CC BY
5NAUKI MEDYCZNE I NAUKI O ZDROWIU | МЕДИЦИНСКИЕ НАУКИ
PREDICTING THE REFRACTIVE OUTCOME AFTER CATARACT SURGERY, COMPARING SRKII, SRK/T AND HAIGIS
Davis Rascevskis* Eriks Elksnis*,** Ilze Lace*,** Guna Laganovska *,** Riga Stradins University, Riga, Latvia.* Departmant of Ophthalmology, Pauls Stradins Clinical University Hospital, Riga, Latvia.**
Introduction. Prospective study was made to evaluate and compare predictability and accuracy of three IOL power calculation formulas (SRK/T, SRK II and Haigis) for IOL power calculation.
The postoperative refractive expectations of patients having cataract surgery have been increased due to the advances in technology. Therefore, accurate intraocular lens (IOL) power calculation is very important to attain the postoperative target refraction. [6] Results between each formula could not differ, but frequently there are patients with a radical difference in calculations.
Methods. Prospective study included 25 eyes of 25 patients who underwent phacoemulsification with IOL implantation. Postoperative refraction and refraction predicted by the SRK II, SRK/T and Haigis formulas were analyzed. The mean estimation error (EE), mean absolute estimation error (AEE) and the percentage of eyes within target refraction (EWTR) (± 0.50 D and ± 1.00 D) for all three formulas were compared. Analysis was repeated in three groups with three subgroups in each group. These groups formed based on axial length (AL) (group 1.1: < 23 mm, group 1.2: 23 - 24 mm, group 1.3: > 24 mm), keratometry (K value) (group 2.1: < 43 D, group 2.2: 43 - 45 D, group 2.3: > 45 D) and anterior chamber depth (ACD) (group 3.1: < 3 mm, group 3.2: 3 - 3.5 mm, group 3.3: > 3.5 mm).
Results. In the overall study group, the smallest mean AEE (0.33 ± 0.17) was provided by the SRK/T formula. The highest percentage of EWTR ± 0.50 D and ± 1.00 D was also found by using SRK/T (80% and 100%).
SRK/T provided the smallest mean AEE (0.55 ± 0.25 and 0.26 ± 0.18) for groups 1.1 (n = 8) and 1.3 (n = 7), however, there was no statistically significant difference between all three formulas in group 1.3 (P = 0.22). In group 1.2 (n = 10), the smallest mean AEE (0.37 ± 0.26) was obtained using SRK II.
Haigis provided the smallest mean AEE (0.31 ± 0.09) in group 2.1 (n = 4). In both, group 2.2 (n = 11) and 2.3 (n = 10) the smallest mean AEE was found by using SRK/T (0.43 ± 0.18) and 0.21 ± 0.09).
In all subgroups of group 3 (group 3.1 (n = 4), group 3.2 (n = 14) and group 3.3 (n = 7), SRK/T showed the smallest mean AEE (0.26 ± 0.12), (0.40 ± 0.19) and (0.23 ± 0.09), however, no significant difference was found between all three formulas (P
= 0.17, P = 0.24 and P = 0.31)
SRK/T provided the highest percentage of EWTR ±0.50D and ±1.00D in all subgroups, except 2.2, where Haigis showed better percentage of EWTR ±0.50D. (54% < 63%)
Conclusions. Better results can be obtained using SRK/T formula in almost every eye, except SRK II formula may be preferred in eyes with moderate AL and Haigis formula in eyes with K value under 43D.
