Advertisement

Accuracy of Intraocular Lens Power Formulas Modified for Patients with Keratoconus

      Purpose

      To assess the accuracy of intraocular lens (IOL) power formulas modified specifically for patients with keratoconus (Holladay 2 with keratoconus adjustment and Kane keratoconus formula) compared with normal IOL power formulas (Barrett Universal 2, Haigis, Hoffer Q, Holladay 1, Holladay 2, Kane, and SRK/T).

      Design

      Retrospective consecutive case series.

      Participants

      A total of 147 eyes of 147 patients with keratoconus.

      Methods

      Data from patients with keratoconus who had preoperative IOLMaster biometry were included. A single eye per qualifying patient was randomly selected. The predicted refraction was calculated for each of the formulas and compared with the actual refractive outcome to give the prediction error. Subgroup analysis based on the steepest corneal power measured by biometry (stage 1: ≤48 diopters [D], stage 2: >48 D and ≤53 D, and stage 3: >53 D) was performed.

      Main Outcome Measure

      Prediction error.

      Results

      On the basis of the mean absolute prediction error (MAE), the formulas were ranked as follows: Kane keratoconus formula (0.81 D), SRK/T (1.00 D), Barrett Universal 2 (1.03 D), unmodified Kane (1.05 D), Holladay 1 (1.18 D), unmodified Holladay 2 (1.19 D), Haigis (1.22 D), Hoffer Q (1.30 D), and Holladay 2 with keratoconus adjustment (1.32 D). The Kane keratoconus formula had a statistically significant lower MAE compared with all formulas (P < 0.01). In stage 3 keratoconus, all nonmodified formulas had a hyperopic mean prediction error ranging from 1.72 to 3.02 D.

      Conclusions

      The Kane keratoconus formula was the most accurate formula in this series. The SRK/T was the most accurate of the traditional IOL formulas. All normal IOL formulas resulted in hyperopic refractive outcomes that worsened as the corneal power increased. Suggestions for target refractive aims in each stage of keratoconus are given.

      Abbreviations and Acronyms:

      D (diopter), ELP (effective lens position), IOL (intraocular lens), MAE (mean absolute prediction error), MedAE (median absolute prediction error), SRK/T (Sanders, Retzlaff, Kraff), ULIB (User Group for Laser Interference Biometry)
      Keratoconus is a progressive disorder characterized by central or paracentral corneal thinning and ectasia. Intraocular lens (IOL) power calculation in these eyes represents a significant challenge.
      Kamiya et al
      • Kamiya K.
      • Iijima K.
      • Nobuyuki S.
      • et al.
      Predictability of Intraocular Lens Power Calculation for Cataract with Keratoconus: a multicenter study.
      reported on 71 patients with keratoconus comparing the Haigis, Hoffer Q, Holladay 1, Holladay 2, and SRK/T formulas and found that that the SRK/T formula was the most accurate with 36% of eyes within 0.50 diopters (D) of the final manifest refraction. Savini et al
      • Savini G.
      • Abbate R.
      • Hoffer K.J.
      • et al.
      Intraocular lens power calculation in eyes with keratoconus.
      also found that the SRK/T was the most accurate formula in 41 patients (compared with Barrett Universal 2, Haigis, Hoffer Q, and Holladay 1), with 43.9% of eyes within 0.50 D. Both studies found that all formulas resulted in a hyperopic refractive surprise that worsened with more advanced stages of the disease. Suggestions regarding an appropriate myopic refractive target to avoid unwanted postoperative hyperopic error have been proposed.
      • Watson M.P.
      • Anand S.
      • Bhogal M.
      • et al.
      Cataract surgery outcome in eyes with keratoconus.
      These refractive results in keratoconus studies are significantly worse than the 75% to 80% of eyes within 0.50 D usually seen in nonkeratoconic eyes,
      • Melles R.B.
      • Holladay J.T.
      • Chang W.J.
      Accuracy of intraocular lens calculation formulas.
      for which there are many reasons. First, the calculation of corneal power in IOL power calculation is based on an assumed ratio of the anterior to posterior corneal power that is not maintained in keratoconus.
      • Camps V.J.
      • Piñero D.P.
      • Caravaca-Arens E.
      • et al.
      New approach for correction of error associated with keratometric estimation of corneal power in keratoconus.
      Second, IOL power formulas typically use the corneal power in the estimation of the effective lens position (ELP), meaning an inaccurate corneal power will lead to errors in ELP. Finally, keratometers—which assume that the corneal curvature is constant along a given meridian—are inaccurate in keratoconic eyes where the corneal curvature is often variable along a particular meridian and the principal power meridians may not be orthogonal.
      Two formulas have specific adjustments to counter these issues seen in keratoconus eyes. Within the Holladay IOL Consultant software package, it is possible to designate a patient as having keratoconus, which leads to a change in the refractive prediction. It does so by allowing the software to “differentiate a normal [non-ectatic] steep K-reading in a small anterior segment from a patient with keratoconus.”
      • Holladay J.T.
      Holladay IOL Consultant Software & Surgical Outcomes Assessment. 1105.2019 ed.
      This likely changes the underlying ELP algorithm to account for the independence between anterior segment size and axial length. The Kane keratoconus formula (available at www.iolformula.com) was developed using a purely theoretical modification to the original Kane formula.
      • Connell B.J.
      • Kane J.X.
      Comparison of the Kane formula with existing formulas for intraocular lens power selection.
      ,
      • Kane J.X.
      Kane Formula.
      It uses a modified corneal power derived from anterior corneal radii of curvature that better represents the true anterior/posterior ratio in keratoconic eyes while also aiming to minimize the effect of corneal power on the ELP calculation. There is no requirement for additional variables beyond those included in the original formula when using both the Holladay 2 with keratoconus adjustment and the Kane keratoconus, and both work using the same IOL constants as the original formulas.
      Given these keratoconus-specific formulas have never been assessed in a published study, the aim of this study was to test whether either of these formulas would result in improved refractive outcomes compared with preexisting methods.

