Dr.Anuj Kumar Singal,Dr.Vardhaman Kankaria,Dr.Shrutika Kankaria
AIM
To assess the accuracy of 7 Intraocular lens (IOL) power formulas (Barrett Universal II, SRK/T, Holladay 1, Hoffer Q, Haigis, Hill-Radial Basis Function and Ladas) using optical biometry in patients undergoing Tecnis ZCB00 Implantation
METHODS
The retrospective study included 153 eyes undergoing uneventful phacoemulsification with Tecnis ZCB00 IOL implantation. Refractive outcome was determined using data from IOL master & compared with actual refractive outcome to give predicted error. Eyes were divided into subgroups based on axial length (AL) as short(<22mm), medium(22 to 24.5mm) and long(>24.5mm)
RESULTS
SRK/T had least mean predictive error over entire AL range. Least accurate was Haigis but showed good accuracy in short eyes (p<0.05). For AL <22 mm, Hills RBF and Haigis was most accurate (p<0.05) and for AL>22mm, Barrett’s and Hills RBF was most accurate (p<0.05).
CONCLUSION
The Hill RBF formula is most accurate in eyes with AL <22mm while Barrett’s is the most accurate in AL>22 mm.
KEYWORDS: Axial Length, Barrett’s formula, Hill’s RBF Formula, SRK/T Formula, biometry,
Introduction
With the advancement of cataract surgery and IOL (Intraocular lens) quality, it is not surprising that patients expect an emmetropic result. This end point of cataract surgery involves having a value of both spherical correction and astigmatism to be near zero and is possible with the latest surgical techniques.
Though astigmatism is dependent on subjective factors such as the pre op astigmatism, length of incision, shape of incision, axis of astigmatism and the distance of incision from cornea centre, the spherical error is an objective value and is dependent on factors such as axial length (AL), corneal curvature, anterior chamber (AC) depth, placement of IOL and surgeon factors.
Different IOL formula exist where these values can be incorporated to get a desired IOL power which when placed in place of natural crystalline lens is supposed to give a target postoperative spherical error. Several formula have come up over the years and no ideal formula that would work over different axial length is known. Furthermore, refractive surprises are known to occur in the extreme ends of axial length.
Aristodemou et al had performed the largest IOL study where he incorporated 8108 eyes to assess the accuracy of Hoffer Q, Holladay 1 and SRK/T formula on two different IOLs.[1] In one of the IOLs that he tested, it was found that Hoffer Q was the most accurate for small eyes with axial length less than 21mm. For eyes with medium length axial length (23.5 to 26mm), Holladay 1 was most accurate and for long eyes with axial length more than 26mm, SRK/T was the most accurate formula. In contrast, a similar study by Narvaez et al,[2] found no difference between Holladay 1, Hoffer Q and SRK/T on the 643 eyes that they studied.
Recently some new formulas have come up such as Hill – radial basis function (Hill RBF formula), Barrett’s universal II formula and Ladas super formula that are supposed to be much more accurate than the previous formula. In a recent study by Kane et al[3] that compared seven IOL power formula found that Barrett universal II formula was more accurate predictor of actual postoperative refraction than the other formulas. They further evaluated the three new methods for IOL calculation – Hill-RBF, FullMonte method and Ladas Super formula with Holladay 1 and Barrett Universal II formulas on 3122 eyes and found that Barrett Universal II was the most accurate followed by Hill-RBF formula and Ladas Super formula.[4]
As far as the authors are aware after a detailed pubmed and internet search, this is the first study that compares all the seven latest formulas.
Material and Methods
This retrospective study was conducted at a tertiary level eye hospital in western India. Data of 153 eyes of 112 patients who underwent uneventful phacoemulsification with Tecnis ZCB00 IOL implantation over 18 months was collected and evaluated. The eyes were divided into three groups each depending on axial length as short (≤22.0 mm), medium (>22.0 to <24.5 mm) and long (≥24.5 mm). Tenets of declaration of Helsinki (1964, amended 2013) were adhered to.