Keywords. IOL power, calculation formula, Zeiss IOLMaster Introduction
The final refractive result is dependent upon the accuracy of the biometric data and its appropriate use in the relevant calculations. [3] The postoperative refractive expectations of patients having cataract surgery have been increased due to the advances in technology. Therefore, accurate intraocular lens (IOL) power calculation is very important to attain the postoperative target refraction. [6] Biometry, as well as formula use, is a vital part of the cataract process because the effective IOL power largely controls the final refractive error. That refractive error is the 'take home' element of the surgery, and will remain virtually unchanged for the rest of the patient's life. [10] An incorrect IOL is a leading cause of successful litigation in ophthalmology. [9] Aim of the study
The aim of this research was to analyze predictability and accuracy for IOL power calculation, evaluate and compare estimation error of three IOL power calculation formulas (SRK/T, SRK II and Haigis). Materials and methods
In this prospective study, which was conducted in Pauls Stradins Clinical University Hospital, Ophthalmology clinic between September 15 and December 15, 2015, included 25 eyes of 25 patients who underwent phacoemulsification with IOL implantation. Patients with good quality biometry measurements and best - corrected visual acuities (BCVA) greater than 32/40 after cataract surgery were included in this study. To assess axial length, corneal power and anterior chamber depth each patient underwent biometry measurement using Carl Zeiss IOL Master v5 optical biometer. It is more accurate and reproducible than contact ultrasound in providing accurate AL measurements. The risk of corneal lesion and transmission of infection from patient to patient are
also excluded, as it is non-contact technique. [1] Patients were divided into three groups with three subgroups in each group. These groups formed based on axial length (AL), keratometry (K value) and anteriror chamber depth (ACD). First group was made according to value of AL and divided in shorter eyes -under 23 mm, normal eyes - 23 to 24 mm and longer eyes -above 24 mm. Second group showed corneal power in diopters distributed in subgroups - under 43 D, 43 to 45 D and above 45 D. Subgroups of anterior chamber depth under 3 mm, 3 to 3.5 mm and above 3.5 mm were included in third group.
Formulas SRK II, SRK/T and Haigis were used to calculate IOL power. The aim in IOL power selection was a value that would provide a postoperative refraction nearest to zero (plano), staying on the side of myopia.
All operations were performed under topical anaesthesia by one experienced surgeon. A standard phacoemulsification was made through a temporal clear corneal incision. The monoblock foldable hydrophobic acrylic IOL with A-constant of 118.4 was inserted into the capsular bag using an injector system.
After first postoperative month, all patients underwent ophthalmological examination to evaluate uncorrected visual acuity (UCVA) and BCVA.
The estimation error (EE) was defined as the difference between the postoperative refractive error and the preoperative refractive error predicted by IOL Master biometer using different formulas. The absolute estimation error (AEE) was defined as the absolute value of the EE.
To compare differences in mean EE and mean AEE between formulas in the overall study group, as well as all groups and subgroups, the Friedman Anova Test (comparing multiple related samples) was used. Data were analyzed using SPSS statistical program version 21.0 and Microsoft Office Excel 2007. The obtained results were compared using The Friedman Anova (comparing multiple related samples) test. The p value < 0.05 was accepted as statistically valid.
Results
Twenty five eyes of 25 patients were included in the study. The mean age of patients was 74.44 ± 6.32 years (range, 62 - 85 years). 56% females and 44% males. The mean AL was 23.51 ± 1.19 mm (range, 22.02 - 27.56 mm). The mean K value was 44.6 ± 1.11 D (range, 42.59 - 45.99 D). The mean ACD was 3.24 ± 0.31 mm (range, 2.6 - 3.83 mm). Characteristics of patients are shown in Table 1.
In the overall study group [Table 2], the smallest mean AEE (0.33 D ± 0.17) was provided by the SRK/T formula. The highest percentage of EWTR ± 0.50 D and ± 1.00 D was also found by using SRK/T (80% and 100%).
SRK/T provided the smallest mean AEE (0.55 D ± 0.25 and 0.26 D ± 0.18) for groups 1.1 (n = 8) and 1.3 (n = 7), however, there was no statistically significant difference between all three formulas in group 1.3 (P = 0.22) [Table 3, Table 5]. In group 1.2 (n = 10), the smallest mean AEE (0.37 D ± 0.26) was obtained using SRK II [Table 4].
Haigis provided the smallest mean AEE (0.31 D ± 0.09) in group 2.1 (n = 4) [Table 6]. In both, group 2.2 (n = 11) and 2.3 (n = 10) the smallest mean AEE was found by using SRK/T (0.43 D ± 0.18 and 0.21 D ± 0.09) [Table 7, Table 8].
In all subgroups of group 3 (group 3.1 (n = 4), group 3.2 (n = 14) and group 3.3 (n = 7), SRK/T showed the smallest mean AEE (0.26 D ± 0.12), (0.40 D ± 0.19) and (0.23 D ± 0.09), however, no significant difference was found between all three formulas (P = 0.17, P = 0.24 and P = 0.31) [Table 9, Table 10, Table 11].