      Methods

      This study was performed under institutional review board approval and conformed to the tenets of the Declaration of Helsinki. A retrospective chart review was conducted on consecutive cataract surgeries performed from October 2008 to March 2019 at 4 centers in Melbourne and Brisbane, Australia. Inclusion criteria were keratoconus diagnosed by a corneal subspecialist based on corneal topographical findings, preoperative biometry using partial coherence interferometry with the IOLMaster 500 (software versions 3.02, 5.2.1, and 7.7) or 700 (software version 1.50) (Carl Zeiss Meditec AG, Jena, Germany) with subjective manifest refraction performed by an orthoptist using a lane length of 6 m at least 28 days postoperatively. Exclusion criteria included previous corneal or other intraocular surgery, intraoperative or postoperative complications, and postoperative corrected distance visual acuity worse than 20/40. If both eyes of 1 patient were eligible for inclusion in the study, then one was randomly selected as suggested in editorials regarding IOL formula studies by Hoffer et al
      • Hoffer K.J.
      • Aramberri J.
      • Haigis W.
      • et al.
      Protocols for studies of intraocular lens formula accuracy.
      and Wang et al.
      • Wang L.
      • Koch D.D.
      • Hill W.
      • Abulafia A.
      Pursuing perfection in intraocular lens calculations: III. Criteria for analyzing outcomes.
      The Haigis,
      • Haigis W.
      • Lege B.
      • Miller N.
      • Schneider B.
      Comparison of immersion ultrasound biometry and partial coherence interferometry for intraocular lens calculation according to Haigis.
      Hoffer Q,
      • Hoffer K.J.
      The Hoffer Q formula: a comparison of theoretic and regression formulas.
      Holladay 1,
      • Holladay J.T.
      • Prager T.C.
      • Chandler T.Y.
      • et al.
      A three-part system for refining intraocular lens power calculations.
      and SRK/T
      • Retzlaff J.A.
      • Sanders D.R.
      • Kraff M.C.
      Development of the SRK/T intraocular lens implant power calculation formula.
      formulas were calculated using Excel spreadsheets (Microsoft Corp, Redmond, WA). The original Kane formula
      • Kane J.X.
      Kane Formula.
      and Kane keratoconus formula were calculated by Kane. The Barrett Universal 2 formula (hereafter referred to as the “Barrett”) was calculated using the APACRS website.
      • Barrett G.
      Barrett Universal II Formula.
      The Holladay 2 formula was calculated using the Holladay IOL Consultant software package
      • Holladay J.T.
      Holladay IOL Consultant Software & Surgical Outcomes Assessment. 1105.2019 ed.
      with and without the keratoconus option selected. The Hill-RBF 2.0 was not included in the study because it has a maximum keratometry value of 52 D, which would have resulted in exclusion of 32 patients (21.6% of the total cohort). The Olsen formula was not included because lens thickness was not available for all patients and the use of the C-constant approach
      • Olsen T.
      • Hoffmann P.
      C constant: new concept for ray tracing-assisted intraocular lens power calculation.
      without the dependence on corneal power is required in keratoconus eyes.
      The IOL constants from the User Group for Laser Interference Biometry (ULIB) website
      • Haigis W.
      Optimized IOL constants for the ZEISS IOLMaster.
      were used for the calculation of each formula given the relatively small number of cases for each IOL implanted as recommended by Savini et al.
      • Savini G.
      • Abbate R.
      • Hoffer K.J.
      • et al.
      Intraocular lens power calculation in eyes with keratoconus.
      The ULIB constants were used to calculate the predicted refraction for each of the formulas. The prediction error was then calculated as the actual postoperative refraction minus the formulas’ predicted refraction using the IOL power implanted. The mean prediction error, mean absolute prediction error (MAE), standard deviation of the prediction error, median absolute prediction error (MedAE), and percentages of eyes that had a prediction error of ±0.25 D, ±0.50 D, ±1.00 D, and ±2.00 D were calculated for each formula. Subgroup analysis was then undertaken by classifying patients by severity of keratoconus according to the same criteria used by Savini et al,
      • Savini G.
      • Abbate R.
      • Hoffer K.J.
      • et al.
      Intraocular lens power calculation in eyes with keratoconus.
      which is a modification of those proposed by Krumeich et al.
      • Krumeich J.H.
      • Daniel J.
      • Knalle A.
      Live-epikeratophakia for keratoconus.
      Patients were classified as stage 1 when the steepest corneal meridian according to optical biometry was ≤48 D, stage 2 when the steepest corneal meridian was between 48 and 53 D, and stage 3 when the steepest corneal meridian was >53 D.
      The differences in mean absolute error between formulas were assessed using the Friedman test, and in the event of a significant result, post hoc analysis was undertaken using the Wilcoxon test with Bonferroni correction as recommended by Aristodemou et al.
      • Aristodemou P.
      • Knox Cartwright N.E.
      • et al.
      Statistical analysis for studies of intraocular lens formula accuracy.
      The Cochran Q test with Bonferroni correction was performed to evaluate whether percentage of eyes within a prediction error of ±0.50 D differed significantly between formulas. A P value less than 0.05 was considered statistically significant. Statistical analysis was performed in R (R Foundation, Vienna, Austria).

