User login
Cataract surgery is one of the most common ambulatory procedures performed in the US.1-3 With the aging of the US population, the number of Americans with cataracts is projected to increase from 24.4 million in 2010 to 38.7 million in 2030.4
Approximately 20% of all cataract patients have preoperative astigmatism of > 1.5 diopters (D), underscoring the importance of training residents in the placement of toric intraocular lenses (IOLs).5 However, the implantation of toric IOLs is more challenging than monofocal IOLs, requiring precise surgical alignment of the IOL.6 Successful toric IOL implantation also requires accurate calculation of the IOL cylinder power and target axis of alignment. It is unclear which toric IOL calculation formula offers the most accurate refractive predictions, and practitioners have designed strategies to apply different formulae depending on the biometric dimensions of the target eye.7-9
Previous studies of resident-performed cataract surgery using toric IOLs6,10-13 and studies that compare the performance of the Barrett and Holladay toric formulae have been limited by their small sample sizes (< 107 eyes).7,14-16 Moreover, none of the studies that evaluate the comparative effectiveness of these biometric formulae were conducted at a teaching hospital.7,14-16
Given the added complexity of toric IOL placement and variable surgical experience of residents as ophthalmologists-in-training, it is important to assess outcomes in teaching hospitals.13 The primary aims of this study were to assess the visual and refractive outcomes of cataract surgery using toric IOLs in a US Department of Veterans Affairs (VA) teaching hospital and to compare the relative accuracy of the Holladay 2 or Barrett toric biometric formulae in predicting postoperative refraction outcomes.
Methods
The Providence VA Medical Center (PVAMC) Institutional Review Board approved this study. This retrospective chart review included patients with cataract and corneal astigmatism who underwent cataract surgery using Acrysof toric IOLs, model SN6AT (Alcon) at the PVAMC teaching hospital between November 2013 and May 2018.
Only 1 eye was included from each study subject to avoid compounding of data with the use of bilateral eyes.17 In addition, bilateral cataract surgery was only performed on some patients at the PVAMC, so including both eyes from eligible patients would disproportionately weigh those patients’ outcomes. If both eyes had cataract surgery and their postoperative visual acuities were unequal, we chose the eye with the better postoperative visual acuity since refraction accuracy decreases with worsening best-corrected visual acuity (BCVA). If both eyes had cataract surgery and the postoperative visual acuity was the same, the first operated eye was chosen.17,18
Exclusion criteria included worse than 20/40 BCVA, posterior capsular rupture, sulcus IOL, history of corneal disease, history of refractive surgery (laser-assisted in situ keratomileusis [LASIK]/photorefractive keratectomy [PRK]), axial length not measurable by the Lenstar optical biometer (Haag-Streit USA), or no postoperative refraction within 3 weeks to 4 months.19,20
Patient age, race/ethnicity, gender, preoperative refraction, preoperative BCVA, postoperative refraction, postoperative BCVA, and IOL power were recorded from patient charts (Table 1). Preoperative and postoperative refractive values were converted to spherical equivalents. The preoperative biometry and most of the postoperative refractions were performed by experienced technicians certified by the Joint Commission on Allied Health Personnel in Ophthalmology. The main outcomes for the assessment of surgeries included the postoperative BCVA, postoperative spherical equivalent refraction, and postoperative residual refractive astigmatism.
Axial length (AL), preoperative anterior chamber depth (ACD), preoperative flat corneal front power (K1), preoperative steep corneal front power (K2), lens thickness, horizontal white-to-white (WTW) corneal diameter, and central corneal thickness (CCT) were recorded from the Lenstar biometric device. Predicted postoperative refractions for the Holladay 2 formula were calculated using Holladay IOL Consultant software (Holladay Consulting). Predicted postoperative refractions for the Barrett toric IOL formula were calculated using the online Barrett toric formula calculator.21 Since previous studies have shown that both the Holladay and Barrett formulae account for posterior corneal astigmatism, a comparison of refractive outcomes in eyes with against-the-rule astigmatism vs with-the-rule astigmatism was not performed.14 An estimated standardized value for surgically-induced astigmatism was entered into both formulae; 0.3 diopter (D) was chosen based on previously published averages.22-24
A formula’s prediction error is defined as the predicted postoperative refraction minus the actual postoperative refraction. The mean absolute prediction error (MAE), defined as the mean of the absolute values of the prediction errors, and the median absolute prediction error (MedAE), defined as the median of the absolute values of the prediction errors, were used to assess the overall accuracy of each formula. Also, the percentages of eyes with postoperative refraction within ≥ 0.25 D, ≥ 0.50 D, and ≥ 1.0 D were calculated for both formulae. Two-tailed t tests were performed to compare the MAE between the formulae. Subgroup analyses were performed for short eyes (AL < 22 mm), medium length eyes (AL = 22-25 mm), and long eyes (AL > 25 mm). Statistical analysis was performed using STATA 11 (STATA Corp). The preoperative corneal astigmatism and postoperative refractive astigmatism of all the cases were compared in double-angle plots to assess how well the toric IOL neutralized the corneal astigmatism.
Results
Of 325 charts reviewed during the study period, 34 patients were excluded due to lack of postoperative refraction within the designated follow-up period, 5 for worse than 20/40 postoperative BCVA (4 had preexisting ocular disease), 2 for complications, and 1 for missing data. We included 283 eyes from 283 patients in the final study. Resident ophthalmologists were the primary surgeons in 87.6% (248/283) of the cases.
