From South Australian Institute of Ophthalmology, The Queen Elizabeth Hospital, University of Adelaide, Woodville South, South Australia.
The authors have no proprietary or financial interest in the materials presented herein.
The authors thank Drs Arthur Karagiannis, Peter Cooper, Darcy Economos, and Neil Gehling, who performed the surgeries for this study, and the staff at The Queen Elizabeth Hospital for their support and assistance.
Study concept and design (J.L.); data collection (J.L.); analysis and interpretation of data (J.L., M.G.); drafting of the manuscript (J.L.); critical revision of the manuscript (M.G.); statistical expertise (J.L.); administrative, technical, or material support (M.G.)
Correspondence: John Landers, MBBS, MPH, PhD, The Queen Elizabeth Hospital, 28 Woodville Rd, Woodville South, South Australia 5011. Tel: 618 8222 7579; Fax: 618 8222 6233; E-mail: email@example.com
During cataract surgery, an intraocular lens (IOL) is usually implanted, the power of which is derived from preoperative measurements and calculations. Any deviation from target refraction may be associated with factors relating to the accuracy of keratometry or biometry measurement, or to the calculation used to derive IOL power.1,2 However, in the absence of evidence of measurement or calculation error, the outcome of the first eye may be used as a basis for adjustments to lens power when selecting an IOL for the second eye.3–5 Although this may seem the correct approach, evidence suggests that this method may be flawed.3–5
We designed the current study to investigate the relationship between the refractive outcome of each eye and factors that may modify this relationship.
Patients and Methods
Following ethical approval from the Human Research and Ethics Committee of The Queen Elizabeth Hospital, we reviewed case notes of patients who underwent cataract surgery and IOL implantation in each eye between January 2006 and December 2006. The recruitment of patients conformed to the tenets of the Declaration of Helsinki. Inclusion criteria were bilateral phacoemulsification cataract surgery during which an IOL was implanted in the capsular bag. Exclusion criteria were final best spectacle-corrected visual acuity (BSCVA) of 20/30 (logMAR 0.18) or worse, preoperative sphere >5.00 diopters (D) and cylinder >3.00 D (to ensure reasonable homogeneity and accuracy using an SRK-T formula), previous corneal diseases (eg, dystrophy, injury, irregular astigmatism, current or previous pterygium), previous corneal refractive surgery, (eg, photorefractive keratectomy, LASIK), IOL not in-the-bag, no IOL, previous anterior segment or vitreoretinal surgery including the preoperative presence of intraocular silicone oil, or secondary procedure occurring at the same time as phacoemulsification (eg, trabeculectomy, limbal relaxing incisions).
Eighty-one patients were found to be eligible. Data were collected from case notes regarding preoperative keratometry, refractive error and biometry, type and power of IOL used, postoperative refractive error, and BSCVA. Surgery was performed by one of four ophthalmologists; however, every patient underwent surgery in both eyes by the same ophthalmologist. Postoperative refractive error was determined with subjective refraction at 6 weeks after the procedure. Keratometry was performed using a Humphrey 599 automatic refractor keratometer (Carl Zeiss Meditec, Dublin, Calif). Biometry was performed using an OcuScan RxP immersion ultrasound biometer (Alcon Laboratories Inc, Ft Worth, Tex). The output from this device displays the calculated target refraction for a variety of powers for a variety of IOLs. This was used to determine the IOL target refraction, which was selected for the surgery with an SRK-T formula. To compare the baseline parameters among pairs of eyes and illustrate their preoperative similarity, the IOL power was recalculated for the same IOL type in each eye and adjusted for the exact power required to achieve emmetropia. This parameter was only used in the table of descriptive parameters (Table 1), but not in any analysis. The difference between the final postoperative refractive spectacle correction required to achieve BSCVA and the target refraction was calculated. This outcome variable was termed “the difference from target refraction,” and it was believed that this value should be the same for any eye regardless of the intended target refraction (eg, if a refractive outcome is more myopic than the target refraction, the difference from target refraction should occur regardless of whether the intended target was emmetropia or −1.00 or +2.00 D, etc). A comparison was then made between eyes, and details of any associations were determined statistically.