      Results

      A total of 147 eyes of 147 patients were eligible for inclusion in the study. Two eyes were excluded because their average keratometry readings exceeded 60 D, which is the maximum allowable entry into the Barrett Universal 2 calculator. The demographics of the study cohort are displayed in Table 1. The number of eyes implanted with each respective IOL and the IOL constants used are displayed in Table 2. According to the modified Krumeich classification, 84 eyes were classified as stage 1, 37 eyes were classified as stage 2, and 25 eyes were classified as stage 3. The number of cases from each author were as follows: B.C. 51, G.R.S. 27, E.C. 27, P.B. 22, and J.C.M. 20.
      Table 1Demographics of the Patient Cohort
      DemographicMeanSDRange
      Axial length (mm)24.381.85(20.5–31.8)
      Anterior chamber depth (mm)3.410.41(2.32–4.46)
      Keratometry (steep) (D)48.424.86(40.5–61.7)
      Keratometry (flat) (D)44.833.69(34.2–58.6)
      IOL power (D)15.907.89(−10 to 34)
      Postoperative spherical equivalent (D)−0.501.48(−6.25 to 5.125)
      Female gender (%)53.4
      D = diopters; IOL = intraocular lens; SD = standard deviation.
      Anterior chamber depth measured from epithelium to lens.
      Table 2Number of and Constants Used for the Different Intraocular Lens Models Used in the Study
      IOL TypenA-ConstantpACDsfa0a1a2
      J&J ZCB007119.35.802.02−1.300.210.25
      Alcon SN60WF291195.641.84−0.7690.2340.217
      Alcon SN6ATx52119.25.811.98−0.3230.2130.208
      J&J AR40M1118.75.391.620.4720.0770.174
      MBI PreciSAL302A11118.95.511.751.3200.4000.100
      Oculentis LU313MF304118.25.111.330.8700.4000.100
      Rayner 623T6118.85.391.641.2000.4000.100
      Zeiss Asphina 40910118.35.121.360.3220.1620.158
      Zeiss AT Torbi709M26118.55.371.591.1300.4000.100
      Table 3 shows the results for all patients. The Kane keratoconus formula had the lowest MAE and highest percentage of eyes within ±0.25 D, ±0.50 D, and ±1.00 D, which was statistically significant compared with all other formulas (P < 0.01). The SRK/T had the second lowest MAE, which was statistically significant compared with all formulas (P < 0.01) except the Barrett and original Kane formula. The original Kane formula had the second highest percentage of eyes within ±0.50 D, which was statistically significant compared with the original Haigis, Hoffer Q, Holladay 2, and Holladay 2 with keratoconus adjustment (P < 0.05) but not the Barrett, Holladay 1, or SRK/T. The Barrett and Holladay 1 had the equal third highest percentage of eyes within ±0.50 D, which was statistically significant compared with the Holladay 2 with keratoconus adjustment (P < 0.05) but none of the other formulas. Figure 1 shows the percentage of eyes within prediction errors of 0.25 D, 0.50 D, 1.00 D, and 2.00 D for each of the formulas.
      Table 3Outcomes for All Patients Sorted by Mean Absolute Error