The postoperative expected spherical error was obtained readily from a Partial coherence interferometry (PCI) based biometer, the Zeiss IOLMaster 500TM for the SRK/T, Holladay 1, Haigis and Hoffer Q formulas. The same was calculated for Barrett’s Universal II formula from Asia-Pacific Association of Cataract and Refractive Surgeons (APACRS) website[5], the Hill-RBF data from rbf calculator website[6] developed by the creator himself, Dr Warren Hill and the Ladas super formula again from the developer website.[7]
Preoperative biometry measurements were accessed directly from the database of the PCI-based biometer used to assess all patients. The remaining variables were obtained from the electronic medical record. Subjective refraction was performed 14 days postoperatively on all included patients by the double blinded optometric staff.
Any eye without a formal refraction (even with a CDVA better than 6/6) was excluded. No details of lens thickness or white-to-white diameter were used in the study.
The patient’s post-operative spherical refractive error after 2 weeks of surgery was noted. A deviation of the expected spherical error from the actual postoperative spherical error was calculated. Mean absolute error was calculated between the two for different groups in axial length and statistical analysis done using SPSS software 22.0. p value ≤0.05 was taken significant.
Inclusion criteria were uneventful phacoemulsification cataract surgery with in-the-bag insertion of Tecnis ZCB00 IOL (Abbott Medical Optics, AMO Inc., Santa Ana, CA, division of Johnson & Johnson) and preoperative biometry performed using the IOLMaster (version 5.4, Carl Zeiss Meditec AG), PCI based biometer. If patients had bilateral phacoemulsification cataract extraction with insertion of the study IOL, both eyes were chosen for inclusion. Exclusion criteria were incomplete PCI biometry, corneal astigmatism more than 3.0 diopters (D), signs of keratoconus or irregular astigmatism on corneal topography; complicated cataract surgery, additional procedures during cataract surgery, postoperative corrected distance visual acuity (CDVA) worse than 6/12, refraction performed before 14 days postoperatively, postoperative complications, and incomplete documentation.
Kruskal-Wallis test was done to compare the mean absolute errors between different formulas in different axial length groups. Prediction analysis tables were made to observe the spherical errors in ±0.5D to ±2D in 0.5D difference amongst different formulas.
Results
The age of the patients ranged from 29 years to 84 years with a mean age of 59.93 years and a standard deviation of 9.77 years. Males dominated our study with 55.6% eyes being males (n=85). Right eye and left eye were nearly same with right eye being 76.
Overall formulae accuracy
Overall the deviation of the calculated post-operative spherical error was statistically significant with actual post-operative spherical error in all seven formulae but SRK/T demonstrated the maximum correlation (r=0.905). This was followed by Hill RBF, Holladay, Ladas and Barrett’s (TABLE 1) which had very similar correlations (r=0.866 to r=0874). Haigis had the worst correlation value with r=0.786. As there was a heavy overlap in the confidence interval amongst the formulas, there was no statistical significance between the formulas in overall calculation of 153 eyes and all formulas can be considered for IOL calculation taking into consideration the possible refractive error.
A predictive error table involving all 153 eyes was made (Table 2). Hill’s RBF formula showed the maximum eyes having an error upto ±0.5D (64.1%). Both Hill’s RBF and SRK/T formulas showed equal efficacy of having a predictive error of ±1D in 92.2% eyes followed by Barrett’s and Holladay 1 (90.2% eyes). The least predictive error was found to be with Hoffer’s Q in both ±0.50D (58.8%) and ±1.0D group (87.6% eyes)
Short Axial Length eyes (≤22mm)
The mean deviation from expected outcome was seen maximum with Hoffers Q followed by Holladay 1 (Table 3) and least with Hill RBF followed by Haigis. A predictive error table demonstrated that Haigis had the maximum prediction of 33.2% within ±0.5D group while SRK/T and Hill’s RBF had a predictability of 100% in ±1D while the predictability of Haigis decreased to 77.7% (Table 4). These results were limited due to the fact that patients with axial length were very less for a more realistic comparison (n=9)
Medium Axial Length eyes (22mm – 24.5mm)
For eyes with medium axial length (22 to 24.5mm), none of the formulas had a significant mean error. This means that the formulas were accurate in calculation (Table 5). The least mean error was seen with Barrett’s followed by Hill RBF while the maximum error was seen with Haigis (0.11D). Ladas super formula showed a minimal hyperopic mean error (0.03D) though insignificant statistically (p=0.56).
