Outcomes of One Dioptre Toric IOL Correction in Low Corneal Astigmatism: A Comparison with Non-toric IOLs Authors: Ben Connell, Jack Kane Published: 2026 ABSTRACT IntroductionToric intraocular lenses (IOLs) effectively correct astigmatism, but their benefit at low levels of corneal astigmatism remains uncertain. We compared visual and refractive outcomes between non-toric and toric IOLs in patients with low corneal astigmatism.MethodsThis was a real-world comparative cohort study. We analysed 804 eyes undergoing cataract surgery for which the Barrett Toric Calculator recommended an IOL with 1 dioptre (D) of toricity. Postoperative uncorrected distance visual acuity (UDVA), refractive astigmatism and corrected distance visual acuity (CDVA) were compared between non-toric and toric IOL groups, with pre-specified subgroup analyses by astigmatism magnitude and axis.ResultsBaseline corneal astigmatism was lower in the non-toric group than in the toric group (median 0.4 versus 0.7 D; P < 0.001), reflecting surgeon selection within the Barrett-recommended 1 D toric population. Postoperative refractive astigmatism was lower in the toric group (mean 0.31 versus 0.36 D; P = 0.022), a difference smaller than the 0.25 D increment in which subjective refraction was recorded, but UDVA was comparable overall (median 79 versus 78 letters; P = 0.122) as was CDVA (median 85 versus 85 letters; P = 0.235). In the with-the-rule (WTR) astigmatism subgroup, UDVA was better with non-toric IOLs (median 80 versus 77 letters; P = 0.002); against-the-rule and oblique subgroups showed no UDVA differences.ConclusionsIn eyes with low corneal astigmatism, 1 D toric IOLs reduced postoperative refractive astigmatism, particularly in eyes with against-the-rule (ATR) astigmatism and at higher levels of baseline corneal astigmatism, but did not improve uncorrected or corrected visual acuity overall. In eyes with WTR astigmatism, non-toric IOLs achieved better uncorrected vision. These findings suggest a statistically significant refractive benefit of low-power toric IOLs that is most evident in ATR eyes, while the functional visual benefit in this population remains uncertain and axis-dependent.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1007/s40123-026-01483-1. Key Summary Points Why carry out this study?Toric IOLs are well established for correcting high corneal astigmatism, but their value in low corneal astigmatism, the more common clinical scenario, has remained uncertain, particularly for the lowest commercially available toric power (1 D).This study asked whether the visual and refractive benefits of toric IOLs seen in higher-astigmatism eyes extend to eyes for which only a 1 D toric correction is recommended.What was learned from the study?In this real-world cohort of 804 eyes, 1 D toric IOLs reduced postoperative refractive astigmatism compared with non-toric IOLs, particularly in eyes with against-the-rule (ATR) astigmatism and at higher levels of baseline corneal astigmatism, but did not improve uncorrected or corrected visual acuity overall.The refractive benefit was axis-dependent: it was not seen in eyes with with-the-rule (WTR) astigmatism, where non-toric IOLs in fact achieved better uncorrected vision, suggesting that corneal astigmatism axis, not just magnitude, should inform the decision to implant a low-power toric IOL. Why carry out this study?Toric IOLs are well established for correcting high corneal astigmatism, but their value in low corneal astigmatism, the more common clinical scenario, has remained uncertain, particularly for the lowest commercially available toric power (1 D).This study asked whether the visual and refractive benefits of toric IOLs seen in higher-astigmatism eyes extend to eyes for which only a 1 D toric correction is recommended.What was learned from the study?In this real-world cohort of 804 eyes, 1 D toric IOLs reduced postoperative refractive astigmatism compared with non-toric IOLs, particularly in eyes with against-the-rule (ATR) astigmatism and at higher levels of baseline corneal astigmatism, but did not improve uncorrected or corrected visual acuity overall.The refractive benefit was axis-dependent: it was not seen in eyes with with-the-rule (WTR) astigmatism, where non-toric IOLs in fact achieved better uncorrected vision, suggesting that corneal astigmatism axis, not just magnitude, should inform the decision to implant a low-power toric IOL. Introduction Postoperative astigmatism is a key determinant of visual quality after cataract surgery, directly influencing uncorrected distance visual acuity (UDVA) and the spectacle independence patients can achieve [1]. In eyes with clinically significant corneal astigmatism, implantation of a non-toric intraocular lens (IOL) leaves substantial residual refractive astigmatism, reduces uncorrected visual acuity, and increases spectacle dependence; toric IOLs largely correct these deficits, although at the cost of additional preoperative planning, precise intraoperative alignment and the occasional need for secondary repositioning. The case for toric IOLs in high corneal astigmatism is well established [2]. Whether the same benefit extends to low corneal astigmatism, by far the more common scenario, affecting approximately 64% of cataract-age patients with ≤ 1 D on routine biometry [3], remains unresolved [4]. This question is clinically pressing in Australia and other settings where privately insured patients have ready access to premium IOLs, and where the most widely available toric