      FormulaMAEMESTDEVMedAE
      Kane keratoconus0.81
      P < 0.01.
      0.041.210.50
      SRK/T1.00
      P < 0.01.
      0.311.490.60
      Barrett1.030.381.570.62
      Kane (original)1.050.531.540.59
      Holladay 11.180.651.690.67
      Holladay 2 (original)1.190.491.730.75
      Haigis1.220.611.720.74
      Hoffer Q1.300.781.810.75
      Holladay 2 (keratoconus adjustment)1.320.641.830.81
      All values expressed in diopters.
      MAE = mean absolute prediction error; ME = mean prediction error; MedAE = median absolute prediction error; STDEV = standard deviation of the prediction error.
      Kane keratoconus formula statistically significant compared with all other formulas. SRK/T formula statistically significant compared with all formulas below it except the Barrett and Kane (original).
      P < 0.01.
      Figure thumbnail gr1
      Figure 1Stacked histograms comparing the percentage of cases within a given diopter (D) range of predicted spherical equivalent refraction outcome for the entire data set.
      The Holladay 2 with keratoconus adjustment was less accurate than the original Holladay 2 without any adjustment with a lower percentage of eyes within each prediction error band (27.4% within ±0.50 D vs. 37.0%) and higher MAE (1.32 vs. 1.19 D), MedAE (0.81 vs. 0.75 D), and standard deviation of the prediction error (1.83 vs. 1.73 D). The difference in mean absolute error and percentage of eyes within ±0.25 D, ±0.50 D, and ±1.00 D was statistically significant (P < 0.01). The Kane keratoconus formula resulted in more accurate predictions compared with the original Kane formula with a higher percentage of eyes within all prediction error bands (50.0% within ±0.50 D vs. 45.2% for the original Kane formula), lower MAE (0.81 vs. 1.05 D), MedAE (0.50 vs. 0.59 D), and standard deviation of the prediction error (1.21 vs. 1.54 D). The difference in mean absolute error and percentage of eyes within ±0.25 D, ±0.50 D, and ±1.00 D between the Kane keratoconus formula and original Kane formula was statistically significant (P < 0.01). Table 4 shows the percentage of eyes that resulted in a hyperopic surprise compared with a myopic surprise for each formula. All formulas except the Kane keratoconus formula resulted in more patients with a hyperopic refractive surprise.
      Table 4Percentage of Eyes with a Hyperopic versus Myopic Prediction Error Sorted According to Alphabetical Order
      FormulaOverallEyes within ±0.50 D
      HyperopicMyopicHyperopicMyopic
      Barrett53.446.620.520.5
      Haigis54.145.913.720.5
      Hoffer Q59.640.415.818.5
      Holladay 159.640.421.219.9
      Holladay 2 (keratoconus adjustment)52.147.912.315.1
      Holladay 2 (original)50.749.315.821.2
      Kane (original)56.843.221.923.3
      Kane keratoconus45.254.821.928.1
      SRK/T53.446.619.919.9
      D = diopters.

