

ORIGINAL ARTICLE 

Year : 2018  Volume
: 2
 Issue : 2  Page : 8589 

Comparison of SRK/T and Haigis formulae in the prediction of refractive outcome after phacoemulsification
Mona N Mansour, Rehab M Kamel, Hanan S Hegazy
Department of Ophthalmology, Faculty of Medicine (for Girls), AlAzhar University, Cairo, Egypt
Date of Submission  08Jul2018 
Date of Acceptance  26Jul2018 
Date of Web Publication  21Sep2018 
Correspondence Address: Mona N Mansour Department of Ophthalmology, Faculty of Medicine (for Girls), AlAzhar University, Nasr City, 11754 Cairo Egypt
Source of Support: None, Conflict of Interest: None
DOI: 10.4103/sjamf.sjamf_21_18
Purpose To compare the accuracy of SRK/T and Haigis formulae used for intraocular lens (IOL) power calculation by a partial coherence interferometer in patients undergoing phacoemulsification surgery. Patients and methods A prospective interventional clinical study included 70 eyes of 60 patients, who underwent uncomplicated phacoemulsification with IOL implantation from October 2015 to December 2017. Preoperative axial length (AL), corneal curvature (keratometry), and preoperative anterior chamber depth (preoperative ACD) were measured using Nidek ALscan optical biometer and the IOL power was determined using both SRK/T and Haigis formulae. The difference between the predicted value and the postoperative spherical equivalent was calculated for both the formulae by the end of the followup (3 months postoperatively). Results The mean errors of the two formulae were SRK/T: −0.225±0.61 D and Haigis: 0.171±0.68 D; the mean absolute errors of the two formulae were 0.534±0.36 and 0.533±0.44 D, respectively. There was no statistically significant difference between the mean error of the two formulas used in the overall performance, but was significant in eyes with an AL of more than 25 mm. The proportion of patients having a prediction error within ±0.50 D of SRK/T formula (54.29%) was comparable to those of Haigis (55.71%) and the prediction errors within ±1.0 D were 87.14 and 85.71%, respectively. There is a weak correlation between the mean AL, keratometry and the Haigis–SRK/T prediction differences (r^{2}=0.273). Conclusion The calculation of IOL power using SRK/T and Haigis formulae resulted in an accurate postoperative refraction. In long AL subcategory, the mean absolute error of Haigis was less compared with the SRK/T formula.
Keywords: axial lengthscan, biometry, Haigis formula, intraocular lens power, partial coherence interferometer, SRK/T formula
How to cite this article: Mansour MN, Kamel RM, Hegazy HS. Comparison of SRK/T and Haigis formulae in the prediction of refractive outcome after phacoemulsification. Sci J AlAzhar Med Fac Girls 2018;2:859 
How to cite this URL: Mansour MN, Kamel RM, Hegazy HS. Comparison of SRK/T and Haigis formulae in the prediction of refractive outcome after phacoemulsification. Sci J AlAzhar Med Fac Girls [serial online] 2018 [cited 2018 Dec 9];2:859. Available from: http://www.sjamf.eg.net/text.asp?2018/2/2/85/241826 
Introduction   
Accurate biometry is essential for achieving good surgical outcomes and patient satisfaction after cataract surgery. Highaccuracy technologies for ocular axial length (AL) measurements using partial coherence interferometry, along with precise keratometry, and the new prediction formulae have provided improvements on IOL calculation [1].
Formulae for IOL power calculation had past four generations, the first generation formulae were theoretical and based on the same fundamental constant with no respect to anterior chamber depth (ACD). The second generation was designed by combining linear regression analysis and stepwise adjustment for long and short eyes according to ACD [2].
Thirdgeneration formulas, such as the Hoffer Q, Holladay 1, and SRK/T recognize that postoperative ACD varies with AL and corneal curvature [3]. The Haigis formula, a fourthgeneration formula, uses three constants to calculate the effective lens position. It takes into account the preoperative ACD and uses three constants (a0, a1, and a2), which are analogous to surgeon factor, ACD and AL, respectively.
Different studies to evaluate the predictive accuracy of various IOL power calculation formulae with different results [4],[5].
This study was conducted to evaluate the accuracy of the SRK/T and Haigis formulae for IOL power calculation in patients undergoing cataract surgery in eyes with different ALs.