option corrects approximately 1 D at the IOL plane (≈0.7 D at the corneal plane). The existing evidence base is limited. Prior studies of toric IOLs for low astigmatism have typically been small (often < 100 eyes) [5–9], frequently lacked non-toric comparison groups [5, 8], and were predominantly retrospective [5, 6, 8–10]. Heterogeneity in toric calculators, incision placement (superior [10], temporal [5, 7], or steep-axis [9]), and accounting for surgically induced astigmatism further confounds cross-study comparison. Despite these limitations, comparative analyses have generally favoured toric IOLs, with improvements in UDVA [6, 7, 9, 10] and reductions in postoperative refractive astigmatism [7, 9, 10]. Posterior corneal astigmatism and the astigmatic axis are now recognised as important determinants of total corneal astigmatism [11], and the UDVA gain with 1 D toric correction may be greatest in eyes with against-the-rule (ATR) astigmatism rather than with-the-rule (WTR) astigmatism [10]. We addressed these gaps in a real-world cohort of more than 800 eyes, comparing visual outcomes between non-toric and toric IOLs in low corneal astigmatism. By restricting analysis to eyes for which the Barrett toric IOL formula recommended a 1 D toric (the lowest astigmatic correction widely available), this is, to our knowledge, one of the largest studies to date on this question. Our objective was to determine whether the benefits of toric IOLs in clinical trials extend to routine clinical practice, and thereby to inform decision-making for the most common refractive scenario in cataract surgery. Methods Study Design and Setting This was a real-world comparative cohort study conducted at a high-volume, single surgeon private ophthalmology practice in Melbourne, Australia, between 2014 and March 2023. We included consecutive eyes for which the Barrett toric IOL calculator (https://www.apacrs.org/barr) recommended a toric intraocular lens (IOL) with 1 dioptre (D) of toricity and compared visual and refractive outcomes between eyes that received a non-toric IOL and those that received the recommended 1 D toric IOL. This study was conducted in accordance with the Declaration of Helsinki. Ethical approval was obtained from the Royal Victorian Eye and Ear Hospital Human Research Ethics Committee (approval number 23/1583HL). Given the retrospective nature of the study, the use of fully de-identified patient data, and that all cataract surgery and IOL selection was performed as standard of care independent of this research, the requirement for individual informed consent was waived by the ethics committee. No identifiable patient information is included in this publication. Inclusion and Exclusion Criteria Eligible patients were aged ≥ 18 years with preoperative biometry indicating a Barrett-recommended 1 D toric IOL. Eyes were excluded for incomplete or missing biometry, intraoperative complications (anterior or posterior capsule tear, vitreous loss, or zonular dehiscence), or ocular comorbidities that could affect refractive outcomes (previous laser vision correction, corneal ectasia including keratoconus, corneal scarring or irregular astigmatism, or pterygium). Eyes without a postoperative examination between 3 weeks and 6 months were also excluded. Where both eyes from a patient met the inclusion criteria, one eye was randomly selected. Baseline characteristics are summarised in Table 1.Table 1Baseline characteristics of eyes receiving non-toric and toric intraocular lenses included in the final analysisNon-toric (n = 365)Toric (n = 439)P valueMedian (IQR)Age at surgery (years)74.0 (68.0, 79.0)73.0 (67.0, 79.0)0.119aPreoperative CDVA (letters read)76.0 (72.0, 81.0)77.0 (71.0, 82.0)0.988aAxial length (mm)23.6 (23.0, 24.5)23.6 (23.0, 24.3)0.311aAnterior chamber depth (mm)3.1 (2.9, 3.4)3.1 (2.9, 3.4)0.574aLens thickness (mm)4.6 (4.3, 4.9)4.6 (4.3, 4.9)0.931aMean corneal power (dioptres)43.5 (42.6, 44.5)43.7 (42.6, 44.8)0.165aCentral corneal thickness (µm)548.0 (524.0, 572.0)549.0 (526.0, 570.0)0.917aHorizontal white-to-white (mm)12.0 (11.7, 12.3)12.0 (11.7, 12.4)0.177aPupil diameter (mm)3.7 (3.2, 4.4)3.7 (3.1, 4.4)0.935aIOL power (dioptres)22.0 (19.5, 23.0)21.5 (20.0, 23.0)0.450aCorneal astigmatism (dioptres)0.4 (0.2, 0.6)0.7 (0.4, 1.0)< 0.001aPredicted postoperative spherical equivalent (dioptres)−0.3 (−0.6, −0.1)−0.3 (−0.5, −0.1)0.412aSurgically induced astigmatism (dioptres)0.4 (0.2, 0.6)0.4 (0.2, 0.6)0.411aPostoperative spherical equivalent (dioptres)−0.5 (−1.0, −0.3)−0.5 (−1.0, −0.3)0.945aPostoperative IOL rotation (degrees)–4.0 (2.0, 7.0)Categorical, n (%)Male/female sex129/236 (35.3/64.7)179/260 (40.8/59.2)0.126bFuchs endothelial dystrophy13 (3.6)9 (2.1)0.201bPseudoexfoliation syndrome11 (3.0)3 (0.7)0.014cDiabetes28 (7.7)34 (7.7)1.000bLeft eye180 (49.3)224 (51.0)0.671bProcedure technique0.047bFLACS264 (72.3)288 (65.6)Phacoemulsification101 (27.7)151 (34.4)Anaesthetic0.932bTopical316 (86.6)382 (87.2)Peribulbar48 (13.2)54 (12.3)Sub-Tenon’s1 (0.3)1 (0.2)General0 (0.0)1 (0.2)CDVA Corrected distance visual acuity, IOL Intraocular lens, IQR Interquartile rangeaWilcoxon rank-sum testbPearson chi-square test; cFisher exact test. Bold P < 0.05 Table 1Baseline characteristics of eyes receiving non-toric and toric intraocular lenses included in the final analysisNon-toric (n = 365)Toric (n = 439)P valueMedian (IQR)Age at surgery (years)74.0 (68.0, 79.0)73.0 (67.0, 79.0)0.119aPreoperative CDVA (letters read)76.0 (72.0, 81.0)77.0 (71.0, 82.0)0.988aAxial length (mm)23.6 (23.0, 24.5)23.6 (23.0, 24.3)0.311aAnterior chamber depth (mm)3.1 (2.9, 3.4)3.1 (2.9, 3.4)0.574aLens thickness (mm)4.6 (4.3, 4.9)4.6 (4.3, 4.9)0.931aMean corneal power (dioptres)43.5 (42.6, 44.5)43.7 (42.6, 44.8)0.165aCentral corneal thickness (µm)548.0 (524.0, 572.0)549.0 (526.0, 570.0)0.917aHorizontal white-to-white (mm)12.0 (11.7, 12.3)12.0 (11.7, 12.4)0.177aPupil diameter (mm)3.7 (3.2, 4.4)3.7 (3.1, 4.4)0.935aIOL power (dioptres)22.0 (19.5, 23.0)21.5 (20.0, 23.0)0.450aCorneal astigmatism (dioptres)0.4 (0.2, 0.6)0.7 (0.4, 1.0)< 0.001aPredicted postoperative spherical equivalent (dioptres)−0.3 (−0.6, −0.1)−0.3 (−0.5, −0.1)0.412aSurgically induced astigmatism (dioptres)0.4 (0.2, 0.6)0.4 (0.2, 0.6)0.411aPostoperative spherical equivalent (dioptres)−0.5 (−1.0, −0.3)−0.5 (−1.0, −0.3)0.945aPostoperative IOL rotation (degrees)–4.0 (2.0, 7.0)Categorical, n (%)Male/female sex129/236 (35.3/64.7)179/260 (40.8/59.2)0.126bFuchs endothelial dystrophy13 (3.6)9 (2.1)0.201bPseudoexfoliation syndrome11 (3.0)3 (0.7)0.014cDiabetes28 (7.7)34 (7.7)1.000bLeft eye180 (49.3)224 (51.0)0.671bProcedure technique0.047bFLACS264 (72.3)288 (65.6)Phacoemulsification101 (27.7)151 (34.4)Anaesthetic0.932bTopical316 (86.6)382 (87.2)Peribulbar48 (13.2)54 (12.3)Sub-Tenon’s1 (0.3)1 (0.2)General0 (0.0)1 (0.2)CDVA Corrected distance visual acuity, IOL Intraocular lens, IQR Interquartile rangeaWilcoxon rank-sum testbPearson chi-square test; cFisher exact test. Bold P < 0.05 Baseline characteristics of eyes receiving non-toric and toric intraocular lenses included in the final analysis CDVA Corrected distance visual acuity, IOL Intraocular lens, IQR Interquartile range aWilcoxon rank-sum test bPearson chi-square test; cFisher exact test. Bold P < 0.05 Preoperative Assessment and Surgical Procedure Preoperative biometry was performed using the IOLMaster 700 (software version 1.88 or later; Carl Zeiss Meditec AG, Jena, Germany). Measurements were entered into the Barrett Toric Calculator (Barrett Universal II formula). All wounds were clear corneal temporal. A surgically induced astigmatism centroid of 0.15 D for horizontal corneal flattening, derived from a prior dataset of the same surgeon’s cases, was entered in the formula. Anterior corneal keratometry was used in preference to total corneal power, as the Barrett toric calculator already accounts for the expected contribution of posterior corneal astigmatism. IOL toricity was selected to minimise predicted postoperative refractive astigmatism, irrespective of any axis inversion. Cataract surgery was performed under standardised conditions using either conventional phacoemulsification or femtosecond laser-assisted cataract surgery (FLACS), with a 2.4 mm clear corneal incision (slightly enlarged when manufacturer-specified for higher-powered IOLs). The non-toric group received an AcrySof SN60WF or Clareon CNA0T0; the toric group received an AcrySof SN6AT2 or Clareon CNA0T2 (Alcon Laboratories Inc, Fort Worth, TX, USA). Toric alignment was achieved by one of three approaches: without preoperative marking; manually, by preoperative marking of the 0° and 180° meridians with a pendulum-weighted marker (patient seated), followed by intraoperative placement of radial limbal marks at the intended IOL axis using a Mendez degree gauge (Storz Ophthalmic Instruments); or digitally, using Callisto (Carl Zeiss Meditec AG) or Verion (Alcon Laboratories Inc) image-overlay systems. Postoperative Assessment Postoperative assessments were performed by orthoptists between 3 weeks and 6 months after surgery. Uncorrected distance visual acuity (UDVA) was measured first on the logMAR chart at 6 m, followed by manifest refraction to determine corrected distance visual acuity (CDVA). Repeat keratometry was obtained with the IOLMaster 700, and a dilated slit-lamp examination assessed rotational stability of the IOL by comparing IOL alignment marks against a graticule under a narrow slit beam. Statistical Analysis Analyses were performed using Stata (StataCorp, College Station, TX, USA). Visual acuity outcomes (UDVA and CDVA) were summarised as both means and medians, with the proportion of eyes achieving 70 letters (approximately Snellen 6/12) and 85 letters (approximately 6/6) also reported. To avoid bias from intentional myopic targets, UDVA analyses excluded eyes with a refractive target exceeding 1 D of myopia. Postoperative refractive astigmatism analyses were restricted to eyes achieving predetermined CDVA thresholds. Because visual acuity and most biometry variables were not normally distributed, medians were compared using the Wilcoxon rank-sum test and means using the t-test for unequal variances (Welch’s t-test); categorical variables were compared using Fisher’s exact test, or the Pearson chi-square test where expected cell counts were sufficiently large; P< 0.05 was considered statistically significant. Pre-specified subgroup analyses examined the influence of corneal astigmatism axis and magnitude on outcomes. By axis, eyes were classified as against-the-rule (ATR; steep corneal axis 0°–30° or 150°–180°), with-the-rule (WTR; 60°–120°), or oblique (31°–59° or 121°–149°). By magnitude, eyes were stratified by preoperative corneal astigmatism (≤ 0.5 D versus > 0.5 D). Surgically induced corneal astigmatism was calculated using Naeser’s polar value method [12]. Results Cohort and Baseline Characteristics The final analysis included 804 eyes, 365 in the non-toric and 439 in the toric IOL group (Table 1). Median age (74 versus 73 years; P = 0.119) and sex