A prediction error analysis table demonstrated Hill’s RBF to be most accurate in both ±0.25D (37.4%) and ±0.5D (65.2%) group while Holladay 1 to be most accurate(91.3%) in ±1D error group (Table 6) followed by SRK/T (90.4%) and Ladas (90.4%).
Longer Axial Length (≥24.5mm)
For longer Axial length (≥24.5mm) also there was no significant mean error in any of the formulas. The least mean error was seen with Hill RBF followed by Barrett’s while the maximum was seen with Hoffer Q (Table 7).
A predictive analysis table showed that Hills’s RBF was much more accurate up to ±0.25D error (55.2% eyes) but Holladay 1 was much more accurate in ±0.5D group followed by Hills RBF again being 100% accurate with an error of up to +/-1D (Table 8). All formula showed good accuracy within 1D limit error but worse performance was seen with Hoffer’s Q (89.7% eyes).
Discussion
Tecnis ZCB00 is a single piece Biconvex, Aspheric UV blocking hydrophobic acrylic surface, square optic design, 6mm diameter intraocular lens designed and manufactured by Abbott Medical Optics, Santa Ana, CA. The recommended A-constant using optical biometry is 119.3, the theoretical AC Depth is 5.4mm and Surgeon factor is 1.68mm.[8]
SRK/T Is a theoretical approach and upgradation to the original SRK formula developed by Donald R. Sanders, John A. Retzlaff and Manus C. Kraff.[9,10] The original SRK formula was a regressive formula and used the following equation to calculate IOL power: P = A – 2.5L – 0.9K, where P is the implant power for emmetropia; L is the axial length (mm); K is the average keratometry (D); and A is a constant specific for each IOL. Over years, it was found that the original formula was best used between 22 mm and 24.5 mm and thus SRK/T came in which optimizes the prediction of postoperative ACD, retinal thickness AL correction, and corneal refractive index.[11]
The Haigis formula utilizes preoperative anterior chamber depth (ACD) and axial length (L) in its linear regression. It also uses three IOL constants for the prediction of the effective lens position (ELP) of the intraocular lens (IOL), ELP ≈ a0 + a1 ACD + a2 L.[12]
The Ladas super formula was developed by Ladas et al using numerical computing environment to create 3-D surfaces of pre-existing IOL formulas such as SRK/T, Hoffer Q, Holladay 1 and Haigis.[13]
Barrett Universal II formula uses a theoretical model eye in which anterior chamber depth is related to axial length and keratometry. A relationship between the A constant and Lens factor is used to determine ACD. It uses location of principle plane of refraction of the IOL and retains it as a relevant variable.[14]
Holladay 1 uses postoperative stabilized refraction value, dioptric power of implanted IOL and preoperative corneal and AL measurements to calculate a personalized surgeon factor.[15]
Hoffer Q utilizes a personalized ACD apart from AL and corneal curvature. This personalized ACD (pACD) is developed by using a series of a particular IOL style. It includes a factor that increases the ACD with increasing AL, a factor that increases the ACD with increasing corneal curvature, a factor that moderates the change in ACD for extremely long and short eyes and a constant added to the ACD.[16]
The Hill-RBF formula employs pattern recognition and multi-dimensional data interpolation. It is self-validating method artificial intelligence programme that uses the radial basis function to calculate IOL power.[17]
One of the oldest comparison study was done by Hoffer et al in 1993 and again in 2000. His analysis concluded that the Hoffer Q formula provided the most reliable results in short eyes (AL < 22.0 mm) while the SRK/T formula was best in long eyes (AL > 26.0 mm).[14,18] and usually both can be used for normal axial length eyes between 22 to 26mm. With the addition of new formulas, Hoffer Q has started becoming obsolete. Infact in our study of 153 eyes, Hoffer Q had the second worst accuracy after Haigis formula. Also Hoffer Q was found to be least accurate for both short eyes of AL <22mm and long eyes with AL >26mm.