Patients and methods   
A prospective, comparative study was carried out at AlZahraa University Hospital from October 2015 to December 2017 and included 70 eyes of 60 patients. The study adhered to the tenets of the Declaration of Helsinki and was approved by the ethics board of AlAzhar University and an informed written consent was taken from each participant in the study. Patients were scheduled for phacoemulsification and routine preoperative ocular examination was done.
Inclusion criteria were patients with cataract of any type, normal anterior and posterior segment; and uneventful surgery with ‘in the bag’ monofocal IOL implantation in all patients. Patients with a history of intraocular operation, inflammation, retinopathy, severe corneal degeneration, corneal opacity, vitreous pathology, developmental and acquired macular diseases were excluded. Biometry by ALscan (Nidek Co. Ltd, Gamagori, Japan) was used in all eyes. AL, ACD, and corneal refractive power were measured by the ALscan. Calculation of the IOL power to be implanted was done using SRK/T and Haigis formulae and the Aconstant was maintained at 118. The predicted refraction value targeted myopia (−0.5 to −1). All patients underwent uneventful phacoemulsification surgery using the same standard phacoemulsification protocol (Abbot Medical Optics Inc., CA, USA). The Phacochop technique was the standard phacoemulsification technique in all patients, through a 3.2 mm clear corneal tunnel incision at the 12 o’clock position and a 5.0–5.5 mm capsulorhexis. Phacoemulsification was followed by inthebag implantation of the foldable IOL, and surgical wounds were hydrated with no sutures. Subconjuctival gentamicin and dexamethasone injections were given at the end of surgery. Surgery was performed in all cases by the same surgeon to avoid errors of personalized surgeon factor. All the patients were examined at 1 day, 1 week, 1 month, and 3 months postoperatively. The actual postoperative spherical equivalence was recorded 3 months following the surgery. The mean error (ME) is calculated as: formula predicted postoperative refraction − actual postoperative refractive error by the end of followup. The absolute value of ME is the mean absolute errors (MAEs).
Data were statistically described in terms of mean, SD, median, and range when appropriate. Comparison between the two studied equations was done using paired ttest. Multivariate linear regression analysis was used to test for the preferential effect of the independent variable(s) on MAE for each equation and the deviation between them. A P value of less than 0.05 was considered statistically significant. All statistical calculations were done using the computer program IBM statistical package for the social sciences (IBM Corp., Armonk, New York, USA) release 22 for Microsoft Windows.
Results   
A total of 60 patients and 70 eyes were included in this study, who underwent uneventful cataract surgeries (phacoemulsification) with implantation of foldable IOL. The age of the patients ranged from 14 to 70, mean (51.69±11.9) years. The mean AL and mean keratometry reading are shown in [Table 1]. [Table 1] also shows the ME and MAE for both the SRKT and Haigis formulae.
The majority of patients (87.14% of SRK/T formula and 85.71% of Haigis formula) were within 1 D of the predicted refractive error ([Table 2]).  Table 2 Percentage of cases predicted to within 0.25, 0.50, 1.00, and 1.50 D for SRKT and Haigis formulae
Click here to view 
The SRK/T formula caused nearly the same ME as the Haigis formula, but tend to be higher in myopes (35 eyes with AL >25 mm) ([Table 3]).
In this study, there was no statistically significant difference between the MAEs of the two formulae used in the overall performance, but is significant in the myopic subcategory (35 eyes with an AL >25 mm) ([Table 4]).
To verify the biometric variables that could be involved in the discrepancy between the Haigis and SRK/T predictions multivariate linear regression analysis was used to test for the preferential effect of the independent variables (AL and keratometry) on MAE for each equation and the intraindividual difference between both formulae predictions ([Table 5]).  Table 5 Correlations between preoperative biometric data and the mean absolute error for each equation and the deviation between Haigis and SRK/T predictions
Click here to view 
Although the correlation between the mean AL, keratometry, and the Haigis–SRK/T prediction differences is significant (P=0.00), it is of weak correlation (r^{2}=0.273). The linear regression parameter for variable AL=−0.056±0.011 which indicates that a decrease of 1 mm on AL implies an increase of 0.056 D on the difference between formulas prediction, for variable K=0.067±0.021 which indicates that an increase of 1 D on K implies an increase of 0.067 D on the difference between formulae prediction. Other variables showed nonsignificant correlations, and no significant multivariate linear regression parameters ([Table 4]).
Discussion   
According to the WHO, cataract causes 27–45 million cases of blindness in the world. Cataract surgery is the most common eye surgery in the world [6].