distribution (male 35.3% versus 40.8%, female 64.7% versus 59.2%; P = 0.126) were comparable between groups. As expected by design, median preoperative corneal astigmatism was higher in the toric group (0.7 D) than in the non-toric group (0.4 D; P < 0.001), reflecting the surgeon’s pattern of selecting toric IOLs preferentially in eyes with greater anterior corneal astigmatism within the Barrett-recommended 1 D toric population. Other biometric parameters were comparable between groups, with two exceptions: pseudoexfoliation syndrome (PXF) was more frequent in the non-toric group (3.0% versus 0.7%; P = 0.014), and femtosecond laser-assisted cataract surgery (FLACS) was used more frequently in the non-toric group than in the toric group (72.3% versus 65.6%; P = 0.047). Uncorrected Distance Visual Acuity Overall postoperative UDVA did not differ between groups (median 79 versus 78 letters; P = 0.122; Table 2, Fig. 1A). Stratification by astigmatic axis revealed a single subgroup difference: in eyes with with-the-rule (WTR) astigmatism, UDVA was higher in the non-toric group (median 80 letters) than in the toric group (median 77 letters; P = 0.002). No corresponding difference was observed in eyes with against-the-rule (ATR) or oblique astigmatism, nor when stratifying by corneal astigmatism magnitude (≤ 0.5 D versus > 0.5 D). The WTR finding was not attributable to differences in spherical refractive outcomes, which were identical between groups (median spherical equivalent −0.5 D in each; P = 0.945).Table 2Comparison of final uncorrected distance visual acuity (UDVA) between groupsNon-toric (n = 365)Toric (n = 439)P valuePostoperative UDVA (letters read)Mean (95% CI)77.7 (76.7, 78.8)76.9 (76.0, 77.8)0.238aMedian (IQR)79.0 (74.0, 84.0)78.0 (73.0, 83.0)0.122bBy corneal astigmatism axisATR77.0 (73.0, 83.0)77.0 (72.0, 83.0)0.893bOblique78.5 (74.0, 84.0)80.0 (75.0, 85.0)0.287bWTR80.0 (75.0, 84.0)77.0 (70.0, 83.0)0.002bReading ≥ 70 letters, n (%)Total267 (85.9)312 (83.2)0.398cBy corneal astigmatism axisATR40 (81.6)55 (79.7)1.000cOblique90 (84.9)114 (89.8)0.320cWTR132 (89.2)143 (79.9)0.023cReading ≥ 85 letters, n (%)Total72 (23.2)73 (19.5)0.260cBy corneal astigmatism axisATR7 (14.3)14 (20.3)0.470cOblique26 (24.5)33 (26.0)0.880cWTR36 (24.3)26 (14.5)0.033cUDVA analyses were restricted to 311 non-toric and 375 toric eyes after excluding eyes with a refractive target exceeding 1 D of myopia; percentages for UDVA rows are calculated on these denominatorsat-test for unequal variancesbWilcoxon rank-sum testcFisher exact testBold P < 0.05. UDVA summarised as mean, median and percentage of eyes reading ≥ 70 and ≥ 85 letters, overall and stratified by corneal astigmatism axis. Outcomes stratified by corneal astigmatism magnitude are provided in Supplementary Table S1. Eight non-toric eyes had no assignable corneal astigmatism axis and are included in the total rows only. ATR, against-the-rule; IQR, interquartile range; WTR, with-the-ruleFig. 1Postoperative outcomes in the non-toric (light grey bars) and toric (dark grey bars) IOL groups. A Cumulative percentage of eyes achieving each level of uncorrected distance visual acuity (UDVA) or better, in letters read, in the 311 non-toric and 375 toric eyes retained in the uncorrected acuity analysis after exclusion of eyes with a refractive target exceeding 1 D of myopia. B Distribution of postoperative refractive astigmatism, shown as the percentage of eyes in each dioptre range, in the 359 non-toric and 425 toric eyes achieving a corrected distance visual acuity of 70 letters or better. UDVA, uncorrected distance visual acuity; IOL, intraocular lens Table 2Comparison of final uncorrected distance visual acuity (UDVA) between groupsNon-toric (n = 365)Toric (n = 439)P valuePostoperative UDVA (letters read)Mean (95% CI)77.7 (76.7, 78.8)76.9 (76.0, 77.8)0.238aMedian (IQR)79.0 (74.0, 84.0)78.0 (73.0, 83.0)0.122bBy corneal astigmatism axisATR77.0 (73.0, 83.0)77.0 (72.0, 83.0)0.893bOblique78.5 (74.0, 84.0)80.0 (75.0, 85.0)0.287bWTR80.0 (75.0, 84.0)77.0 (70.0, 83.0)0.002bReading ≥ 70 letters, n (%)Total267 (85.9)312 (83.2)0.398cBy corneal astigmatism axisATR40 (81.6)55 (79.7)1.000cOblique90 (84.9)114 (89.8)0.320cWTR132 (89.2)143 (79.9)0.023cReading ≥ 85 letters, n (%)Total72 (23.2)73 (19.5)0.260cBy corneal astigmatism axisATR7 (14.3)14 (20.3)0.470cOblique26 (24.5)33 (26.0)0.880cWTR36 (24.3)26 (14.5)0.033cUDVA analyses were restricted to 311 non-toric and 375 toric eyes after excluding eyes with a refractive target exceeding 1 D of myopia; percentages for UDVA rows are calculated on these denominatorsat-test for unequal variancesbWilcoxon rank-sum testcFisher exact testBold P < 0.05. UDVA summarised as mean, median and percentage of eyes reading ≥ 70 and ≥ 85 letters, overall and stratified by corneal astigmatism axis. Outcomes stratified by corneal astigmatism magnitude are provided in Supplementary Table S1. Eight non-toric eyes had no assignable corneal astigmatism axis and are included in the total rows only. ATR, against-the-rule; IQR, interquartile range; WTR, with-the-rule Comparison of final uncorrected distance visual acuity (UDVA) between groups UDVA analyses were restricted to 311 non-toric and 375 toric eyes after excluding eyes with a refractive target exceeding 1 D of myopia; percentages for UDVA rows are calculated on these denominators at-test for unequal variances bWilcoxon rank-sum test cFisher exact test Bold P < 0.05. UDVA summarised as mean, median and percentage of eyes reading ≥ 70 and ≥ 85 letters, overall