Extremely hyperopic eyes with axial length of <22mm create the maximum problem with none of the formulas in our study being accurate. Hills RBF as well as Haigis were the most reliable of all seven formulas in our study. Other studies have also demonstrated the high level of accuracy of the Haigis formula amongst the older formulas in extreme hyperopia.[19,20]
Very few studies have compared the newer formulas for eyes with shorter axial length. A very recent study by Gokce et al in 2017 that evaluated 87 eyes also had very similar results to our study. They also found hills rbf to be statistically more accurate than Hoffer Q.[21]
Thus Hills RBF and Haigis are the only formulas that should be used in present times for short eyes.
Classically all formulas give comparable refractive outcomes for eyes with normal axial length (22mm to 26mm)[18] though some studies have demonstrated the efficiency of Holladay 1 formula over others especially for eyes with slightly longer axial length between 23.50 mm and 25.99 mm[19]. Unfortunately these studies have not compared the recent formulas. Our study found that Barrett’s had the least amount of mean spherical error followed by Hills RBF formula. SRK/T, Ladas and Holladay 1 were also found to give comparable results but with higher mean spherical error. The least accurate formula in this group was found to be Haigis.
Olsen et al in their study on 2043 eyes found no difference between the Haigis, Hoffer Q, Holladay 1, and SRK/T formulas, except in very long eyes with AL greater than 27.0 mm, for which the SRK/T was the most accurate formula.[22] SRK/T has been considered to be highly reliable for long eyes such as that by Aristodemou et al which is considered to be the largest study ever done on over 8000 eyes.[1] These studies unfortunately did not compare the latest formulas such as Hill RBF and the Barrett’s formula.
Our study demonstrated that Both Hill RBF and Barrett’s formula achieve much higher accuracy as compared to older formulas including SRK/T. This is similar to the study done by Roberts et al[17] that also compared these new IOL formula with the previous existing formula in 400 eyes and found that Hill – radial basis function and more specifically Barrett’s universal II formula showed much less mean numerical error. The same was also demonstrated by Chinese authors Zhang et al which found Barrett’s to be most accurate in 407 eyes.[23]
Overall SRK/T was found to give most reliable results in our study. Though Hill’s RBF gives excellent results in all group of axial length with the best results in shorter axial length. Barrett’s gives reliable results in eyes with slightly longer axial length (>22mm). A study by Cooke et al also recommended Barrett’s as it was found to be the most accurate formula amongst nine different IOL formulas.[24] Shajari et al also evaluated nine IOL formula and again found Barrett’s to be most accurate followed by Hill-RBF and SRK/T formula.[25]
Haigis was the least reliable of formulas when used in all 153 eyes of our study but it was found that Haigis gave reliable results in short axial length eyes and its use should be restricted to that.
References
1. 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.
2. Narvaez J, Zimmerman G, Stulting RD, Chang DH. Accuracy of intraocular lens power prediction using the Hoffer Q, Holladay 1, Holladay 2, and SRK/T formulas. J Cataract Refract Surg 2006; 32:2050–2053.
3. Kane JX, Van Heerden A, Atik A, Petsoglou C. Intraocular lens power formula accuracy: Comparison of 7 formulas. J Cataract Refract Surg. 2016 Oct;42(10):1490-1500.
4. Kane JX, Van Heerden A, Atik A, Petsoglou C. Accuracy of 3 new methods for intraocular lens power selection. J Cataract Refract Surg. 2017 Mar;43(3):333-339.
5. https://www.apacrs.org/barrett_universal2105/. Accessed over Aug-Sept 2018
6. https://rbfcalculator.com/online/index.html. Accessed over Aug-Sept 2018
7. https://www.iolcalc.com/home. Accessed over Aug-Sept 2018
8. TECNIS® Foldable Posterior Chamber Intraocular Lens [package insert]. Santa Ana, Calif. Abbott Medical Optics Inc. Available on https://www.precisionlens.net/tecnis-1-piece-iol-model-zcb00. Accessed on 13/09/18.