The ability to predict postoperative refraction with accuracy is important to achieve successful result of cataract surgery with IOL implantation. Errors in predicted refraction after IOL implantation are mostly a result of AL measurement error. With the advent of partial coherence interferometry, the AL determination has become more accurate. Therefore, a need has arisen for even more precise IOL formulae [1].
Different studies have evaluated the predictive accuracy of various IOL power calculation formulae; Wang and Chang [7] showed similar performance of the Haigis, Hoffer Q, Holladay 1, and SRK/T in medium eye length. In another study by Wang et al. [8], SRK/T and Haigis performed equally well and outperformed the Hoffer Q and Holladay 1 in 34 eyes between 25 and 28 mm. Contrary to Moschos et al. [4] and Roh et al. [9] who reported that the Haigis formula was more accurate than the other formulae, Kapadia et al. [10], Maclaren et al. [11], Aristodemou et al. [12], and ElNafees et al. [2] reported that the SRK/T formula was more accurate than the other formulae in long eyes.
In this study, we found that similar performance of SRK/T and Haigis (MAE: 0.534 for SRK/T and 0.533 for Haigis), with a little tendency of myopic shift for SRK/T formula (ME=−0.225) and hyperopic shift for the Haigis formula (ME=0.171) ([Table 1]). Doshi et al. [13] in their study also reported that Haigis formula had a little tendency for hyperopic results, in contrarst to Dalto et al. [5] who found a significant myopic shift using the Haigis formula and a hyperopic shift with the SRK/T formula.
In this study, we found that about half of cases being within ±0.50 D of predicted refraction (54.29% for SRK/T formula, 55.71% for Haigis formula) and the majority were within ±1.00 D (87.14% for SRK/T formula, 85.71% for Haigis formula) ([Table 2]). These results were similar to the study done by Sharma et al. [1] who achieved a prediction accuracy within 1.00 D of 78% for SRK/T formula, 86% for Haigis formula; ElNafees et al. [2] achieved a prediction accuracy of 83.01% (for both SRK/T, Haigis); Kapadia et al. [10] achieved a prediction accuracy 67.85% for SRK/T formula and 68% for Haigis formula. Zhu et al. [14] showed a prediction accuracy of 38.8 and 45.6% (SRK/T and Haigis, respectively) for refractive error within 0.50 D.
In this study, by comparing the two studied equations using paired ttest, there was no statistically significant difference between the MAEs of the two formulae used in the overall performance, but the difference is significant in the myopic subcategory (35 eyes with an AL >25 mm, P=0.012) ([Table 3] and [Table 4]). Doshi et al. [13] in their study found that in eyes with an AL of more than 24.5 (40 eyes), there was no statistically significant difference between MAE of Haigis, Holladay 1, Hoffer Q, and SRK/T formulae (P>0.05). Dalto et al. [5] reported that a difference between SRK/T and Haigis formulae predictability was only found in their study of 108 eyes, when myopia is target for eyes with K less than 43.5.
In this study, the prediction error of SRK/T and Haigis formulae was weakly, nonsignificant, negatively correlated with AL and keratometry (P>0.05), and although the Haigis–SRK/T prediction differences is significantly correlated with both AL and keratometry (P<0.05), it is of weak correlation (r^{2}=0.273) ([Table 5]). The linear regression parameter for variable AL=−0.056±0.011 indicates that a decrease of 1 mm on AL implies an increase of 0.056 D on the difference between formulas prediction and for variable K=0.067±0.021 which indicates that an increase of 1 D on K implies an increase of 0.067 D on the difference between formulae prediction ([Table 5]). This is in contrast to Dalto et al. [5] who reported a strong correlation between preoperative keratometry and the difference between SRK/T and Haigis formulas’ predictability, when myopia is target for eyes with K less than 43.5. Zhu et al. [14] in their study on 103 eyes with an AL of at least 26 mm found that the prediction error of SRK/T formula was positively correlated with AL and corneal astigmatism (P<0.01), while for Holladay and Haigis formulas, in addition to the previous two factors, the errors were also positively correlated with mean corneal curvature (P<0.01).
Kane et al. [15] who studied intraocular lens power formula accuracy with comparison of seven formulas on 3241 patients found that MAE of Haigis and SRK/T formulas were 0.420 and 0.413, respectively, compared with 0.533 and 0.534 in our study that may be attributed to a large number of cases compared with the small number of eyes in our study.