Table 1: Descriptive Parameters of 162 Eyes of 81 Patients Who Received IOL Implantation After Cataract Surgery
The 6.12 Statistical Analysis System (SAS Institute Inc, Cary, NC) was used including descriptive statistics, Student t test, paired t test, chi-square test, analysis of variance (ANOVA), simple linear regression, and multiple linear regression. Keratometry, anterior chamber depth, axial length, adjusted IOL power, and difference from target refraction were considered as continuous variables. Intraocular lens type and surgeon were considered as nominal discrete variables. The ametropia group was considered as a dichotomous discrete variable. Adjustments for IOL type and operating surgeon were made with grouped dummy variables in a multiple regression equation. Test statistics and P values are presented. A P value <.05 was considered statistically significant.
Within our sample of 81 patients, 27 (33%) were men and 54 (67%) were women. The average age was 75±8.4 years (range: 57 to 93 years); 75±9.1 years for men and 74±8.1 years for women (t=0.13; P=.90). Of the 162 eyes, 30 (18%) eyes received a MA30AC IOL (Alcon Laboratories, Ft Worth, Tex), 22 (14%) received an AR40e IOL (Abbott Medical Optics [AMO], Santa Ana, Calif), 26 (16%) received an Akreos Adapt IOL (Bausch & Lomb, Rochester, NY), 27 (17%) received an SN60WF IOL (Alcon Laboratories), 29 (18%) received a Tecnis IOL (AMO), and 28 (17%) received an Akreos Adapt AO IOL (Bausch & Lomb) (Table 1). Eighteen pairs (22%) of eyes received the same type of IOL and 63 pairs (78%) received different IOLs. Second eyes were more likely to have received aspheric IOLs (Table 1). The median time between procedures was 89 days (range: 53 to 138 days).
Preoperatively, bilateral eyes were similar. The inter-eye difference was 0.003±0.58 D (t=0.04; P=.97) for mean keratometry, 0.02±0.29 mm (t=0.47; P=.64) for anterior chamber depth, 0.02±0.16 mm (t=1.15; P=.25) for axial length, and 0.07±0.8 D (t=0.77; P=.45) for adjusted IOL power predicted preoperatively to achieve emmetropia (Table 1).
We observed an average difference from target refraction of −0.24±0.77 D for the first eye and −0.18±0.67 D for the second eye (t=0.64; P=.52). This was not associated with IOL type (F=0.75; P=.61). A statistically significant inter-eye association for the difference from target refraction (t=3.02; P=.003) was noted, such that if the first eye showed a large difference from target refraction postoperatively, then the second eye often showed this as well. However, there was only a modest correlation for this association (r2=0.10), with the variability of the difference from target refraction in the first eye contributing to only 10% of the variability in the second (Fig). This association was not modified by age, sex, keratometry, biometry (anterior chamber depth or axial length), operating surgeon, or IOL type. It also was not associated with any inter-eye IOL combinations.
Figure. Difference Between Calculated Target Refraction and Refractive Error After Phacoemulsification for Each Eye Showing Trend Line for the Linear Association Between the First Eye and the Second Eye.
If we were to consider only the direction in which the difference from target refraction occurred (more myopic or more hypermetropic than target refraction), this would still result in a 63% inter-eye concordance, with 37% of eye pairs behaving independently (Table 2).
Table 2: Relationship Between First and Second Eyes that Were More Myopic or Hypermetropic Postoperatively Compared to Preoperative Intended Refraction
When predicting the final refractive outcome after cataract surgery, it would seem intuitive that the experiences from the first eye should assist the surgeon when choosing the power of an IOL for the second eye. Our study demonstrated that the refractive outcome of the first eye following cataract surgery was a poor predictor for the second eye. Concordance was seen in some patients; however, non-concordance was frequently seen. Previous works using phacoemulsification or extracapsular cataract extraction and a variety of IOLs have suggested this to be the case.3–5 Any adjustments made to IOL power for the second eye based on the refractive outcome of the first eye did not yield a more accurate outcome and, in some instances, led to more variable refractive results. This inherent unpredictability may be accounted for by variability in measurable factors, such as keratometry, axial length measurement, preoperative IOL calculation, and postoperative subjective refraction.1,2 However, it may also be due to less quantifiable factors such as capsular bag contraction.6,7
The limitations of our study included the retrospective nature of data collection. This means that IOL selection was not done in a masked fashion for the second eye and individual surgeons could have altered the IOL power for the second eye based on the results of the first. However, this should not affect the results given that the analysis was based on the value of the difference between observed refraction and target refraction and not on the actual final refractive outcome. Furthermore, the study was designed and carried out after patients had undergone surgery; therefore, clinicians could have no knowledge of the study prior to IOL selection. The type of IOL was also not consistent across patients in our study. This may have had an impact if one particular IOL was not optimized for a surgeon compared with another. However, at The Queen Elizabeth Hospital, an internal audit has found that the best outcomes are achieved simply by using the manufacturer’s recommended a-constant; therefore, no IOL would have been optimized. Furthermore, adjustments were made to our analysis to account for the potential effect of the different IOLs and surgeon. Measurement of axial length was made using immersion ultrasound, which has a lower axial resolution than partial coherence interferometry8 and may have contributed to measurement error.