9. Sanders D, Retzlaff J, Kraff M, et al. Comparison of the accuracy of the Binkhorst, Colenbrander, and SRK implant power prediction formulas. J Am Intraocul Implant Soc. 1981;7(4):337-340.
10. Sanders DR, Retzlaff J, Kraff MC. Comparison of empirically derived and theoretical aphakic refraction formulas. Arch Ophthalmol. 1983;101(6):965-967.
11. Retzlaff JA, Sanders DR, Kraff MC. Development of the SRK/T intraocular lens implant power calculation formula. J Cataract Refract Surg. 1990.
12. Haigis W. Challenges and approaches in modern biometry and IOL calculation. Saudi Journal of Ophthalmology. 2012;26(1):7-12.
13. Ladas JG, Siddiqui AA, Devgan U, Jun AS. A 3-D “Super Surface” Combining Modern Intraocular Lens Formulas to Generate a “Super Formula” and Maximize Accuracy. JAMA Ophthalmol. 2015 Dec;133(12):1431-6.
14. Barrett GD. An improved universal theoretical formula for intraocular lens power prediction. J Cataract Refract Surg. 1993 Nov;19(6):713-20.
15. Jack T. Holladay, Kathryn H. Musgrove, Thomas C. Prager, John W. Lewis, Thomas Y. Chandler, Richard S. Ruiz. A three-part system for refining intraocular lens power calculations J Cataract Refract Surg. 1988 Jan; 14(1): 17-24.
16. Hoffer KJ. The Hoffer Q formula: a comparison of theoretic and regression formulas. J Cataract Refract Surg 1993; 19: 700-12; errata 1994; 20: 677.
17. Roberts TV, Hodge C, Sutton G, Lawless M; contributors to the Vision Eye Institute IOL outcomes registry. Comparison of Hill-radial basis function, Barrett Universal and current third generation formulas for the calculation of intraocular lens power during cataract surgery. Clin Exp Ophthalmol. 2018 Apr;46(3):240-246.
18. Hoffer KJ. Clinical results using the Holladay 2 intraocular lens power formula. J Cataract Refract Surg 2000; 26: 1233-7.
19. Moschos MM, Chatziralli IP, Koutsandrea C. Intraocular lens power calculation in eyes with short axial length. Indian J Ophthalmol. 2014 Jun;62(6):692-4.
20. MacLaren RE, Bourne RR, Restori M, Allan BD. Biometry and formula accuracy with intraocular lenses used for cataract surgery in extreme hyperopia. Am J Ophthalmol 2007; 143: 920-31.
21. Gökce SE, Zeiter JH, Weikert MP, Koch DD, Hill W, Wang L. Intraocular lens power calculations in short eyes using 7 formulas. J Cataract Refract Surg. 2017Jul;43(7):892-897.
22. Olsen T, Hoffmann P. C constant: new concept for ray tracing–assisted intraocular lens power calculation. J Cataract Refract Surg 2014; 40:764–773.
23. Zhang Y, Ying Liang X, Liu S, Lee JWY, Bhaskar S, Lam DSC. Accuracy of intraocular lens power calculation formulas for highly myopic eyes. J Ophthalmol 2016.
24. Cooke DL, Cooke TL. Comparison of 9 intraocular lens power calculation formulas. J Cataract Refract Surg. 2016 Aug;42(8):1157-64.
25. Shajari M, Kolb CM, Petermann K, Böhm M, Herzog M, de’Lorenzo N, Schönbrunn S,Kohnen T. Comparison of 9 modern intraocular lens power calculation formulas for a quadrifocal intraocular lens. J Cataract Refract Surg. 2018 Aug;44(8):942-948.