and stratified by corneal astigmatism axis. Outcomes stratified by corneal astigmatism magnitude are provided in Supplementary Table S1. Eight non-toric eyes had no assignable corneal astigmatism axis and are included in the total rows only. ATR, against-the-rule; IQR, interquartile range; WTR, with-the-rule Postoperative outcomes in the non-toric (light grey bars) and toric (dark grey bars) IOL groups. A Cumulative percentage of eyes achieving each level of uncorrected distance visual acuity (UDVA) or better, in letters read, in the 311 non-toric and 375 toric eyes retained in the uncorrected acuity analysis after exclusion of eyes with a refractive target exceeding 1 D of myopia. B Distribution of postoperative refractive astigmatism, shown as the percentage of eyes in each dioptre range, in the 359 non-toric and 425 toric eyes achieving a corrected distance visual acuity of 70 letters or better. UDVA, uncorrected distance visual acuity; IOL, intraocular lens Because toric and non-toric eyes differed substantially in baseline corneal astigmatism, a propensity-score-matched analysis was performed, matching eyes on baseline corneal astigmatism. After matching, there was no significant difference between lens types in uncorrected visual acuity (adjusted difference 0.02 letters, 95% CI −2.99 to 3.02, P = 0.99) or in the proportion achieving 85 letters or better (adjusted difference 2.1%, 95% CI −5.9% to 10.1%, P = 0.61). Postoperative Refractive Astigmatism Refractive astigmatism was lower in the toric group overall (Table 3, Fig. 1B). Among eyes achieving the 70-letter threshold, mean postoperative refractive astigmatism was lower in the toric group (0.31 versus 0.36 D; P = 0.022), with the Wilcoxon rank-sum test also detecting a statistically significant distributional difference between groups despite equal medians (0.25 D [IQR 0.00–0.50] in both groups; P = 0.037). No axis subgroup reached significance at this threshold, although among eyes with corneal astigmatism > 0.5 D a higher proportion of toric eyes achieved residual astigmatism ≤ 0.5 D (79.6% versus 89.6%; P = 0.012; Supplementary Table S1). Among eyes achieving the 85-letter CDVA threshold, however, the toric benefit was significant in eyes with corneal astigmatism > 0.5 D (73.9% versus 91.8% achieving residual astigmatism ≤ 0.5 D; P < 0.001) and in eyes with ATR astigmatism (75.9% versus 92.3%, P = 0.049; median 0.50 versus 0.25 D, P = 0.032), but not in eyes with corneal astigmatism ≤ 0.5 D, oblique astigmatism, or WTR astigmatism (Table 3, Supplementary Table S1).Table 3Comparison of refractive astigmatism at the final visit between groupsNon-toric (n = 365)Toric (n = 439)P valueCDVA ≥ 70 letters readRefractive astigmatism (dioptres)Mean (95% CI)0.36 (0.32, 0.39)0.31 (0.28, 0.33)0.022aMedian (IQR)0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.037bBy corneal astigmatism axisATR0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.470bOblique0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.125bWTR0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.140bRefractive astigmatism ≤ 0.5 D, n (%)Total308 (85.8)384 (90.4)0.058cBy corneal astigmatism axisATR46 (80.7)68 (88.3)0.232cOblique102 (86.4)135 (92.5)0.152cWTR153 (86.9)181 (89.6)0.427cRefractive astigmatism ≤ 1.0 D, n (%)Total350 (97.5)422 (99.3)0.075cBy corneal astigmatism axisATR54 (94.7)77 (100.0)0.075cOblique117 (99.2)146 (100.0)0.447cWTR171 (97.2)199 (98.5)0.481cCDVA ≥ 85 letters readRefractive astigmatism (dioptres)Mean (95% CI)0.34 (0.30, 0.38)0.27 (0.24, 0.31)0.009aMedian (IQR)0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.015bBy corneal astigmatism axisATR0.50 (0.00, 0.50)0.25 (0.00, 0.50)0.032bOblique0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.178bWTR0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.192bRefractive astigmatism ≤ 0.5 D, n (%)Total203 (86.4)246 (92.8)0.018cBy corneal astigmatism axisATR22 (75.9)48 (92.3)0.049cOblique68 (87.2)88 (94.6)0.107cWTR108 (88.5)110 (91.7)0.520cRefractive astigmatism ≤ 1.0 D, n (%)Total229 (97.4)263 (99.2)0.156cBy corneal astigmatism axisATR26 (89.7)52 (100.0)0.043cOblique78 (100.0)93 (100.0)1.000cWTR119 (97.5)118 (98.3)1.000cRefractive astigmatism summarised as mean, median and percentage of eyes achieving ≤ 0.5 D and ≤ 1.0 D, at two corrected distance visual acuity (CDVA) thresholds (≥ 70 and ≥ 85 letters), overall and stratified by corneal astigmatism axis. Analyses at the ≥ 70-letter threshold include 359 non-toric and 425 toric eyes, and at the ≥ 85-letter threshold 235 non-toric and 265 toric eyes; percentages are calculated on these denominators. Non-toric eyes with no assignable corneal astigmatism axis (eight at the ≥ 70-letter threshold and six at the ≥ 85-letter threshold) are included in the total rows only. Outcomes stratified by corneal astigmatism magnitude are provided in Supplementary Table S1. ATR, against-the-rule; IQR, interquartile range; WTR, with-the-rule.at-test for unequal variancesbWilcoxon rank-sum test; c Fisher exact testBold P < 0.05 Table 3Comparison of refractive astigmatism at the final visit between groupsNon-toric (n = 365)Toric (n = 439)P valueCDVA ≥ 70 letters readRefractive astigmatism (dioptres)Mean (95% CI)0.36 (0.32, 0.39)0.31 (0.28, 0.33)0.022aMedian (IQR)0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.037bBy corneal astigmatism axisATR0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.470bOblique0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.125bWTR0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.140bRefractive