med. 2020;8(22):1553.
Kieval JZ, Al-Hashimi S, Davidson RS, et al. Prevention and management of refractive prediction errors following cataract surgery. J Cataract Refract Surg. 2020;46(8):1189-1197.
Hoffer KJ. Accuracy f ultrasound intraocular lens calculation. Arch Ophthalmol. 1981;99(11): 1819-1823.
Siddiqui A, Devgan U. Mastering lens calculations: New formula and comparisons. Current Ophthalmology Reports. 2018.
Retzlaff JA, Sanders DR, Kraff MC. Development of the SRK/T intraocular lens implant power calculation formula [published correction appears in J Cataract Refract Surg 1990 Jul;16(4):528]. J Cataract Refract Surg. 1990;16(3):333-340.
Hoffer KJ. Clinical results using the Holladay 2 intraocular lens power formula. J Cataract Refract Surg. 2000;26(8):1233-1237.
Aristodemou P, Knox Cartwright NE, Sparrow JM, Johnston RL. Formula choice: Hoffer Q, Holladay 1, or SRK/T and refractive outcomes in 8108 eyes after cataract surgery with biometry by partial coherence interferometry. J Cataract Refract Surg. 2011;37(1):63-71.
Olsen T, Hoffmann P. C constant: new concept for ray tracing-assisted intraocular lens power calculation. J Cataract Refract Surg. 2014;40(5):764-773.
Holladay JT, Lynn MJ, Waring GO 3rd, Gemmill M, Keehn GC, Fielding B. The relationship of visual acuity, refractive error, and pupil size after radial keratotomy. Arch Ophthalmol. 1991;109(1):70-76.
Villegas EA, Alcón E, Artal P. Minimum amount of astigmatism that should be corrected. J Cataract Refract Surg. 2014;40(1):13-19.
Kane JX, Van Heerden A, Atik A, Petsoglou C. Intraocular lens power formula accuracy: Comparison of 7 formulas. J Cataract Refract Surg. 2016;42(10):1490-1500.
Melles RB, Holladay JT, Chang WJ. Accuracy of Intraocular Lens Calculation Formulas. Ophthalmology. 2018;125(2):169-178.
Darcy K, Gunn D, Tavassoli S, Sparrow J, Kane JX. Assessment of the accuracy of new and updated intraocular lens power calculation formulas in 10 930 eyes from the UK National Health Service. J Cataract Refract Surg. 2020;46(1):2-7.
Savini G, Hoffer KJ, Balducci N, Barboni P, Schiano-Lomoriello D. Comparison of formula accuracy for intraocular lens power calculation based on measurements by a swept-source optical coherence tomography optical biometer. J Cataract Refract Surg. 2020;46(1):27-33.
Pereira A, Popovic MM, Ahmed Y, et al. A comparative analysis of 12 intraocular lens power formulas. Int Ophthalmol. 2021;41(12):4137-4150.
Ryu S, Jun I, Kim TI, Kim EK, Seo KY. Accuracy of the Kane Formula for Intraocular Lens Power Calculation in Comparison with Existing Formulas: A Retrospective Review. Yonsei Med J. 2021;62(12):1117-1124.
Chang P, Qian S, Wang Y, et al. Accuracy of new-generation intraocular lens calculation formulas in eyes with variations in predicted refraction [published online ahead of print, 2022 Jul 8]. Graefes Arch Clin Exp Ophthalmol. 2022;10.
Kothari SS, Reddy JC. Recent developments in the intraocular lens formulae: An update [published online ahead of print, 2022 Jul 1]. Semin Ophthalmol. 2022;1-8.
Haigis W. Occurrence of erroneous anterior chamber depth in the SRK/T formula. J Cataract Refract Surg. 1993;19(3):442-446.
Taroni L, Hoffer KJ, Barboni P, Schiano-Lomoriello D, Savini G. Outcomes of IOL power calculation using measurements by a rotating Scheimpflug camera combined with partial coherence interferometry. J Cataract Refract Surg. 2020;46(12):1618-1623.
Gökce SE, Montes De Oca I, Cooke DL, Wang L, Koch DD, Al-Mohtaseb Z. Accuracy of 8 intraocular lens calculation formulas in relation to anterior chamber depth in patients with normal axial lengths. J Cataract Refract Surg. 2018;44(3):362-368.
Hogarty DT, Mackey DA, Hewitt AW. Current state and future prospects of artificial intelligence in ophthalmology: a review. Clin Exp Ophthalmol. 2019;47(1):128-139.
Dutt S, Sivaraman A, Savoy F, Rajalakshmi R. Insights into the growing popularity of artificial intelligence in ophthalmology. Indian J Ophthalmol. 2020;68(7):1339-1346.
Nemeth G, Kemeny-Beke A, Modis L Jr. Comparison of accuracy of different intraocular lens power calculation methods using artificial intelligence. Eur J Ophthalmol. 2022;32(1):235-241.
a href="https://www.escrs.org/about-escrs/registries/eurequo/wp-content/uploads/2020/02/EUREQUO_Annual_Report2018_final.pdf">https://www.escrs.org/about-escrs/registries/eurequo/wp-content/uploads/2020/02/EUREQUO_Annual_Report2018_final.pdf
Carmona-González D, Castillo-Gómez A, Palomino-Bautista C, et al. Comparison of the accuracy of 11 intraocular lens power calculation formulas. Eur J Ophthalmol. 2021;31(5):2370-2376
Hoffer KJ, Savini G. Update on Intraocular Lens Power Calculation Study Protocols: The Better Way to Design and Report Clinical Trials. Ophthalmology. 2021;128(11):e115-e120.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Comparing the accuracy of different intraocular lens power calculation formulas using IOL-Master 700

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Results

Discussion

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1