Table 1: Comparison of deviation from expected refractive error with post op refractive error
|
Deviation from expected spherical refractive error |
Post op spherical refractive error |
||
| SRK/T deviation | r |
.905 |
|
| p |
.000 |
||
| N |
153 |
||
| Holladay 1 deviation | r |
.869 |
|
| p |
.000 |
||
| N |
153 |
||
| Hoffers Q deviation | r |
.807 |
|
| p |
.000 |
||
| N |
153 |
||
| Haigis deviation | r |
.786 |
|
| p |
.000 |
||
| N |
153 |
||
| Barrett’s deviation | r |
.866 |
|
| p |
.000 |
||
| N |
153 |
||
| Hill RBF deviation | r |
.874 |
|
| p |
.000 |
||
| N |
153 |
||
| Ladas deviation | r |
.867 |
|
| p |
.000 |
||
| N |
153 |
||
r- Relative risk, p- p value, N- no of eyes evaluated
Table 2: Overall Predictive error amongst different formulas
|
Predictive error |
SRK/T | Holladay 1 |
Hoffers Q |
Haigis |
Barrett’s |
Hill’s RBF |
Ladas |
| ±0.25D |
34.6 |
37.3 |
30.1 |
32.7 |
35.3 |
39.2 |
28.8 |
| ±0.5D |
63.4 |
63.4 |
58.8 |
60.1 |
62.1 |
64.1 |
60.1 |
| ±1D |
92.2 |
90.2 |
87.6 |
88.2 |
90.2 |
92.2 |
90.8 |
| ±2D |
98.7 |
98.7 |
98.7 |
98.7 |
98.7 |
98.7 |
98.7 |
| >±2D |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
Table 3: Predictive error in short eyes (<22mm) amongst different formulas
Table 4: Paired sample test of post op sphere with different formula in short eyes (<22mm) Paired Samples Test |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Mean |
SD |
95% Confidence Interval of the Difference |
t |
dif |
p |
||||||||||||||||||||||||||||||||||||||||||||||||||
|
Lower |
Upper |
||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Pair 1 | SRK/T – Post op Sphere
and Target Sphere |
-.77333 |
.24980 |
-.96535 |
-.58132 |
-9.287 |
8 |
.000 |
|||||||||||||||||||||||||||||||||||||||||||||||
| Pair 2 | Holladay 1 – Post op Sphere
and Target Sphere |
-.79889 |
.29918 |
-1.02886 |
-.56892 |
-8.011 |
8 |
.000 |
|||||||||||||||||||||||||||||||||||||||||||||||
| Pair 3 | Hoffers Q Post op Sphere
and Target Sphere |
-.91889 |
.31227 |
-1.15892 |
-.67886 |
-8.828 |
8 |
.000 |
|||||||||||||||||||||||||||||||||||||||||||||||
| Pair 4 | Haigis Q Post op Sphere
and Target Sphere |
-.63889 |
.35491 |
-.91170 |
-.36608 |
-5.400 |
8 |
.001 |
|||||||||||||||||||||||||||||||||||||||||||||||
| Pair 5 | Barrett’s Post op Sphere
and Target Sphere |
-.72333 |
.26669 |
-.92833 |
-.51834 |
-8.137 |
8 |
.000 |
|||||||||||||||||||||||||||||||||||||||||||||||
| Pair 6 | Hills RBF Post op Sphere
and Target Sphere |
-.60778 |
.28115 |
-.82389 |
-.39167 |
-6.485 |
8 |
.000 |
|||||||||||||||||||||||||||||||||||||||||||||||
| Pair 7 | Ladas Post op Sphere
and Target Sphere |
-.70667 |
.33019 |
-.96047 |
-.45286 |
-6.421 |
8 |
.000 |
|||||||||||||||||||||||||||||||||||||||||||||||
| Table 5: Paired sample test of post op sphere with different formula in medium eyes (22-24.5mm)
Paired Samples Test |
||||||||||||
|
Mean |
SD |
Std. Error Mean |
95% Confidence Interval of the Difference |
T value | diff | P value | ||||||
|
Lower |
Upper |
|||||||||||