       Subgroup Analysis

      Table 5 shows the mean prediction error, MAE, and percentage of eyes within 0.50 and 1.00 D for each formula for each subgroup classification. For stage 1 keratoconus, the Kane keratoconus formula and Kane original formula have identical results because the modification of the formula does not commence until a corneal power >48 D. The Kane keratoconus formula and Kane original formula had the equal lowest MAE and highest percentage of eyes within ±0.50 D compared with all other formulas (P < 0.05 for all). The Barrett had the third lowest MAE, which was statistically significant compared with the Haigis and Holladay 2 (P < 0.05) but not any other formula. No statistically significant difference was found between the percentage of eyes within ±0.50 D for the Barrett and the remaining formulas. For stage 2 keratoconus, the Kane keratoconus formula had the lowest MAE, which was statistically significant compared with the Haigis, Hoffer Q, and Holladay 2 with keratoconus adjustment (P < 0.01) but not the Barrett, Holladay 1, Holladay 2 (original), Kane (original), and SRK/T. The Kane keratoconus formula had a higher percentage of eyes within ±0.50 D, which was statistically significant compared with the Haigis, Hoffer Q, Holladay 1, Holladay 2 with keratoconus adjustment, and SRK/T (P < 0.01) but not the Barrett, Holladay 2 (original), or Kane (original) formula. For stage 3 keratoconus, the Kane keratoconus formula had the lowest MAE, which was statistically significant compared with all formulas (P < 0.01 for all except SRK/T P < 0.05). The Kane keratoconus formula had a higher percentage of eyes within ±0.50 D, which was statistically significant compared with all formulas (P < 0.01). The SRK/T had the second lowest MAE and second highest percentage of eyes within ±0.50 D in this group, which was statistically significant over all formulas (P < 0.01 for both measures) except the Barrett and Haigis.
      Table 5Outcomes for Patients According to the Stage of Keratoconus
      FormulaStage 1Stage 2Stage 3
      ±0.50 D±1.00 DMAEME±0.50 D±1.00 DMAEME±0.50 D±1.00 DMAEME
      Kane keratoconus60.790.50.49
      P < 0.05.
      −0.1843.256.81.08
      P < 0.01.
      0.5324.048.01.44
      P < 0.01.
      0.02
      Kane (original)60.790.50.49
      P < 0.05.
      −0.1837.851.41.231.004.012.02.642.22
      Barrett53.689.30.54−0.2537.854.11.210.894.016.02.451.72
      SRK/T52.488.10.56−0.2329.748.61.130.5112.020.02.321.86
      Holladay 152.485.70.56−0.1837.851.41.311.128.012.03.072.71
      Hoffer Q52.485.70.57−0.1916.248.61.581.470.012.03.363.02
      Haigis51.285.70.58−0.2616.243.21.541.344.016.02.882.43
      Holladay 2 (original)46.482.10.62−0.3835.151.41.271.058.012.03.012.62
      Holladay 2 (keratoconus adjustment)38.181.00.64−0.3616.240.51.591.418.016.03.192.88
      D = diopters; MAE = mean absolute prediction error; ME = mean prediction error.
      Significance values are compared with all formulas in stage 1, compared with the Hoffer Q, Haigis and Holladay 2 (keratoconus adjustment) in stage 2 and compared with all formulas in stage 3 except the SRK/T, which was statistically significant but P < 0.05.
      P < 0.05.
      ∗∗ P < 0.01.