The median postoperative BCVA was 20/20, and 92% of patients had a postoperative BCVA of 20/25 or better. The prediction outcomes of the toric SN6AT IOLs are shown in Table 2. The Barrett toric formula had a lower MAE than the Holladay 2 formula, but this difference was not statistically significant. The Barrett toric formula also predicted a higher percentage of eyes with postoperative refraction within ≥ 0.25 D (53.2%), ≥ 0.5 D (77.3%), and ≥ 1.0 D (96.1%). For both formulae, > 95% of eyes had prediction errors that fell within 1.0 D.
While the Barrett formula demonstrated a lower MAE in all 3 AL groups, no statistically significant differences were found between the Barrett and Holladay formulae (P = .94, P = .49, and P = .08 for short, medium, and long eyes, respectively). Both formulae produced the lowest MAE in the long AL group: Barrett had a MAE of 0.221 D and Holladay 2 had one of 0.329 D. The Barrett formula produced its highest percentage of eyes with prediction errors falling within 0.25 D and 0.5 D in the long AL group. In comparison, both formulae had the highest MAEs in the short AL group (Barrett toric, 0.598 D; Holladay 2, 0.613 D) and produced the lowest percentage of eyes with prediction errors falling within ≥ 0.25 D and ≥ 0.5 D in the short AL group.
A cumulative histogram of the preoperative corneal and postoperative refractive astigmatism magnitude is shown in Figure 1. The same data are presented as double-angle plots in the Appendix, which shows that the centroid values for preoperative corneal astigmatism were greatlyreduced when compared with the postoperative refractive astigmatism (mean absolute value of 1.77 D ≥ 0.73 D to 0.5 D ≥ 0.50 D).
Preoperative corneal astigmatism and postoperative refractive astigmatism were compared since preoperative refractive astigmatism has noncorneal contributions, including lenticular astigmatism, and there is minimal expected change between preoperative and postoperative corneal astigmatism.14 For comparison, double-angle plots of postoperative refractive astigmatism prediction errors for the Holladay and Barrett formulae are shown in Figure 2.
Discussion
To our knowledge, this is the largest study of resident-performed cataract surgery using toric IOLs, the largest study that compared the performance of the Barrett toric and Holladay 2 formulae, and the first that compared these formulae in a teaching hospital setting. This study found no significant difference in the predictive accuracy of the Barrett and Holladay 2 biometric formulae for cataract surgery using toric IOLs. In addition, our refractive outcomes were consistent with the results of previous toric IOL outcome studies conducted in teaching and nonteaching hospital settings.6,10-13
In 4 previous studies that compared the MAE of the Barrett and Holladay formulae for toric IOLs, the Barrett formula produced a lower MAE than the Holladay 2 formula.7,14-16 However, this difference was significant in only 2 of the studies, which had sample sizes of only 68 and 107 eyes.14,16 Furthermore, the Barrett toric formula produced the lower MAE for the entire AL range, though this was not statistically significant at our sample size. In addition, both formulae produced the lowest MAE in the long AL group and the highest MAE in the short AL group. The unique anatomy and high IOL power needed in short eyes may explain the challenges in attaining accurate IOL power predictions in this AL group.19,25
Limitations
The sample size of this study may have prevented us from detecting statistically significant differences in the performance of the Barrett and Holladay formulae. However, our findings are consistent with studies that compare the accuracy of these formulae in teaching and nonteaching hospital settings. Second, the study was conducted at a VA hospital, and a high proportion of patients were male; thus, our findings may not be generalizable to patients who receive cataract surgery with toric IOLs in other settings.
Conclusions
In a single VA teaching hospital, the Barrett and Holladay 2 biometric formulae provide similar refractive predictions for cataract surgery using toric IOLs. Larger studies would be necessary to detect smaller differences in the relative performance of the biometric formulae.
1. Schein OD, Cassard SD, Tielsch JM, Gower EW. Cataract surgery among Medicare beneficiaries. Ophthalmic Epidemiol. 2012;19(5):257-264.
2. Congdon N, O’Colmain B, Klaver CC, et al. Causes and prevalence of visual impairment among adults in the United States. Arch Ophthalmol. 2004;122(4):477-485.
3. Congdon N, Vingerling JR, Klein BE, et al. Prevalence of cataract and pseudophakia/aphakia among adults in the United States. Arch Ophthalmol. 2004;122(4):487-494.
4. National Eye Institute. Cataract tables: cataract defined. https://www.nei.nih.gov/learn-about-eye-health/resources-for-health-educators/eye-health-data-and-statistics/cataract-data-and-statistics/cataract-tables. Updated February 7, 2020. Accessed February 10, 2020.
5. Ostri C, Falck L, Boberg-Ans G, Kessel L. The need for toric intra-ocular lens implantation in public ophthalmology departments. Acta Ophthalmol. 2015;93(5):e396-e397.
6. Sundy M, McKnight D, Eck C, Rieger F 3rd. Visual acuity outcomes of toric lens implantation in patients undergoing cataract surgery at a residency training program. Mo Med. 2016;113(1):40-43.
7. Ferreira TB, Ribeiro P, Ribeiro FJ, O’Neill JG. Comparison of methodologies using estimated or measured values of total corneal astigmatism for toric intraocular lens power calculation. J Refract Surg. 2017;33(12):794-800.
