is an associate professor of ophthalmology at the Cullen Eye Institute, Department of Ophthalmology, Baylor College of Medicine. Her areas of research include various aspects in cataract surgery, refractive surgery, diagnostic devices, optics, and wavefront technology and its use in refractive and cataract surgery.
is an ophthalmology resident at Baylor College of Medicine in Houston, Texas. He completed his medical school training at the University of Washington and his undergraduate studies at Brigham Young University.
is a fellow in Cornea, Anterior Segment, and Refractive Surgery at the Cullen Eye Institute, Department of Ophthalmology, Baylor College of Medicine. She received her M.D. at the Washington University School of Medicine in St. Louis. After doing an internship in Internal Medicine at the Barnes Jewish Hospital, she completed ophthalmology residency training at the University of Michigan Kellogg Eye Center in Ann Arbor, Michigan.
is an Associate Professor at the Cullen Eye Institute, Baylor College of Medicine, where he specializes in corneal, cataract, and refractive surgery. His research interests include biomedical optics, anterior segment imaging, intraocular lens technology, and wavefront applications in cataract and refractive surgery. He is the Residency Program Director at the Cullen Eye Institute, Baylor College of Medicine, and the Medical Director of the Lion’s Eye Bank of Texas.
is Professor and the Allen, Mosbacher, and Law Chair in Ophthalmology at the Cullen Eye Institute, Baylor College of Medicine, where he specializes in cataract and refractive surgery. His research interests include optics of cataract and refractive surgery, intraocular lens technology, anterior segment imaging, and surgical techniques in cataract and refractive surgery. He is Editor Emeritus of the Journal of Cataract and Refractive Surgery and past president of the American Ophthalmological Society, American Society of Cataract and Refractive Surgery, and International Intraocular Implant Club.
4.1 Introduction
Modern intraocular lens (IOL) power calculation formulas tend to produce inaccurate outcomes in eyes with long axial lengths. In eyes greater than 25.0 mm, current formulas often suggest IOLs of insufficient power, which leave patients with postoperative hyperopia [1–6]. There is no consensus on which of the IOL power calculation formulas is most accurate in long eyes, but none are ideal in their unmodified form. Many surgeons just empirically aim for a more myopic postoperative outcome in order to try to avoid hyperopic surprises, but axial length optimization formulas derived using regression analysis now allow for more precise and predictable refractive outcomes in these patients [7].
It has been suggested that the increased incidence of posterior staphylomas in eyes with axial high myopia and the resultant inaccurate measurement of preoperative axial length is the main reason for postoperative hyperopic outcomes in patients with axial high myopia [8]. When ultrasonic biometric methods are used, axial length can be inadvertently measured to the depth of a staphyloma rather than to the fovea, leading to the selection of an IOL of insufficient power. The advent of optical coherence biometry permits more accurate measurements in the presence of posterior staphylomas because the patient fixates along the direction of the measuring beam. The instrument is therefore more likely to display an accurate axial length to the center of the macula.
However, minimizing or eliminating the impact of posterior staphylomas on IOL calculations does not necessarily prevent hyperopic surprises in long eyes. MacLaren et al. [2] evaluated the accuracy of biometry in eyes with negative-powered or zero-powered IOLs using the SRK/T formula. They reported consistent hyperopic errors among all three methods of biometry (A-scan, B-scan, and optical), which suggests that the potential error induced by a posterior staphyloma is not solely responsible for hyperopic outcomes in these eyes.
Recently, Wang et al. [7] evaluated the accuracy of refractive error prediction of 4 IOL power calculation formulas (Holladay 1, Haigis, SRK/T, and Hoffer Q) in eyes with axial length greater than 25.0 mm and developed regression formulas to optimize axial length and improve prediction accuracy. Subsequently gathered unpublished data validate the superior predictive accuracy of lens power calculation formulas when optimized axial lengths are used compared to unoptimized axial lengths and another method of IOL power calculation, the Barrett Universal formula [9], which is intended to be used in long eyes without axial length modification.