astigmatism ≤ 0.5 D, n (%)Total308 (85.8)384 (90.4)0.058cBy corneal astigmatism axisATR46 (80.7)68 (88.3)0.232cOblique102 (86.4)135 (92.5)0.152cWTR153 (86.9)181 (89.6)0.427cRefractive astigmatism ≤ 1.0 D, n (%)Total350 (97.5)422 (99.3)0.075cBy corneal astigmatism axisATR54 (94.7)77 (100.0)0.075cOblique117 (99.2)146 (100.0)0.447cWTR171 (97.2)199 (98.5)0.481cCDVA ≥ 85 letters readRefractive astigmatism (dioptres)Mean (95% CI)0.34 (0.30, 0.38)0.27 (0.24, 0.31)0.009aMedian (IQR)0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.015bBy corneal astigmatism axisATR0.50 (0.00, 0.50)0.25 (0.00, 0.50)0.032bOblique0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.178bWTR0.25 (0.00, 0.50)0.25 (0.00, 0.50)0.192bRefractive astigmatism ≤ 0.5 D, n (%)Total203 (86.4)246 (92.8)0.018cBy corneal astigmatism axisATR22 (75.9)48 (92.3)0.049cOblique68 (87.2)88 (94.6)0.107cWTR108 (88.5)110 (91.7)0.520cRefractive astigmatism ≤ 1.0 D, n (%)Total229 (97.4)263 (99.2)0.156cBy corneal astigmatism axisATR26 (89.7)52 (100.0)0.043cOblique78 (100.0)93 (100.0)1.000cWTR119 (97.5)118 (98.3)1.000cRefractive astigmatism summarised as mean, median and percentage of eyes achieving ≤ 0.5 D and ≤ 1.0 D, at two corrected distance visual acuity (CDVA) thresholds (≥ 70 and ≥ 85 letters), overall and stratified by corneal astigmatism axis. Analyses at the ≥ 70-letter threshold include 359 non-toric and 425 toric eyes, and at the ≥ 85-letter threshold 235 non-toric and 265 toric eyes; percentages are calculated on these denominators. Non-toric eyes with no assignable corneal astigmatism axis (eight at the ≥ 70-letter threshold and six at the ≥ 85-letter threshold) are included in the total rows only. Outcomes stratified by corneal astigmatism magnitude are provided in Supplementary Table S1. ATR, against-the-rule; IQR, interquartile range; WTR, with-the-rule.at-test for unequal variancesbWilcoxon rank-sum test; c Fisher exact testBold P < 0.05 Comparison of refractive astigmatism at the final visit between groups Refractive astigmatism summarised as mean, median and percentage of eyes achieving ≤ 0.5 D and ≤ 1.0 D, at two corrected distance visual acuity (CDVA) thresholds (≥ 70 and ≥ 85 letters), overall and stratified by corneal astigmatism axis. Analyses at the ≥ 70-letter threshold include 359 non-toric and 425 toric eyes, and at the ≥ 85-letter threshold 235 non-toric and 265 toric eyes; percentages are calculated on these denominators. Non-toric eyes with no assignable corneal astigmatism axis (eight at the ≥ 70-letter threshold and six at the ≥ 85-letter threshold) are included in the total rows only. Outcomes stratified by corneal astigmatism magnitude are provided in Supplementary Table S1. ATR, against-the-rule; IQR, interquartile range; WTR, with-the-rule. at-test for unequal variances bWilcoxon rank-sum test; c Fisher exact test Bold P < 0.05 In the same propensity-score-matched analysis, there was no significant difference between lens types in residual refractive astigmatism (adjusted difference −0.05 D, 95% CI −0.12 to 0.01, P = 0.11), nor in the proportion with refractive astigmatism of 0.50 D or less (adjusted difference 6.3%, 95% CI −0.7% to 13.3%, P = 0.08). In a logistic regression model incorporating a lens-type by baseline corneal astigmatism interaction, the toric lens was associated with a progressively higher probability of achieving residual refractive astigmatism of 0.50 D or less as baseline corneal astigmatism increased. Although the interaction term did not itself reach significance (χ2(1) = 2.70, P = 0.10), the adjusted between-lens difference favoured the toric lens significantly at higher baseline astigmatism: the absolute increase in the probability of residual astigmatism ≤ 0.50 D was 8.8% at 1.00 D, reaching significance at baseline corneal astigmatism of 0.75 D (+8.8%, P = 0.01), 1.00 D (+13.5%, P = 0.02) and 1.25 D (+18.8%, P = 0.03). Corrected Distance Visual Acuity CDVA outcomes were comparable between groups across most measures (Table 4). There was no difference in CDVA when subdivided by corneal astigmatism into ≤ 0.5 D or > 0.5 D groups (Supplementary Table S1).Table 4Comparison of final corrected distance visual acuity (CDVA) between groupsNon-toric (n = 365)Toric (n = 439)P valuePostoperative CDVA (letters read)Mean (95% CI)84.7 (84.0, 85.4)84.3 (83.7, 85.0)0.420aMedian (IQR)85.0 (83.0, 89.0)85.0 (83.0, 89.0)0.235bReading ≥ 70 letters, n (%)Total359 (98.4)425 (96.8)0.180cReading ≥ 85 letters, n (%)Total235 (64.4)265 (60.4)0.244cCDVA summarised as mean, median and percentage of eyes reading ≥ 70 and ≥ 85 letters. Axis subgroup data were not available for CDVA; outcomes stratified by corneal astigmatism magnitude are provided in Supplementary Table S1. IQR, interquartile rangeat-test for unequal variancesbWilcoxon rank-sum testcFisher exact testBold P < 0.05 Table 4Comparison of final corrected distance visual acuity (CDVA) between groupsNon-toric (n = 365)Toric (n = 439)P valuePostoperative CDVA (letters read)Mean (95% CI)84.7 (84.0, 85.4)84.3 (83.7, 85.0)0.420aMedian (IQR)85.0 (83.0, 89.0)85.0 (83.0, 89.0)0.235bReading ≥ 70 letters, n (%)Total359 (98.4)425 (96.8)0.180cReading ≥ 85 letters, n (%)Total235 (64.4)265 (60.4)0.244cCDVA summarised as mean, median and percentage of eyes reading ≥ 70 and ≥ 85 letters. Axis subgroup data were not available for CDVA; outcomes stratified by corneal astigmatism magnitude are provided in Supplementary Table S1. IQR, interquartile rangeat-test for unequal variancesbWilcoxon rank-sum