SRK/T provided the highest percentage of EWTR ± 0.50 D and ± 1.00 D in all subgroups, except 2.2, where Haigis showed better percentage of EWTR ± 0.50 D. (54% < 63%)
Discussion
It has to be acknowledged that calculation of IOL power is not absolute due to the large number of individual variations in human eyes and no single formula has been found to be useful in all circumstances. [7] With current IOL formulas there is great accuracy on IOL power calculations, provided that the measurements they require are precise. [2] It goes without saying that high quality results expected from excellent surgery and high-end refractive devices will be spoiled or even made impossible by erroneous IOL powers calculated by the wrong formula from low quality measurement data. Surgeons should check the plausibility of all measured data, for instance ensuring that axial length correlates with refraction and that corneal radii measurements correlate with astigmatism. Most surgeons have developed their own plan for deciding on the clinical needs of their patients. It has often been recommended to aim patients for mild postoperative myopia (- 0.5 to - 1.5 D) so if the error is on the plus side, they will be emmetropic and if on the minus side, they will have reading vision. This is necessary because of the larger range of IOL power errors generally experienced. [1]
There are several studies assessing different IOL power calculation formulas in different eyes. In a recent study evaluating refractive outcome of IOL Master, there was no significant difference in AEE of different formulas for AL from 21.50 to 23.50 mm. [4] Roh YR et al. have reported that calculation with Haigis had more predictable outcomes for eyes with an AL shorter than 22.00 mm. [8] Another study has reported that Haigis was leading with success in eyes with AL shorter than 22.00 mm. Besides it was significantly more accurate than other formulas in patients with an ACD less than 2.40 mm. [5] In the study of Aristodemou et al., lowest AEE was obtained using SRK/T formula for eyes with AL of 27.00 mm and longer. [4] Wang et al. have found that in eyes with an AL longer than 25.00 mm, Haigis was more accurate than other formulas. [11]
In our entire study group, with statistically significant difference comparing with other two formulas, SRK/T provided the smallest mean AEE (0.33 D ± 0.17). Accordingly, SRK/T had the highest percentage of eyes within target refraction, showing 80% in ± 0.50 D range and 100% in ± 1.00 D range.
Conclusion
To prevent IOL power errors, simple steps and attention to detail can be very useful, as well as calculation by the right formula from great quality measurements. Based on data used in our study, better results can be obtained using SRK/T formula in almost every eye, except SRK II formula may be preferred in eyes with moderate AL and Haigis formula in eyes with K value under 43 D.
Annex:
Table 1
Characteristics of patients and preoperative measurements.
Parameter Mean ± SD Range
Age, years 74.44 ± 6.32 62 - 85
Sex -male, n (%) 11 (44%) -
Sex - female, n (%) 14 (56%) -
Laterality -right eye, n (%) 12 (48%) -
Laterality -left eye, n (%) 13 (52%) -
Axial length, mm 23.51 ± 1.19 22.02 - 27.56
K value, D 44.6 ± 1.11 42.59 - 45.99
ACD, mm 3.24 ± 0.31 2.6 - 3.83
Corneal astigmatism, D -0.68 ± 0.45 -1.9 - (-0.2)
IOL power, D 20.8 ± 3.49 8.5 - 25
SD - standard deviation; K - mean corneal power; D -diopters; IOL - intraocular lens; ACD - anterior chamber depth.
Table 2
Comparison of mean absolute estimation error (AEE), estimation error (EE), and percentage of eyes within target refraction
(EWTR) between three formulas in overall study group (n = 25).
SRK II SRK/T Haigis P* value
Mean AEE ± SD (range), D 0.52 ± 0.45 (0 - 1.65) 0.33 ± 0.17 (0.06 - 0.64) 0.56 ± 0.34 (0.03 - 1.36) 0.029
Mean EE ± SD (range), D 0.19 ± 0.66 (-1.65 - 1.40) 0.02 ± 0.38 (-0.64 - 0.61) -0.55 ± 0.35 (-1.36 - 0.07) 0.000
EWTR ± 0.50 D (%) 60 80 56
EWTR ± 1.00 D (%) 88 100 84
* The Friedman Anova (comparing multiple related samples) Test.
Table 3
Comparison of mean absolute estimation error (AEE), estimation error (EE), and percentage of eyes within target refraction (EWTR) between three formulas in group 1.1 (AL < 23 mm, n = 8).
SRK II SRK/T Haigis P* value
Mean AEE ± SD (range), D 0.84 ± 0.38 (0.46 - 1.40) 0.32 ± 0.19 (0.08 - 0.61) 0.55 ± 0.25 (0.11 - 0.87) 0.021
Mean EE ± SD (range), D 0.84 ± 0.38 (0.46 - 1.40) 0.16 ± 0.36 (-0.34 - 0.61) -0.55 ± 0.25 (-0.87 - (-0.11)) 0.000
EWTR ± 0.50 D (%) 37 75 50
EWTR ± 1.00 D (%) 75 100 100
* The Friedman Anova (comparing multiple related samples) Test.