4.2 Methods
4.2.1 Development and Validation of Axial Length Optimization Formulas
Consecutive cases of eyes with axial length greater than 25.0 mm that underwent cataract extraction by phacoemulsification and intraocular lens implantation by the same surgeon (D.D.K.) at Cullen Eye Institute, Baylor College of Medicine, Houston, Texas, USA, from October 2002 to October 2005 were reviewed. All surgeries were performed through a small temporal clear corneal incision, and the surgeon chose the power of the implanted IOL based on the Holladay 1 formula. The inclusion criteria were (1) implantation of an Acrysof SA60AT, Acrysof SN60AT, or Acrysof MA60MA posterior chamber IOL (all Alcon Laboratories, Inc.), (2) biometric measurements by partial coherence interferometry (IOLMaster, Carl Zeiss Meditec, Inc.), (3) no prior ocular surgery and no intraoperative or postoperative complications, and (4) postoperative corrected distance visual acuity of 20/30 or better.
First, the Holladay 1, Haigis, SRK/T, and Hoffer Q formulas were evaluated for accuracy in long eyes using both the manufacturers’ lens constants and back-calculated optimized lens constants. In order to evaluate the accuracy of refractive prediction of these formulas, the refractive prediction error (RPE) was calculated for each eye (actual refraction 3 weeks or more postoperatively minus predicted refraction). A positive RPE indicates a hyperopic outcome. Mean RPE, referred to as mean numerical error (MNE), was calculated for each of the 4 formulas as well as the percentage of eyes with a positive RPE. Lens constants were optimized retrospectively for each formula by obtaining an MNE of zero. This was performed with the IOLMaster device for the Holladay 1, SRK/T, and Hoffer Q formulas and by using multiple regression analysis for the Haigis formula [8]. Then, using the absolute values of the RPEs, the mean absolute error (MAE) was also calculated for each formula.
In order to develop and verify axial length optimization formulas, the eyes used in this main data set were randomized into two groups: one to develop the axial length optimization method and the other to validate the results. In group one, the ideal axial length for each eye, which would have produced an RPE of zero, was back-calculated postoperatively. The manufacturers’ lens constants were used for this calculation. Regression analysis was performed to create a formula relating the optimized axial lengths to the original IOLMaster axial lengths. Group two was then used to assess the predictive accuracy of IOL calculation formulas using axial lengths optimized with the regression equation developed from group one.
Two additional data sets were then used for validation of the accuracy of the method of optimizing axial lengths and for refining the optimization formula. One set consisted of consecutive cases from a single surgeon in another center (Thomas Kohnen MD, PhD, FEBO, Department of Ophthalmology; Johann Wolfgang Goethe University, Frankfurt am Main, Germany) using Acrysof MA60MA IOL implantation for refractive lens exchange through 3.0–3.5 mm unsutured temporal or on-axis posterior limbal tunnel incisions. The second set was again consecutive cases with axial length greater than 25.0 mm who underwent cataract extraction and IOL implantation by the same surgeon (D.D.K) from November 2005 to April 2008 at Cullen Eye institute.
4.2.2 Further Validation of Axial Length Optimization and Comparison to Barrett Universal Formula
Data that are yet unpublished have been gathered more recently to further validate the use of optimized axial lengths when using the Holladay 1, SRK/T, and Haigis formulas for lens power calculation, as well as to compare the accuracy of these optimized formulas to the Barrett Universal Formula [9]. Consecutive cases by 1 surgeon (D.D.K.) on eyes with an axial length greater than 25 mm from April 2010 to December 2012 were analyzed. The patients had no prior ocular surgery, no intraoperative or postoperative complications, and postoperative corrected distance visual acuity of 20/30 or better. The Holladay 1 formula with optimized axial lengths based on the regression formula developed from the previously published study by Wang et al. [7] was used to select the IOL that was implanted. The MNE and percentage of eyes with RPE greater than 0, +0.25, +0.5, and +0.75 D were back-calculated using unoptimized axial lengths, optimized axial lengths, and the Barrett Universal Formula. These groups were compared using paired T-test analysis.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