The results of our study has shown that the SRK/T formula has a performance similar to Haigis formula, with no statistically significant difference between the MAEs of the two formulae in the overall performance, but the difference is significant in the myopic subcategory where the Haigis formula has a lower MAE. The inclusion of the measured ACD into the Haigis formula has allowed for potentially increased accuracy [16]. In fact, the Haigis formula differs in a very important way. Rather than simply move a fixed, formulaspecific outcomes curve up (more IOL power recommended) or down (less IOL power recommended), it uses preoperative ACD and AL to predict the ELP and has three constants [a0+(a1×ACD)+(a2×AL)] for the optimization of results. So in the Haigis formula the IOL power optimization is based on three variables. In contrast, the SRK/T uses two biometric parameters (keratometry, AL) to predict the ELP and one constant for the optimization of the results [17].
Conclusion   
The calculation of IOL power using SRK/T and Haigis formulae resulted in an accurate postoperative refraction. The SRK/T formula showed a performance similar to Haigis formula in the overall performance, but a slight lower performance in myopic subgroups.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References   
1.  Sharma R, Maharajan P, Kotta S, Maharajan S. Prediction of refractive outcome after cataract surgery using partial coherence interferometry: comparison of SRK/T and Haigis formulae. Int Ophthalmol 2014; 34:451–455. 
2.  ElNafees R, Moawad A, Kishk H, Gaafar W. Intraocular lens power calculation in patients with high axial myopia before cataract surgery. Saudi J Ophthalmol 2010; 24:77–80. 
3.  Haigis W. The Haigis formula. In: Shammas HJ, editor. Intaocular lens power calculation. Thorfare, NJ: Slack; 2004. pp. PL 41–57. 
4.  Moschos MM, Chatziralli IP, Koutsandrea C. Intraocular lens power calculation in eyes with short axial length. Indian J Ophthalmol 2014; 62:692–694. 
5.  Dalto RF, Ferreira MA, Queiroz W, Coelho RP, Paula JS, Messias A. Haigis and SRKT formulae accuracy for intentional myopic overcorrection. Int Ophthalmol 2018; 38:1459–1463. 
6.  Razmjoo H, Atarzadeh H, Kargar N, Behfarnia M, Nasrollahi K, Kamali A. The comparative study of refractive index variations between Haigis, SRK/T and HofferQ formulas used for preoperative biometry calculation in patients with the axial length >25 mm. Adv Biomed Res 2017; 6:78. 
7.  Wang JK, Chang SW. Optical biometry intraocular lens power calculation using different formulas in patients with different axial lengths. Int J Ophthalmol 2013; 6:150–154. 
8.  Wang JK, Hu CY, Chang SW. Intraocular lens power calculation using the IOLMaster and various formulas in eyes with long axial length. J Cataract Refract Surg 2008; 34:262–267. 
9.  Roh YR, Lee SM, Han YK, Kim MK, Wee WR, Lee JH. Intraocular lens power calculation using IOLMaster and various formulas in short eyes. Korean J Ophthalmol 2011; 25:151–155. 
10.  Kapadia P, Dalal N, Patel N, Chauhan MD. Intraocular lens power calculation formulas in high refractive errors, what to choose and when? Natl J Integr Res Med 2013; 4:33–36. 
11.  Maclaren RE, Natkunarajah M, Riaz Y, Bourne RR, Restori M, Allan BS. Biometry and formula accuracy with intraocular lenses used for cataract surgery in extreme hyperopia. Am J Ophthalmol 2007; 143:920–931. 
12.  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; 37:63–71. 
13.  Doshi D, Limdi P, Parekh N, Gohil N. A comparative study to assess the predictability of different IOL power calculation formulas in eyes of short and long axial length. J Clin Diagn Res 2017; 11: NC01–NC04. 
14.  Zhu XJ, He WW, Du Y, Qian DJ, Dai JH, Lu Y. Intraocular lens power calculation for high myopic eyes with cataract: comparison of three formulas. Zhonghua Yan Ke Za Zhi 2017; 53:260–265. 
15.  Kane JX, van Heerden A, Atik A, Petsoglou C. Intraocular lens power formula accuracy: comparison of 7 formulas. J Cataract Refract Surg 2016; 42:1490–1500. 
16.  Eom Y, Kang SY, Song JS, Kim YY, Kim HM. 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; 157:818–824. 
17.  Faramarzi A, Aghajani A, Ghiasian L. Accuracy of various intraocular lens power calculation formulas in steep corneas. J Ophthalmic Vis Res 2017; 12:385–389. 
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]