8. Reitblat O, Levy A, Kleinmann G, Abulafia A, Assia EI. Effect of posterior corneal astigmatism on power calculation and alignment of toric intraocular lenses: comparison of methodologies. J Cataract Refract Surg. 2016;42(2):217-225.
9. 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(1):63-71.
10. Moreira HR, Hatch KM, Greenberg PB. Benchmarking outcomes in resident-performed cataract surgery with toric intraocular lenses [published correction appears in: Clin Experiment Ophthalmol. 2013;41(8):819]. Clin Exp Ophthalmol. 2013;41(6):624-626.
11. Retzlaff JA, Sanders DR, Kraff MC. Development of the SRK/T intraocular lens implant power calculation formula [published correction appears in: J Cataract Refract Surg. 1990;16(4):528]. J Cataract Refract Surg. 1990;16(3):333-340.
12. Roensch MA, Charton JW, Blomquist PH, Aggarwal NK, McCulley JP. Resident experience with toric and multifocal intraocular lenses in a public county hospital system. J Cataract Refract Surg. 2012;38(5):793-798.
13. Pouyeh B, Galor A, Junk AK, et al. Surgical and refractive outcomes of cataract surgery with toric intraocular lens implantation at a resident-teaching institution. J Cataract Refract Surg. 2011;37(9):1623-1628.
14. Ferreira TB, Ribeiro P, Ribeiro FJ, O’Neill JG. Comparison of astigmatic prediction errors associated with new calculation methods for toric intraocular lenses. J Cataract Refract Surg. 2017;43(3):340-347.
15. Abulafia A, Hill WE, Franchina M, Barrett GD. Comparison of methods to predict residual astigmatism after intraocular lens implantation. J Refract Surg. 2015;31(10):699-707.
16. Abulafia A, Barrett GD, Kleinmann G, et al. Prediction of refractive outcomes with toric intraocular lens implantation. J Cataract Refract Surg. 2015;41(5):936-944.
17. Wang Q, Jiang W, Lin T, Wu X, Lin H, Chen W. Meta-analysis of accuracy of intraocular lens power calculation formulas in short eyes. Clin Exp Ophthalmol. 2018;46(4):356-363.
18. Melles RB, Holladay JT, Chang WJ. Accuracy of intraocular lens calculation formulas. Ophthalmology. 2018;125(2):169-178.
19. Hoffer KJ. The Hoffer Q formula: a comparison of theoretic and regression formulas. J Cataract Refract Surg. 1993;19(6):700-712.
20. Cooke DL, Cooke TL. Comparison of 9 intraocular lens power calculation formulas. J Cataract Refract Surg. 2016;42(8):1157-1164.
21. American Society of Cataract and Refractive Surgery. Barrett toric calculator. www.ascrs.org/barrett-toric-calculator. Accessed February 5, 2020.
22. Holladay JT, Pettit G. Improving toric intraocular lens calculations using total surgically induced astigmatism for a 2.5 mm temporal incision. J Cataract Refract Surg. 2019;45(3):272-283.
23. Canovas C, Alarcon A, Rosén R, et al. New algorithm for toric intraocular lens power calculation considering the posterior corneal astigmatism. J Cataract Refract Surg. 2018;44(2):168-174.
24. Visser N, Berendschot TT, Bauer NJ, Nuijts RM. Vector analysis of corneal and refractive astigmatism changes following toric pseudophakic and toric phakic IOL implantation. Invest Ophthalmol Vis Sci. 2012;53(4):1865-1873.
25. Narváez 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(12):2050-2053.
Cataract surgery is one of the most common ambulatory procedures performed in the US.1-3 With the aging of the US population, the number of Americans with cataracts is projected to increase from 24.4 million in 2010 to 38.7 million in 2030.4
Approximately 20% of all cataract patients have preoperative astigmatism of > 1.5 diopters (D), underscoring the importance of training residents in the placement of toric intraocular lenses (IOLs).5 However, the implantation of toric IOLs is more challenging than monofocal IOLs, requiring precise surgical alignment of the IOL.6 Successful toric IOL implantation also requires accurate calculation of the IOL cylinder power and target axis of alignment. It is unclear which toric IOL calculation formula offers the most accurate refractive predictions, and practitioners have designed strategies to apply different formulae depending on the biometric dimensions of the target eye.7-9
Previous studies of resident-performed cataract surgery using toric IOLs6,10-13 and studies that compare the performance of the Barrett and Holladay toric formulae have been limited by their small sample sizes (< 107 eyes).7,14-16 Moreover, none of the studies that evaluate the comparative effectiveness of these biometric formulae were conducted at a teaching hospital.7,14-16
Given the added complexity of toric IOL placement and variable surgical experience of residents as ophthalmologists-in-training, it is important to assess outcomes in teaching hospitals.13 The primary aims of this study were to assess the visual and refractive outcomes of cataract surgery using toric IOLs in a US Department of Veterans Affairs (VA) teaching hospital and to compare the relative accuracy of the Holladay 2 or Barrett toric biometric formulae in predicting postoperative refraction outcomes.
Methods
The Providence VA Medical Center (PVAMC) Institutional Review Board approved this study. This retrospective chart review included patients with cataract and corneal astigmatism who underwent cataract surgery using Acrysof toric IOLs, model SN6AT (Alcon) at the PVAMC teaching hospital between November 2013 and May 2018.