Table 4
Comparison of mean absolute estimation error (AEE), estimation error (EE), and percentage of eyes within target refraction (EWTR) between three formulas in group 1.2 (AL 23 - 24 mm, n = 10).
SRK II SRK/T Haigis P* value
Mean AEE ± SD (range), D 0.37 ± 0.26 (0 - 0.70) 0.39 ± 0.20 (0.06 - 0.64) 0.76 ± 0.35 (0.35 - 1.36) 0.045
Mean EE ± SD (range), D -0.1 ± 0.46 (-0.65 - 0.70) -0.18 ± 0.41 (-0.64 - 0.50) -0.76 ± 0.35 (-1.36 - (-0.35)) 0.001
EWTR ± 0.50 D (%) 60 70 30
EWTR ± 1.00 D (%) 100 100 60
* The Friedman Anova (comparing multiple related samples) Test.
Table 5
Comparison of mean absolute estimation error (AEE), estimation error (EE), and percentage of eyes within target refraction (EWTR) between three formulas in group 1.3 (AL > 24 mm, n = 7).
SRK II SRK/T Haigis P* value
Mean AEE ± SD (range), D 0.35 ± 0.57 (0.05 - 1.65) 0.25 ± 0.05 (0.20 - 0.33) 0.26 ± 0.18 (0.03 - 0.50) 0.228
Mean EE ± SD (range), D -0.11 ± 0.68 (-1.65 - 0.20) 0.16 ± 0.21 (-0.31 - 0.33) -0.24 ± 0.22 (-0.50 - 0.07) 0.006
EWTR ± 0.50 D (%) 85 100 100
EWTR ± 1.00 D (%) 85 100 100
* The Friedman Anova (comparing multiple related samples) Test.
Table 6
Comparison of mean absolute estimation error (AEE), estimation error (EE), and percentage of eyes within target refraction (EWTR) between three formulas in group 2.1 (K value < 43 D, n = 4).
SRK II SRK/T Haigis P* value
Mean AEE ± SD (range), D 0.44 ± 0.28 (0.20 - 0.70) 0.36 ± 0.15 (0.23 - 0.50) 0.31 ± 0.09 (0.23 - 0.40) 0.751
Mean EE ± SD (range), D 0.44 ± 0.28 (0.20 - 0.70) 0.36 ± 0.15 (0.23 - 0.50) -0.31 ± 0.09 (-0.40 - (-0.23)) 0.05
EWTR ± 0.50 D (%) 50 100 100
EWTR ± 1.00 D (%) 100 100 100
* The Friedman Anova (comparing multiple related samples) Test.
Table 7
Comparison of mean absolute estimation error (AEE), estimation error (EE), and percentage of eyes within target refraction (EWTR) between three formulas in group 2.2 (K value 43 - 45 D, n = 11).
SRK II SRK/T Haigis P* value
Mean AEE ± SD (range), D 0.59 ± 0.61 (0.05 - 1.65) 0.43 ± 0.18 (0.2 - 0.64) 0.62 ± 0.44 (0.03 - 1.36) 0.534
Mean EE ± SD (range), D -0.003 ± 0.87 (-1.65 - 1.40) -0.25 ± 0.48 (-0.64 - 0.61) -0.61 ± 0.47 (-1.36 - 0.07) 0.002
EWTR ± 0.50 D (%) 54 54 63
EWTR ± 1.00 D (%) 72 100 63
* The Friedman Anova (comparing multiple related samples) Test.
Table 8
Comparison of mean absolute estimation error (AEE), estimation error (EE), and percentage of eyes within target refraction (EWTR) between three formulas in group 2.3 (K value > 45 D, n = 10).
SRK II SRK/T Haigis P* value
Mean AEE ± SD (range), D 0.47 ± 0.30 (0.00 - 0.92) 0.21 ± 0.09 (0.06 - 0.34) 0.58 ± 0.23 (0.11 - 0.87) 0.007
Mean EE ± SD (range), D 0.32 ± 0.46 (-0.26 - 0.92) -0.05 ± 0.24 (-0.34 - 0.32) -0.58 ± 0.23 (-0.87 - (-0.11)) 0.001
EWTR ± 0.50 D (%) 70 100 30
EWTR ± 1.00 D (%) 100 100 100
* The Friedman Anova (comparing multiple related samples) Test.