      Discussion

      Our study shows the ongoing challenge of IOL power calculation in keratoconus and reports on a new method—the Kane keratoconus formula—that results in significantly improved refractive outcomes compared with conventional IOL formulas.
      Our findings relating to older generation formulas mirror those of earlier studies
      • Kamiya K.
      • Iijima K.
      • Nobuyuki S.
      • et al.
      Predictability of Intraocular Lens Power Calculation for Cataract with Keratoconus: a multicenter study.
      ,
      • Savini G.
      • Abbate R.
      • Hoffer K.J.
      • et al.
      Intraocular lens power calculation in eyes with keratoconus.
      that demonstrated the significant hyperopic refractive surprise seen in patients with keratoconus. Our study also confirmed that the SRK/T is the most accurate of the older generation of formulas in keratoconus eyes. We found a similar percentages of eyes within a 0.50 D prediction error for the SRK/T (39.2% in our study) compared with previous studies by Kamiya et al
      • Kamiya K.
      • Iijima K.
      • Nobuyuki S.
      • et al.
      Predictability of Intraocular Lens Power Calculation for Cataract with Keratoconus: a multicenter study.
      (36% within 0.50 D) and Savini et al
      • Savini G.
      • Abbate R.
      • Hoffer K.J.
      • et al.
      Intraocular lens power calculation in eyes with keratoconus.
      (43.9% within 0.50 D). These results are significantly less accurate than those seen in nonkeratoconus eyes where between 75% and 83% of eyes would be expected to be within ±0.50 D.
      • Melles R.B.
      • Holladay J.T.
      • Chang W.J.
      Accuracy of intraocular lens calculation formulas.
      ,
      • Melles R.B.
      • Kane J.X.
      • Olsen T.
      • Chang W.J.
      Update on intraocular lens calculation formulas.
      Many factors contribute to this variability, including corneal power measurements that do not take into account the alteration in the keratometric index
      • Camps V.J.
      • Piñero D.P.
      • Caravaca-Arens E.
      • et al.
      New approach for correction of error associated with keratometric estimation of corneal power in keratoconus.
      seen in keratoconus, poor repeatability of biometry measurements in keratoconus eyes, and ELP models heavily dependent on the corneal power in their calculation.
      Although these issues present challenges in IOL power calculation, some formulas that are able to improve the accuracy of refractive outcomes, either accidentally or by design, do exist. The SRK/T formula has been shown to be the most accurate formula in previous keratoconus studies with less of a tendency toward a hyperopic refractive surprise. The phenomenon has been described by Melles et al,
      • Melles R.B.
      • Holladay J.T.
      • Chang W.J.
      Accuracy of intraocular lens calculation formulas.
      who showed that the SRK/T leads to significant myopic prediction errors in nonkeratoconus eyes as the average keratometry increases. This bias is likely a consequence of the nonphysiologic behavior in corneal height calculation, also referred to as the “cusp phenomenon” described by Haigis
      • Haigis W.
      Occurrence of erroneous anterior chamber depth in the SRK/T formula.
      and investigated by Sheard et al.
      • Sheard R.M.
      • Smith G.T.
      • Cooke D.L.
      Improving the prediction accuracy of the SRK/T formula: The T2 formula.
      The tendency of the SRK/T formula toward myopic prediction errors at higher corneal powers counterbalances the hyperopic tendency seen in patients with keratoconus, thereby reducing the mean prediction error and improving the accuracy of the SRK/T’s predictions in the population with keratoconus.