Only 1 eye was included from each study subject to avoid compounding of data with the use of bilateral eyes.17 In addition, bilateral cataract surgery was only performed on some patients at the PVAMC, so including both eyes from eligible patients would disproportionately weigh those patients’ outcomes. If both eyes had cataract surgery and their postoperative visual acuities were unequal, we chose the eye with the better postoperative visual acuity since refraction accuracy decreases with worsening best-corrected visual acuity (BCVA). If both eyes had cataract surgery and the postoperative visual acuity was the same, the first operated eye was chosen.17,18
Exclusion criteria included worse than 20/40 BCVA, posterior capsular rupture, sulcus IOL, history of corneal disease, history of refractive surgery (laser-assisted in situ keratomileusis [LASIK]/photorefractive keratectomy [PRK]), axial length not measurable by the Lenstar optical biometer (Haag-Streit USA), or no postoperative refraction within 3 weeks to 4 months.19,20
Patient age, race/ethnicity, gender, preoperative refraction, preoperative BCVA, postoperative refraction, postoperative BCVA, and IOL power were recorded from patient charts (Table 1). Preoperative and postoperative refractive values were converted to spherical equivalents. The preoperative biometry and most of the postoperative refractions were performed by experienced technicians certified by the Joint Commission on Allied Health Personnel in Ophthalmology. The main outcomes for the assessment of surgeries included the postoperative BCVA, postoperative spherical equivalent refraction, and postoperative residual refractive astigmatism.
Axial length (AL), preoperative anterior chamber depth (ACD), preoperative flat corneal front power (K1), preoperative steep corneal front power (K2), lens thickness, horizontal white-to-white (WTW) corneal diameter, and central corneal thickness (CCT) were recorded from the Lenstar biometric device. Predicted postoperative refractions for the Holladay 2 formula were calculated using Holladay IOL Consultant software (Holladay Consulting). Predicted postoperative refractions for the Barrett toric IOL formula were calculated using the online Barrett toric formula calculator.21 Since previous studies have shown that both the Holladay and Barrett formulae account for posterior corneal astigmatism, a comparison of refractive outcomes in eyes with against-the-rule astigmatism vs with-the-rule astigmatism was not performed.14 An estimated standardized value for surgically-induced astigmatism was entered into both formulae; 0.3 diopter (D) was chosen based on previously published averages.22-24
A formula’s prediction error is defined as the predicted postoperative refraction minus the actual postoperative refraction. The mean absolute prediction error (MAE), defined as the mean of the absolute values of the prediction errors, and the median absolute prediction error (MedAE), defined as the median of the absolute values of the prediction errors, were used to assess the overall accuracy of each formula. Also, the percentages of eyes with postoperative refraction within ≥ 0.25 D, ≥ 0.50 D, and ≥ 1.0 D were calculated for both formulae. Two-tailed t tests were performed to compare the MAE between the formulae. Subgroup analyses were performed for short eyes (AL < 22 mm), medium length eyes (AL = 22-25 mm), and long eyes (AL > 25 mm). Statistical analysis was performed using STATA 11 (STATA Corp). The preoperative corneal astigmatism and postoperative refractive astigmatism of all the cases were compared in double-angle plots to assess how well the toric IOL neutralized the corneal astigmatism.
Results
Of 325 charts reviewed during the study period, 34 patients were excluded due to lack of postoperative refraction within the designated follow-up period, 5 for worse than 20/40 postoperative BCVA (4 had preexisting ocular disease), 2 for complications, and 1 for missing data. We included 283 eyes from 283 patients in the final study. Resident ophthalmologists were the primary surgeons in 87.6% (248/283) of the cases.
The median postoperative BCVA was 20/20, and 92% of patients had a postoperative BCVA of 20/25 or better. The prediction outcomes of the toric SN6AT IOLs are shown in Table 2. The Barrett toric formula had a lower MAE than the Holladay 2 formula, but this difference was not statistically significant. The Barrett toric formula also predicted a higher percentage of eyes with postoperative refraction within ≥ 0.25 D (53.2%), ≥ 0.5 D (77.3%), and ≥ 1.0 D (96.1%). For both formulae, > 95% of eyes had prediction errors that fell within 1.0 D.
While the Barrett formula demonstrated a lower MAE in all 3 AL groups, no statistically significant differences were found between the Barrett and Holladay formulae (P = .94, P = .49, and P = .08 for short, medium, and long eyes, respectively). Both formulae produced the lowest MAE in the long AL group: Barrett had a MAE of 0.221 D and Holladay 2 had one of 0.329 D. The Barrett formula produced its highest percentage of eyes with prediction errors falling within 0.25 D and 0.5 D in the long AL group. In comparison, both formulae had the highest MAEs in the short AL group (Barrett toric, 0.598 D; Holladay 2, 0.613 D) and produced the lowest percentage of eyes with prediction errors falling within ≥ 0.25 D and ≥ 0.5 D in the short AL group.
A cumulative histogram of the preoperative corneal and postoperative refractive astigmatism magnitude is shown in Figure 1. The same data are presented as double-angle plots in the Appendix, which shows that the centroid values for preoperative corneal astigmatism were greatlyreduced when compared with the postoperative refractive astigmatism (mean absolute value of 1.77 D ≥ 0.73 D to 0.5 D ≥ 0.50 D).