Table 9
Comparison of mean absolute estimation error (AEE), estimation error (EE), and percentage of eyes within target refraction (EWTR) between three formulas in group 3.1 (ACD < 3 mm, n = 4).
SRK II SRK/T Haigis P* value
Mean AEE ± SD (range), D 0.53 ± 0.11 (0.46 - 0.70) 0.26 ± 0.12 (0.08 - 0.34) 0.55 ± 0.37 (0.11 - 0.87) 0.174
Mean EE ± SD (range), D 0.53 ± 0.11 (0.46 - 0.70) -0.06 ± 0.32 (-0.34 - 0.32) -0.55 ± 0.37 (-0.87 - (-0.11)) 0.018
EWTR ± 0.50 D (%) 75 100 50
EWTR ± 1.00 D (%) 100 100 100
* The Friedman Anova (comparing multiple related samples) Test.
Table 10
Comparison of mean absolute estimation error (AEE), estimation error (EE), and percentage of eyes within target refraction (EWTR) between three formulas in group 3.2 (ACD 3 - 3.5 mm, n = 14).
SRK II SRK/T Haigis P* value
Mean AEE ± SD (range), D 0.60 ± 0.43 (0.05 - 1.40) 0.40 ± 0.19 (0.17 - 0.64) 0.66 ± 0.34 (0.23 - 1.36) 0.245
Mean EE ± SD (range), D 0.30 ± 0.69 (-0.65 - 1.40) 0.02 ± 0.45 (-0.64 - 0.61) -0.66 ± 0.34 (-1.36 - (-0.23)) 0.000
EWTR ± 0.50 D (%) 42 64 42
EWTR ± 1.00 D (%) 85 100 71
* The Friedman Anova (comparing multiple related samples) Test.
Table 11
Comparison of mean absolute estimation error (AEE), estimation error (EE), and percentage of eyes within target refraction (EWTR) between three formulas in group 3.3 (ACD > 3.5 mm, n = 7).
SRK II SRK/T Haigis P* value
Mean AEE ± SD (range), D 0.33 ± 0.58 (0.00 - 1.65) 0.23 ± 0.09 (0.06 - 0.33) 0.34 ± 0.23 (0.03 - 0.67) 0.317
Mean EE ± SD (range), D -0.20 ± 0.65 (-1.65 - 0.20) 0.07 ± 0.25 (-0.31 - 0.33) -0.32 ± 0.27 (-0.67 - 0.07) 0.006
EWTR ± 0.50 D (%) 85 100 85
EWTR ± 1.00 D (%) 85 100 100
* The Friedman Anova (comparing multiple related samples) Test.
References
1. Ashok Garg. Intraocular Lens (IOL) Power Calculations // Mastering Intraocular Lenses (IOLs) (Principles, Techniques and Innovations). Jaypee Brothers Medical Publishers Ltd; 2007; 17-106.
2. Myron Yanoff, Jay S. Duker. Refractive surgery // Ophthalmology 4th edition. Elsevier Saunders; 2014; 129.
3. Myron Yanoff, Jay S. Duker. Patient workup for cataract surgery // Ophthalmology 4th edition. Elsevier Saunders; 2014; 341.
4. Aristodemou P, et al. Formula choise: 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(3): 63-70.
5. Eom Y, et al. Comparison of Hoffer Q and Haigis formulae for intraocular lens power calculation according to the anterior chamber depth in short eyes // Am J Ophthalmol 2014; 157: 818-24.
6. Kaya F, et al. Comparison of different formulas for intraocular lens power calculation using a new optical biometer // J Fr Ophtalmol 2015; 38(8): 717-722.
7. R Sheard. Optimising biometry for cataract surgery // Cambridge Ophthalmological Symposium, Eye 2014; 28: 118125.
8. Roh YR, et al. Intraocular lens power calculation using IOL Master and various formulas in short eyes // Korean J Ophthalmol 2011; 25(3): 151-5.
9. Simon JW, Ngo Y, Khan S, et al. Surgical confusions in ophthalmology // Arch Ophthalmol 2007; 125(11): 1515-1522.
10. Spaeth, George L., MD, FACS. Preoperative assessment including biometry // Ophthalmic Surgery: Principles and Practice 4th edition 2012; 35-39.
11. Wang JK, Hu CY, Chang SW. Intraocular lens power calculation using the IOL Master and various formulas in eyes with long axial length // J Cataract Refract Surg 2008; 34(2): 262-7.