      The Kane keratoconus formula and the Holladay 2 formulas both have specific adjustments for use in patients with keratoconus. The Kane keratoconus formula aims to provide a more appropriate corneal power measurement and reduce the influence of corneal power on ELP prediction. The Holladay 2 keratoconus adjustment aims to differentiate a steep keratometry reading in a nonectatic, small anterior segment from a patient with keratoconus, presumably to ensure the ELP is not too affected by the high corneal power reading. In our study, the Kane keratoconus formula resulted in improved outcomes compared with the original Kane formula (P < 0.01), whereas the Holladay 2 resulted in worse outcomes when selecting the keratoconus option (P < 0.01). The Kane keratoconus formula performed especially well in stage 3 keratoconus with elimination of the mean prediction error to a small hyperopic bias of 0.02 D compared with hyperopic biases of between 1.72 and 3.02 D for the other formulas. Additionally, the Kane keratoconus formula also had the lowest standard deviation of the prediction error (1.21 D vs. 1.49–1.83 D), indicating a reduction in both the bias and spread of prediction errors. To avoid a hyperopic refractive result when using a third-generation IOL formula, we would recommend no adjustment to the target aim in stage 1 keratoconus, a more myopic target (between −0.75 and −1.5 D) for stage 2, and a more myopic target (between −2.0 and −3.0 D) for stage 3 keratoconus. These adjustments are less aggressive than those suggested by Watson et al.
      • Watson M.P.
      • Anand S.
      • Bhogal M.
      • et al.
      Cataract surgery outcome in eyes with keratoconus.
      Our study is significant because of the large sample size (twice as many patients as in any previous keratoconus study) and the inclusion of 2 keratoconus-specific formulas in the analysis. However, our study has several limitations in its design. First, inclusion of data using multiple IOL models limited our ability to perform IOL constant optimization, as suggested in an editorial by Hoffer et al.
      • Hoffer K.J.
      • Aramberri J.
      • Haigis W.
      • et al.
      Protocols for studies of intraocular lens formula accuracy.
      Our approach of using the ULIB constants follows the methodology of the work of Savini et al
      • Savini G.
      • Abbate R.
      • Hoffer K.J.
      • et al.
      Intraocular lens power calculation in eyes with keratoconus.
      (who also co-authored the editorial with Hoffer) and the suggestion by Wang et al
      • Wang L.
      • Koch D.D.
      • Hill W.
      • Abulafia A.
      Pursuing perfection in intraocular lens calculations: III. Criteria for analyzing outcomes.
      of using optimized constants from an overall dataset (e.g., ULIB) and not necessarily optimized constants from the subset of atypical eyes studied. Second, we measured corneal power using 2 versions of the IOLMaster (500 and 700), which may affect the outcomes. The IOLMaster 500 and IOLMaster 700 have differences in how they measure corneal power; however, the output for each machine has been shown to correlate extremely closely.
      • Akman A.
      • Asena L.
      • Güngör S.G.
      Evaluation and comparison of the new swept source OCT-based IOLMaster 700 with the IOLMaster 500.
      Third, our inability to determine the total corneal power by direct measurement of both the anterior and posterior corneal surface curvatures means we are unable to comment whether using “total corneal power” would have resulted in improved accuracy. Currently, the measurement of posterior corneal radii of curvature is a contentious issue with disagreement between measurements taken from different biometers and concerns regarding the validity of posterior corneal measurements.
      • Koch D.D.
      The posterior cornea: hiding in plain sight.
      In conclusion, the Kane keratoconus formula resulted in a reduction in MAE of 20% to 39% compared with other modern IOL formulas, suggesting there may be potential to improve refractive outcomes of cataract surgery in patients with keratoconus. Despite this, the results in patients with keratoconus remain significantly less accurate compared with patients without keratoconus, indicating the need for further research to improve the accuracy of IOL power calculation in these eyes.