Preoperative corneal astigmatism and postoperative refractive astigmatism were compared since preoperative refractive astigmatism has noncorneal contributions, including lenticular astigmatism, and there is minimal expected change between preoperative and postoperative corneal astigmatism.14 For comparison, double-angle plots of postoperative refractive astigmatism prediction errors for the Holladay and Barrett formulae are shown in Figure 2.
Discussion
To our knowledge, this is the largest study of resident-performed cataract surgery using toric IOLs, the largest study that compared the performance of the Barrett toric and Holladay 2 formulae, and the first that compared these formulae in a teaching hospital setting. This study found no significant difference in the predictive accuracy of the Barrett and Holladay 2 biometric formulae for cataract surgery using toric IOLs. In addition, our refractive outcomes were consistent with the results of previous toric IOL outcome studies conducted in teaching and nonteaching hospital settings.6,10-13
In 4 previous studies that compared the MAE of the Barrett and Holladay formulae for toric IOLs, the Barrett formula produced a lower MAE than the Holladay 2 formula.7,14-16 However, this difference was significant in only 2 of the studies, which had sample sizes of only 68 and 107 eyes.14,16 Furthermore, the Barrett toric formula produced the lower MAE for the entire AL range, though this was not statistically significant at our sample size. In addition, both formulae produced the lowest MAE in the long AL group and the highest MAE in the short AL group. The unique anatomy and high IOL power needed in short eyes may explain the challenges in attaining accurate IOL power predictions in this AL group.19,25
Limitations
The sample size of this study may have prevented us from detecting statistically significant differences in the performance of the Barrett and Holladay formulae. However, our findings are consistent with studies that compare the accuracy of these formulae in teaching and nonteaching hospital settings. Second, the study was conducted at a VA hospital, and a high proportion of patients were male; thus, our findings may not be generalizable to patients who receive cataract surgery with toric IOLs in other settings.
Conclusions
In a single VA teaching hospital, the Barrett and Holladay 2 biometric formulae provide similar refractive predictions for cataract surgery using toric IOLs. Larger studies would be necessary to detect smaller differences in the relative performance of the biometric formulae.
Cataract surgery is one of the most common ambulatory procedures performed in the US.1-3 With the aging of the US population, the number of Americans with cataracts is projected to increase from 24.4 million in 2010 to 38.7 million in 2030.4
Approximately 20% of all cataract patients have preoperative astigmatism of > 1.5 diopters (D), underscoring the importance of training residents in the placement of toric intraocular lenses (IOLs).5 However, the implantation of toric IOLs is more challenging than monofocal IOLs, requiring precise surgical alignment of the IOL.6 Successful toric IOL implantation also requires accurate calculation of the IOL cylinder power and target axis of alignment. It is unclear which toric IOL calculation formula offers the most accurate refractive predictions, and practitioners have designed strategies to apply different formulae depending on the biometric dimensions of the target eye.7-9
Previous studies of resident-performed cataract surgery using toric IOLs6,10-13 and studies that compare the performance of the Barrett and Holladay toric formulae have been limited by their small sample sizes (< 107 eyes).7,14-16 Moreover, none of the studies that evaluate the comparative effectiveness of these biometric formulae were conducted at a teaching hospital.7,14-16
Given the added complexity of toric IOL placement and variable surgical experience of residents as ophthalmologists-in-training, it is important to assess outcomes in teaching hospitals.13 The primary aims of this study were to assess the visual and refractive outcomes of cataract surgery using toric IOLs in a US Department of Veterans Affairs (VA) teaching hospital and to compare the relative accuracy of the Holladay 2 or Barrett toric biometric formulae in predicting postoperative refraction outcomes.
Methods
The Providence VA Medical Center (PVAMC) Institutional Review Board approved this study. This retrospective chart review included patients with cataract and corneal astigmatism who underwent cataract surgery using Acrysof toric IOLs, model SN6AT (Alcon) at the PVAMC teaching hospital between November 2013 and May 2018.
Only 1 eye was included from each study subject to avoid compounding of data with the use of bilateral eyes.17 In addition, bilateral cataract surgery was only performed on some patients at the PVAMC, so including both eyes from eligible patients would disproportionately weigh those patients’ outcomes. If both eyes had cataract surgery and their postoperative visual acuities were unequal, we chose the eye with the better postoperative visual acuity since refraction accuracy decreases with worsening best-corrected visual acuity (BCVA). If both eyes had cataract surgery and the postoperative visual acuity was the same, the first operated eye was chosen.17,18
Exclusion criteria included worse than 20/40 BCVA, posterior capsular rupture, sulcus IOL, history of corneal disease, history of refractive surgery (laser-assisted in situ keratomileusis [LASIK]/photorefractive keratectomy [PRK]), axial length not measurable by the Lenstar optical biometer (Haag-Streit USA), or no postoperative refraction within 3 weeks to 4 months.19,20
Patient age, race/ethnicity, gender, preoperative refraction, preoperative BCVA, postoperative refraction, postoperative BCVA, and IOL power were recorded from patient charts (Table 1). Preoperative and postoperative refractive values were converted to spherical equivalents. The preoperative biometry and most of the postoperative refractions were performed by experienced technicians certified by the Joint Commission on Allied Health Personnel in Ophthalmology. The main outcomes for the assessment of surgeries included the postoperative BCVA, postoperative spherical equivalent refraction, and postoperative residual refractive astigmatism.