testcFisher exact testBold P < 0.05 Comparison of final corrected distance visual acuity (CDVA) between groups CDVA summarised as mean, median and percentage of eyes reading ≥ 70 and ≥ 85 letters. Axis subgroup data were not available for CDVA; outcomes stratified by corneal astigmatism magnitude are provided in Supplementary Table S1. IQR, interquartile range at-test for unequal variances bWilcoxon rank-sum test cFisher exact test Bold P < 0.05 Toric IOL Alignment and Stability Surgically induced corneal astigmatism did not differ between groups. In the toric IOL group, the median IOL deviation from the intended axis at the postoperative assessment was 4° (interquartile range 2–7°). The IOL was rotated 10° or more from the intended position in 14% of eyes. There was no correlation between absolute rotation and duration since surgery (Spearman correlation coefficient: −0.023, P = 0.641), indicating that the timing of the postoperative assessment, which ranged from 3 weeks to 6 months, did not materially influence the measured rotation. Postoperative IOL misalignment from the intended axis did not differ significantly across the three alignment techniques (none, image-guided, weighted marker; Kendall’s τb = 0.05, P = 0.25). No eyes required secondary repositioning surgery for misalignment. Discussion In this real-world cohort of more than 800 eyes for which the Barrett toric formula recommended an IOL with 1 dioptre of astigmatic correction (the lowest commonly available toric power), eyes receiving the recommended toric IOL achieved lower postoperative refractive astigmatism than eyes receiving a non-toric IOL, despite starting with significantly higher preoperative corneal astigmatism. This refractive benefit was most pronounced in eyes with against-the-rule (ATR) astigmatism and absent in eyes with with-the-rule (WTR) astigmatism. The size of this difference should not be overstated. The between-group difference in mean residual cylinder was 0.05 D, one-fifth of the 0.25 D increment in which subjective refraction was recorded, and a difference of this magnitude is not clinically detectable in an individual eye. The finding is therefore best understood as a statistically significant group-level difference in refractive outcome rather than a clinically detectable advantage for any individual patient. Corneal astigmatism is itself a strong predictor of postoperative astigmatism, so the toric advantage held against an unfavourable baseline. UDVA was superior in the non-toric WTR subgroup. CDVA, in contrast, was comparable between groups. Our results align with the prospective randomised trial by Hienert and colleagues, who demonstrated lower postoperative refractive astigmatism with toric IOLs in low-astigmatism eyes [7]. Their cohort had higher median corneal astigmatism (1.15 D versus 0.4 D in the non-toric and 0.7 D in the toric group in our study) and an order-of-magnitude smaller sample (58 versus 804 eyes), but stronger internal validity from the randomised, double-blind design. We extend their conclusion by showing that the refractive benefit persists in routine clinical practice, in a population at the lower end of the astigmatism spectrum where the absolute gain from toric correction is smallest. The absence of a corresponding UDVA improvement in our cohort is consistent with the findings of Gundersen and colleagues [10], who similarly observed that lower refractive astigmatism with toric IOLs did not translate to better uncorrected vision. Whether the refractive advantage justifies the additional cost and workflow demands of toric implantation in this low-astigmatism population, particularly when functional vision is unchanged, is a question that current evidence cannot fully resolve. The axis-specific effect deserves emphasis. The toric benefit was greatest in eyes with ATR astigmatism, where it reached significance among eyes achieving 85 letters or better of corrected acuity. This is consistent with the established contribution of the posterior cornea to total corneal astigmatism [11, 13–15] and with the recognised limitation of anterior keratometry, which can overestimate WTR and underestimate ATR astigmatism by failing to capture posterior corneal contributions. Because that uncaptured posterior contribution is itself against-the-rule in most eyes [16–18], the cataract-age population most likely to benefit from low-power toric correction is precisely the group in whom anterior keratometry alone may have historically led to undertreatment. The Barrett formula’s incorporation of expected posterior corneal astigmatism appears to translate into a statistically detectable reduction in residual cylinder in these eyes, although, as noted above, the absolute size of that reduction remains small. In WTR eyes, the effect ran the other way, and its direction rather than its magnitude is the likely explanation. Residual astigmatism was of equal magnitude in both groups (median 0.25 D; P = 0.140), yet uncorrected vision was three letters worse with the toric lens, a pattern pointing to a difference in axis rather than amount. A toric selected against an overestimated WTR target, in discrete power steps and without regard to the meridian of the predicted residual, may carry a low-magnitude WTR eye across the neutral point into a small ATR residual, whereas a non-toric eye undergoes no such toric-induced shift. As low ATR astigmatism has been reported to degrade uncorrected distance vision more than WTR astigmatism of equal magnitude [19], the toric eyes