Our study has shown that the degree to which final refractive error differs from target refraction following cataract surgery in the first eye, when calculated using immersion ultrasound biometry, cannot be used reliably as a guide for the second eye. We suggest that the power of IOLs for a pair of eyes be considered independently when attempting to predict refractive outcome.
- Olsen T. Sources of error in intraocular lens power calculation. J Cataract Refract Surg. 1992;18:125–129.
- Norrby S. Sources of error in intraocular lens power calculation. J Cataract Refract Surg. 2008;34:368–376. doi:10.1016/j.jcrs.2007.10.031 [CrossRef]
- Jabbour J, Irwig L, Macaskill P, Hennessy MP. Intraocular lens power in bilateral cataract surgery: whether adjusting for error of predicted refraction in the first eye improves prediction in the second eye. J Cataract Refract Surg. 2006;32:2091–2097. doi:10.1016/j.jcrs.2006.08.030 [CrossRef]
- Olsen T, Logstrup N, Olesen H, Corydon L. Using the surgical result in the first eye to calculate intraocular lens power for the second eye. J Cataract Refract Surg. 1993;19:36–39.
- Murphy GE, Murphy CG. Minimizing anisometropia in bilateral pseudophakia. J Cataract Refract Surg. 1992;18:95–99.
- Kato S, Oshika T, Numaga J, Hayashi Y, Oshiro M, Yuguchi T, Kaiya T. Anterior capsular contraction after cataract surgery in eyes of diabetic patients. Br J Ophthalmol. 2001;85:21–23. doi:10.1136/bjo.85.1.21 [CrossRef]
- Zambarakji HJ, Rauz S, Reynolds A, Joshi N, Simcock PR, Kinnear PE. Capsulorhexis phymosis following uncomplicated phacoemulsification surgery. Eye. 1997;11:635–638.
- Kiss B, Findl O, Menapace R, Wirtitsch M, Drexler W, Hitzenberger CK, Fercher AF. Biometry of cataractous eyes using partial coherence interferometry: clinical feasibility study of a commercial prototype I. J Cataract Refract Surg. 2002;28:224–229. doi:10.1016/S0886-3350(01)01272-X [CrossRef]
Descriptive Parameters of 162 Eyes of 81 Patients Who Received IOL Implantation After Cataract Surgery
|Parameter||First Eye||Second Eye||Test Statistic||PValue|
|Average keratometry (D)||43.68±1.31||43.68±1.36||0.04*||.97|
|Anterior chamber depth (mm)||3.10±0.43||3.11±0.43||0.47*||.64|
|Axial length (mm)||23.08±0.88||23.06±0.89||1.15*||.25|
|Adjusted IOL power using SRK-T formula (D)||21.60±3.37||21.67±3.49||0.77*||.45|
|IOL implanted (n) (%)|
| MA30AC||26 (32)||4 (5)|
| AR40e||13 (16)||9 (11)|
| Adapt||24 (30)||2 (2)|
| SN60WF||10 (12)||17 (21)|
| Tecnis||4 (5)||25 (31)|
| Adapt AO||4 (5)||24 (30)||82.26†||>.001|
|Postoperative difference from target refraction (D)||−0.24±0.77||−0.18±0.67||0.64*||.52|
|Proportion <1.00 D from target refraction postoperatively (%)||81||84||0.20‡||.65|
Relationship Between First and Second Eyes that Were More Myopic or Hypermetropic Postoperatively Compared to Preoperative Intended Refraction
|More Myopic Than Target||More Hypermetropic Than Target|
| More Myopic Than Target||36||14|
| More Hypermetropic Than Target||16||15|