Axial length (AL), preoperative anterior chamber depth (ACD), preoperative flat corneal front power (K1), preoperative steep corneal front power (K2), lens thickness, horizontal white-to-white (WTW) corneal diameter, and central corneal thickness (CCT) were recorded from the Lenstar biometric device. Predicted postoperative refractions for the Holladay 2 formula were calculated using Holladay IOL Consultant software (Holladay Consulting). Predicted postoperative refractions for the Barrett toric IOL formula were calculated using the online Barrett toric formula calculator.21 Since previous studies have shown that both the Holladay and Barrett formulae account for posterior corneal astigmatism, a comparison of refractive outcomes in eyes with against-the-rule astigmatism vs with-the-rule astigmatism was not performed.14 An estimated standardized value for surgically-induced astigmatism was entered into both formulae; 0.3 diopter (D) was chosen based on previously published averages.22-24
A formula’s prediction error is defined as the predicted postoperative refraction minus the actual postoperative refraction. The mean absolute prediction error (MAE), defined as the mean of the absolute values of the prediction errors, and the median absolute prediction error (MedAE), defined as the median of the absolute values of the prediction errors, were used to assess the overall accuracy of each formula. Also, the percentages of eyes with postoperative refraction within ≥ 0.25 D, ≥ 0.50 D, and ≥ 1.0 D were calculated for both formulae. Two-tailed t tests were performed to compare the MAE between the formulae. Subgroup analyses were performed for short eyes (AL < 22 mm), medium length eyes (AL = 22-25 mm), and long eyes (AL > 25 mm). Statistical analysis was performed using STATA 11 (STATA Corp). The preoperative corneal astigmatism and postoperative refractive astigmatism of all the cases were compared in double-angle plots to assess how well the toric IOL neutralized the corneal astigmatism.
Results
Of 325 charts reviewed during the study period, 34 patients were excluded due to lack of postoperative refraction within the designated follow-up period, 5 for worse than 20/40 postoperative BCVA (4 had preexisting ocular disease), 2 for complications, and 1 for missing data. We included 283 eyes from 283 patients in the final study. Resident ophthalmologists were the primary surgeons in 87.6% (248/283) of the cases.
The median postoperative BCVA was 20/20, and 92% of patients had a postoperative BCVA of 20/25 or better. The prediction outcomes of the toric SN6AT IOLs are shown in Table 2. The Barrett toric formula had a lower MAE than the Holladay 2 formula, but this difference was not statistically significant. The Barrett toric formula also predicted a higher percentage of eyes with postoperative refraction within ≥ 0.25 D (53.2%), ≥ 0.5 D (77.3%), and ≥ 1.0 D (96.1%). For both formulae, > 95% of eyes had prediction errors that fell within 1.0 D.
While the Barrett formula demonstrated a lower MAE in all 3 AL groups, no statistically significant differences were found between the Barrett and Holladay formulae (P = .94, P = .49, and P = .08 for short, medium, and long eyes, respectively). Both formulae produced the lowest MAE in the long AL group: Barrett had a MAE of 0.221 D and Holladay 2 had one of 0.329 D. The Barrett formula produced its highest percentage of eyes with prediction errors falling within 0.25 D and 0.5 D in the long AL group. In comparison, both formulae had the highest MAEs in the short AL group (Barrett toric, 0.598 D; Holladay 2, 0.613 D) and produced the lowest percentage of eyes with prediction errors falling within ≥ 0.25 D and ≥ 0.5 D in the short AL group.
A cumulative histogram of the preoperative corneal and postoperative refractive astigmatism magnitude is shown in Figure 1. The same data are presented as double-angle plots in the Appendix, which shows that the centroid values for preoperative corneal astigmatism were greatlyreduced when compared with the postoperative refractive astigmatism (mean absolute value of 1.77 D ≥ 0.73 D to 0.5 D ≥ 0.50 D).
Preoperative corneal astigmatism and postoperative refractive astigmatism were compared since preoperative refractive astigmatism has noncorneal contributions, including lenticular astigmatism, and there is minimal expected change between preoperative and postoperative corneal astigmatism.14 For comparison, double-angle plots of postoperative refractive astigmatism prediction errors for the Holladay and Barrett formulae are shown in Figure 2.
Discussion
To our knowledge, this is the largest study of resident-performed cataract surgery using toric IOLs, the largest study that compared the performance of the Barrett toric and Holladay 2 formulae, and the first that compared these formulae in a teaching hospital setting. This study found no significant difference in the predictive accuracy of the Barrett and Holladay 2 biometric formulae for cataract surgery using toric IOLs. In addition, our refractive outcomes were consistent with the results of previous toric IOL outcome studies conducted in teaching and nonteaching hospital settings.6,10-13
In 4 previous studies that compared the MAE of the Barrett and Holladay formulae for toric IOLs, the Barrett formula produced a lower MAE than the Holladay 2 formula.7,14-16 However, this difference was significant in only 2 of the studies, which had sample sizes of only 68 and 107 eyes.14,16 Furthermore, the Barrett toric formula produced the lower MAE for the entire AL range, though this was not statistically significant at our sample size. In addition, both formulae produced the lowest MAE in the long AL group and the highest MAE in the short AL group. The unique anatomy and high IOL power needed in short eyes may explain the challenges in attaining accurate IOL power predictions in this AL group.19,25
Limitations
The sample size of this study may have prevented us from detecting statistically significant differences in the performance of the Barrett and Holladay formulae. However, our findings are consistent with studies that compare the accuracy of these formulae in teaching and nonteaching hospital settings. Second, the study was conducted at a VA hospital, and a high proportion of patients were male; thus, our findings may not be generalizable to patients who receive cataract surgery with toric IOLs in other settings.