      References

        • Kamiya K.
        • Iijima K.
        • Nobuyuki S.
        • et al.
        Predictability of Intraocular Lens Power Calculation for Cataract with Keratoconus: a multicenter study.
        Sci Rep. 2018; 8: 1312
        • Savini G.
        • Abbate R.
        • Hoffer K.J.
        • et al.
        Intraocular lens power calculation in eyes with keratoconus.
        J Cataract Refract Surg. 2019; 45: 576-581
        • Watson M.P.
        • Anand S.
        • Bhogal M.
        • et al.
        Cataract surgery outcome in eyes with keratoconus.
        Br J Ophthalmol. 2014; 98: 361-364
        • Melles R.B.
        • Holladay J.T.
        • Chang W.J.
        Accuracy of intraocular lens calculation formulas.
        Ophthalmology. 2018; 125: 169-178
        • Camps V.J.
        • Piñero D.P.
        • Caravaca-Arens E.
        • et al.
        New approach for correction of error associated with keratometric estimation of corneal power in keratoconus.
        Cornea. 2014; 33: 960-967
        • Holladay J.T.
        Holladay IOL Consultant Software & Surgical Outcomes Assessment. 1105.2019 ed.
        Holladay Consulting, Bellaire, TX2019
        • Connell B.J.
        • Kane J.X.
        Comparison of the Kane formula with existing formulas for intraocular lens power selection.
        BMJ Open Ophthalmol. 2019; 4 (e000251)
        • Kane J.X.
        Kane Formula.
        (Available at:)
        www.iolformula.com
        Date: 2019
        Date accessed: July 12, 2019
        • Hoffer K.J.
        • Aramberri J.
        • Haigis W.
        • et al.
        Protocols for studies of intraocular lens formula accuracy.
        Am J Ophthalmol. 2015; 160: 403-405.e1
        • Wang L.
        • Koch D.D.
        • Hill W.
        • Abulafia A.
        Pursuing perfection in intraocular lens calculations: III. Criteria for analyzing outcomes.
        J Cataract Refract Surg. 2017; 43: 999-1002
        • Haigis W.
        • Lege B.
        • Miller N.
        • Schneider B.
        Comparison of immersion ultrasound biometry and partial coherence interferometry for intraocular lens calculation according to Haigis.
        Graefe’s Arch Clin Exp Ophthalmol. 2000; 238: 765-773
        • Hoffer K.J.
        The Hoffer Q formula: a comparison of theoretic and regression formulas.
        J Cataract Refract Surg. 1993; 19 (errata, 1994;20:677; errata, 2007;33:2-3): 700-712
        • Holladay J.T.
        • Prager T.C.
        • Chandler T.Y.
        • et al.
        A three-part system for refining intraocular lens power calculations.
        J Cataract Refract Surg. 1988; 14: 17-24
        • Retzlaff J.A.
        • Sanders D.R.
        • Kraff M.C.
        Development of the SRK/T intraocular lens implant power calculation formula.
        J Cataract Refract Surg. 1990; 16 (erratum, 1990;16:528): 333-340
        • Barrett G.
        Barrett Universal II Formula.
        (Available at:)
        http://calc.apacrs.org/barrett_universal2105/
        Date: 2018
        Date accessed: July 13, 2019
        • Olsen T.
        • Hoffmann P.
        C constant: new concept for ray tracing-assisted intraocular lens power calculation.
        J Cataract Refract Surg. 2014; 40: 764-773
        • Haigis W.
        Optimized IOL constants for the ZEISS IOLMaster.
        (Available at:)
        http://ocusoft.de/ulib/c1.htm
        Date accessed: August 19, 2019
        • Krumeich J.H.
        • Daniel J.
        • Knalle A.
        Live-epikeratophakia for keratoconus.
        J Cataract Refract Surg. 1998; 24: 456-463
        • Aristodemou P.
        • Knox Cartwright N.E.
        • et al.
        Statistical analysis for studies of intraocular lens formula accuracy.
        Am J Ophthalmol. 2015; 160: 1085-1086
        • Melles R.B.
        • Kane J.X.
        • Olsen T.
        • Chang W.J.
        Update on intraocular lens calculation formulas.
        Ophthalmology. 2019; 126: 1334-1335
        • Haigis W.
        Occurrence of erroneous anterior chamber depth in the SRK/T formula.
        J Cataract Refract Surg. 1993; 19: 442-446
        • Sheard R.M.
        • Smith G.T.
        • Cooke D.L.
        Improving the prediction accuracy of the SRK/T formula: The T2 formula.
        J Cataract Refract Surg. 2010; 36: 1829-1834
        • Akman A.
        • Asena L.
        • Güngör S.G.
        Evaluation and comparison of the new swept source OCT-based IOLMaster 700 with the IOLMaster 500.
        Br J Ophthalmol. 2016; 100: 1201-1205
        • Koch D.D.
        The posterior cornea: hiding in plain sight.
        Ophthalmology. 2015; 122: 1070-1071

      Linked Article