| Pair 1 | SRK/T – Post op Sphere
and Target Sphere |
-.03870 |
.62809 |
.05857 |
-.15472 |
.07733 |
-.661 |
114 |
.510 |
|||
| Pair 2 | Holladay 1 – Post op Sphere
and Target Sphere |
-.04809 |
.64550 |
.06019 |
-.16733 |
.07116 |
-.799 |
114 |
.426 |
|||
| Pair 3 | Hoffers Q Post op Sphere
and Target Sphere |
-.06296 |
.68912 |
.06426 |
-.19026 |
.06434 |
-.980 |
114 |
.329 |
|||
| Pair 4 | Haigis Q Post op Sphere
and Target Sphere |
-.11930 |
.68751 |
.06411 |
-.24631 |
.00770 |
-1.861 |
114 |
.065 |
|||
| Pair 5 | Barrett’s Post op Sphere
and Target Sphere |
-.01548 |
.64629 |
.06027 |
-.13487 |
.10391 |
-.257 |
114 |
.798 |
|||
| Pair 6 | Hills RBF Post op Sphere
and Target Sphere |
-.01765 |
.62892 |
.05865 |
-.13383 |
.09853 |
-.301 |
114 |
.764 |
|||
| Pair 7 | Ladas Post op Sphere
and Target Sphere |
.03478 |
.64823 |
.06045 |
-.08496 |
.15453 |
.575 |
114 |
.566 |
|||
Table 6: Predictive error in medium length eyes (22mm-24.5mm) amongst different formulas
| SRK/T | Holladay |
Hoffers |
Haigis |
Barrett’s |
Hill |
Ladas |
|
| ±0.25D |
33.9 |
36.5 |
32.2 |
35.7 |
33.9 |
37.4 |
27.8 |
| ±0.5D |
64.3 |
63.5 |
60.9 |
60.9 |
62.6 |
65.2 |
60.9 |
| ±1D |
90.4 |
91.3 |
89.6 |
87.8 |
88.7 |
89.6 |
90.4 |
| ±2D |
98.3 |
98.3 |
98.3 |
98.3 |
98.3 |
98.3 |
98.3 |
| >±2D |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
|
Table 7: Paired sample test of post op sphere with different formula in long eyes (>24.5mm) Paired Samples Test |
|||||||||||||||||
|
Mean |
Std. Deviation |
Std. Error Mean |
95% Confidence Interval of the Difference |
T value | Diff | P value | |||||||||||
|
Lower |
Upper |
||||||||||||||||
| Pair 1 | SRK/T – Post op Sphere
and Target Sphere |
.05966 |
.43516 |
.08081 |
-.10587 |
.22518 |
.738 |
28 |
.467 |
||||||||
| Pair 2 | Holladay 1 – Post op Sphere
and Target Sphere |
.06724 |
.47578 |
.08835 |
-.11373 |
.24822 |
.761 |
28 |
.453 |
||||||||
| Pair 3 | Hoffers Q Post op Sphere
and Target Sphere |
.16621 |
.56467 |
.10486 |
-.04858 |
.38100 |
1.585 |
28 |
.124 |
||||||||
| Pair 4 | Haigis Q Post op Sphere
and Target Sphere |
-.13241 |
.56224 |
.10441 |
-.34628 |
.08145 |
-1.268 |
28 |
.215 |
||||||||
| Pair 5 | Barrett’s Post op Sphere
and Target Sphere |
.03448 |
.48006 |
.08915 |
-.14812 |
.21709 |
.387 |
28 |
.702 |
||||||||
| Pair 6 | Hills RBF Post op Sphere
and Target Sphere |
-.03241 |
.44769 |
.08313 |
-.20271 |
.13788 |
-.390 |
28 |
.700 |
||||||||
| Ladas Post op Sphere
and Target Sphere |
|||||||||||||||||
| Pair 7 | SRK/T – Post op Sphere
and Target Sphere |
.03862 |
.49445 |
.09182 |
-.14946 |
.22670 |
.421 |
28 |
.677 |
||||||||
Table 8: Predictive error in long eyes (>24.5mm) amongst different formulas
| SRK/T | Holladay |
Hoffers |
Haigis |
Barrett’s |
Hill |
Ladas |
|
| ±0.25D |
44.8 |
48.3 |
31.0 |
27.6 |
48.3 |
55.2 |
37.9 |
| ±0.5D |
75.9 |
79.3 |
65.5 |
65.5 |
72.4 |
72.4 |
69.0 |
| ±1D |
96.6 |
96.6 |
89.7 |
93.1 |
96.6 |
100.0 |
96.6 |
| ±2D |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
Leave a Comment