Conclusions
In a single VA teaching hospital, the Barrett and Holladay 2 biometric formulae provide similar refractive predictions for cataract surgery using toric IOLs. Larger studies would be necessary to detect smaller differences in the relative performance of the biometric formulae.
1. Schein OD, Cassard SD, Tielsch JM, Gower EW. Cataract surgery among Medicare beneficiaries. Ophthalmic Epidemiol. 2012;19(5):257-264.
2. Congdon N, O’Colmain B, Klaver CC, et al. Causes and prevalence of visual impairment among adults in the United States. Arch Ophthalmol. 2004;122(4):477-485.
3. Congdon N, Vingerling JR, Klein BE, et al. Prevalence of cataract and pseudophakia/aphakia among adults in the United States. Arch Ophthalmol. 2004;122(4):487-494.
4. National Eye Institute. Cataract tables: cataract defined. https://www.nei.nih.gov/learn-about-eye-health/resources-for-health-educators/eye-health-data-and-statistics/cataract-data-and-statistics/cataract-tables. Updated February 7, 2020. Accessed February 10, 2020.
5. Ostri C, Falck L, Boberg-Ans G, Kessel L. The need for toric intra-ocular lens implantation in public ophthalmology departments. Acta Ophthalmol. 2015;93(5):e396-e397.
6. Sundy M, McKnight D, Eck C, Rieger F 3rd. Visual acuity outcomes of toric lens implantation in patients undergoing cataract surgery at a residency training program. Mo Med. 2016;113(1):40-43.
7. Ferreira TB, Ribeiro P, Ribeiro FJ, O’Neill JG. Comparison of methodologies using estimated or measured values of total corneal astigmatism for toric intraocular lens power calculation. J Refract Surg. 2017;33(12):794-800.
8. Reitblat O, Levy A, Kleinmann G, Abulafia A, Assia EI. Effect of posterior corneal astigmatism on power calculation and alignment of toric intraocular lenses: comparison of methodologies. J Cataract Refract Surg. 2016;42(2):217-225.
9. 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(1):63-71.
10. Moreira HR, Hatch KM, Greenberg PB. Benchmarking outcomes in resident-performed cataract surgery with toric intraocular lenses [published correction appears in: Clin Experiment Ophthalmol. 2013;41(8):819]. Clin Exp Ophthalmol. 2013;41(6):624-626.
11. Retzlaff JA, Sanders DR, Kraff MC. Development of the SRK/T intraocular lens implant power calculation formula [published correction appears in: J Cataract Refract Surg. 1990;16(4):528]. J Cataract Refract Surg. 1990;16(3):333-340.
12. Roensch MA, Charton JW, Blomquist PH, Aggarwal NK, McCulley JP. Resident experience with toric and multifocal intraocular lenses in a public county hospital system. J Cataract Refract Surg. 2012;38(5):793-798.
13. Pouyeh B, Galor A, Junk AK, et al. Surgical and refractive outcomes of cataract surgery with toric intraocular lens implantation at a resident-teaching institution. J Cataract Refract Surg. 2011;37(9):1623-1628.
14. Ferreira TB, Ribeiro P, Ribeiro FJ, O’Neill JG. Comparison of astigmatic prediction errors associated with new calculation methods for toric intraocular lenses. J Cataract Refract Surg. 2017;43(3):340-347.
15. Abulafia A, Hill WE, Franchina M, Barrett GD. Comparison of methods to predict residual astigmatism after intraocular lens implantation. J Refract Surg. 2015;31(10):699-707.
16. Abulafia A, Barrett GD, Kleinmann G, et al. Prediction of refractive outcomes with toric intraocular lens implantation. J Cataract Refract Surg. 2015;41(5):936-944.
17. Wang Q, Jiang W, Lin T, Wu X, Lin H, Chen W. Meta-analysis of accuracy of intraocular lens power calculation formulas in short eyes. Clin Exp Ophthalmol. 2018;46(4):356-363.
18. Melles RB, Holladay JT, Chang WJ. Accuracy of intraocular lens calculation formulas. Ophthalmology. 2018;125(2):169-178.
19. Hoffer KJ. The Hoffer Q formula: a comparison of theoretic and regression formulas. J Cataract Refract Surg. 1993;19(6):700-712.
20. Cooke DL, Cooke TL. Comparison of 9 intraocular lens power calculation formulas. J Cataract Refract Surg. 2016;42(8):1157-1164.
21. American Society of Cataract and Refractive Surgery. Barrett toric calculator. www.ascrs.org/barrett-toric-calculator. Accessed February 5, 2020.
22. Holladay JT, Pettit G. Improving toric intraocular lens calculations using total surgically induced astigmatism for a 2.5 mm temporal incision. J Cataract Refract Surg. 2019;45(3):272-283.
23. Canovas C, Alarcon A, Rosén R, et al. New algorithm for toric intraocular lens power calculation considering the posterior corneal astigmatism. J Cataract Refract Surg. 2018;44(2):168-174.
24. Visser N, Berendschot TT, Bauer NJ, Nuijts RM. Vector analysis of corneal and refractive astigmatism changes following toric pseudophakic and toric phakic IOL implantation. Invest Ophthalmol Vis Sci. 2012;53(4):1865-1873.
25. Narváez 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(12):2050-2053.
1. Schein OD, Cassard SD, Tielsch JM, Gower EW. Cataract surgery among Medicare beneficiaries. Ophthalmic Epidemiol. 2012;19(5):257-264.
2. Congdon N, O’Colmain B, Klaver CC, et al. Causes and prevalence of visual impairment among adults in the United States. Arch Ophthalmol. 2004;122(4):477-485.
3. Congdon N, Vingerling JR, Klein BE, et al. Prevalence of cataract and pseudophakia/aphakia among adults in the United States. Arch Ophthalmol. 2004;122(4):487-494.
4. National Eye Institute. Cataract tables: cataract defined. https://www.nei.nih.gov/learn-about-eye-health/resources-for-health-educators/eye-health-data-and-statistics/cataract-data-and-statistics/cataract-tables. Updated February 7, 2020. Accessed February 10, 2020.
5. Ostri C, Falck L, Boberg-Ans G, Kessel L. The need for toric intra-ocular lens implantation in public ophthalmology departments. Acta Ophthalmol. 2015;93(5):e396-e397.
6. Sundy M, McKnight D, Eck C, Rieger F 3rd. Visual acuity outcomes of toric lens implantation in patients undergoing cataract surgery at a residency training program. Mo Med. 2016;113(1):40-43.
7. Ferreira TB, Ribeiro P, Ribeiro FJ, O’Neill JG. Comparison of methodologies using estimated or measured values of total corneal astigmatism for toric intraocular lens power calculation. J Refract Surg. 2017;33(12):794-800.
8. Reitblat O, Levy A, Kleinmann G, Abulafia A, Assia EI. Effect of posterior corneal astigmatism on power calculation and alignment of toric intraocular lenses: comparison of methodologies. J Cataract Refract Surg. 2016;42(2):217-225.
9. 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(1):63-71.
10. Moreira HR, Hatch KM, Greenberg PB. Benchmarking outcomes in resident-performed cataract surgery with toric intraocular lenses [published correction appears in: Clin Experiment Ophthalmol. 2013;41(8):819]. Clin Exp Ophthalmol. 2013;41(6):624-626.
11. Retzlaff JA, Sanders DR, Kraff MC. Development of the SRK/T intraocular lens implant power calculation formula [published correction appears in: J Cataract Refract Surg. 1990;16(4):528]. J Cataract Refract Surg. 1990;16(3):333-340.
12. Roensch MA, Charton JW, Blomquist PH, Aggarwal NK, McCulley JP. Resident experience with toric and multifocal intraocular lenses in a public county hospital system. J Cataract Refract Surg. 2012;38(5):793-798.
13. Pouyeh B, Galor A, Junk AK, et al. Surgical and refractive outcomes of cataract surgery with toric intraocular lens implantation at a resident-teaching institution. J Cataract Refract Surg. 2011;37(9):1623-1628.
14. Ferreira TB, Ribeiro P, Ribeiro FJ, O’Neill JG. Comparison of astigmatic prediction errors associated with new calculation methods for toric intraocular lenses. J Cataract Refract Surg. 2017;43(3):340-347.
15. Abulafia A, Hill WE, Franchina M, Barrett GD. Comparison of methods to predict residual astigmatism after intraocular lens implantation. J Refract Surg. 2015;31(10):699-707.
16. Abulafia A, Barrett GD, Kleinmann G, et al. Prediction of refractive outcomes with toric intraocular lens implantation. J Cataract Refract Surg. 2015;41(5):936-944.
17. Wang Q, Jiang W, Lin T, Wu X, Lin H, Chen W. Meta-analysis of accuracy of intraocular lens power calculation formulas in short eyes. Clin Exp Ophthalmol. 2018;46(4):356-363.
18. Melles RB, Holladay JT, Chang WJ. Accuracy of intraocular lens calculation formulas. Ophthalmology. 2018;125(2):169-178.
19. Hoffer KJ. The Hoffer Q formula: a comparison of theoretic and regression formulas. J Cataract Refract Surg. 1993;19(6):700-712.
20. Cooke DL, Cooke TL. Comparison of 9 intraocular lens power calculation formulas. J Cataract Refract Surg. 2016;42(8):1157-1164.
21. American Society of Cataract and Refractive Surgery. Barrett toric calculator. www.ascrs.org/barrett-toric-calculator. Accessed February 5, 2020.
22. Holladay JT, Pettit G. Improving toric intraocular lens calculations using total surgically induced astigmatism for a 2.5 mm temporal incision. J Cataract Refract Surg. 2019;45(3):272-283.
23. Canovas C, Alarcon A, Rosén R, et al. New algorithm for toric intraocular lens power calculation considering the posterior corneal astigmatism. J Cataract Refract Surg. 2018;44(2):168-174.
24. Visser N, Berendschot TT, Bauer NJ, Nuijts RM. Vector analysis of corneal and refractive astigmatism changes following toric pseudophakic and toric phakic IOL implantation. Invest Ophthalmol Vis Sci. 2012;53(4):1865-1873.
25. Narváez 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(12):2050-2053.