may read marginally worse despite equivalent residual cylinder. This reinforces that the axis of corneal astigmatism, not magnitude alone, should inform low-power toric IOL selection. Two methodological considerations bear on the interpretation of our findings. First, IOL selection in this real-world cohort was non-randomised, which introduces selection bias: when the Barrett calculator recommends a 1 D toric for an eye with relatively low corneal astigmatism, the predicted refractive gain is small and surgeons may forgo the toric; whereas the same recommendation in an eye with higher corneal astigmatism yields a larger predicted gain, increasing the likelihood of toric implantation. This pattern explains the higher baseline corneal astigmatism in our toric group, a phenomenon also reported by Oshika and colleagues [20]. To mitigate this bias, we performed a subgroup analysis restricted to eyes with ≥ 0.5 D of corneal astigmatism; the absence of UDVA benefit persisted, suggesting the finding is not an artefact of selection. Second, the prior literature is heterogeneous: many studies have used statistical methods that assume a normal distribution of visual acuity data [5], employed older toric calculators that did not account for posterior corneal contributions or relied on small samples. We used non-parametric methods appropriate to the non-normal distribution of acuity data and the Barrett toric calculator, currently the most widely adopted tool in our setting. Baseline differences between groups were modest. Patients receiving non-toric IOLs more frequently had pseudoexfoliation syndrome (associated with increased risk of zonular weakness), and FLACS was used more often in this group. None of these differences would be expected to favour the toric group’s refractive outcomes; if anything, they should bias against the observed result. The central remaining question is the corneal astigmatism threshold at which toric IOLs deliver a meaningful functional benefit. The Hienert trial [7], at a median 1.15 D, found a UDVA advantage; our cohort, at medians of 0.4 D and 0.7 D in the non-toric and toric groups respectively, did not. A prospective trial with matched preoperative astigmatism levels would resolve where this threshold lies. Future work could establish axis inversion directly, through vector analysis of the postoperative refractive axis in a dataset collected for that purpose, which is beyond what a retrospective cohort of this design can support. In the interim, our data show a statistically significant refractive advantage for 1 D toric IOLs in low corneal astigmatism, particularly in ATR eyes, while acknowledging that this advantage is small in absolute terms and is not reflected in uncorrected visual acuity. Limitations. This study has several limitations. IOL selection was non-randomised, introducing selection bias that we sought to mitigate through magnitude-based subgroup analysis and propensity-score matching, though residual confounding cannot be excluded. All surgery was performed by a single surgeon at a single high-volume practice, which improves the internal consistency of surgical technique and of the surgically induced astigmatism centroid used in the toric calculation, but limits the extent to which these findings can be generalised to other surgeons and settings. Three alignment techniques (no marking, image-guided overlay, and weighted pendulum marker) were pooled in the toric group; although postoperative misalignment did not differ significantly between techniques, the study was not powered to exclude smaller differences. Femtosecond laser-assisted cataract surgery was used more frequently in the non-toric group (72.3% versus 65.6%; P = 0.047); because this technique was more common in the comparison group than in the toric group, any effect of FLACS on outcomes would be expected to bias against, rather than account for, the observed toric advantage. We did not perform a formal covariate-adjusted sensitivity analysis for this imbalance, which we acknowledge as a limitation of the present analysis. Restricting the analysis to one randomly selected eye per patient, while necessary to ensure statistical independence between observations, reduced the available sample size from 1026 to 804 eyes. Race and ethnicity data were not collected as part of this retrospective clinical dataset, limiting our ability to assess the generalisability of these findings across different populations. Although the proportion of male and female patients was balanced between groups (Table 1), we did not perform a sex-stratified analysis of baseline characteristics or outcomes; because the dataset was extracted and de-identified specifically for the pre-specified comparison reported here, a new post-hoc breakdown by sex was not undertaken. Finally, both authors have financial relationships with Alcon Laboratories, the manufacturer of the IOLs evaluated in this study. Ben Connell has received speaker fees from Alcon and Johnson & Johnson, and Jack Kane is the inventor of the Kane formula and has received licensing fees from Alcon and other biometer and formula licensees. Given that all study lenses were manufactured by Alcon, these relationships are not expected to have influenced the comparison between study groups; they are disclosed for transparency. Conclusions In this real-world cohort of eyes with low corneal astigmatism, 1 D toric IOLs reduced postoperative refractive astigmatism compared with non-toric IOLs, particularly in eyes with against-the-rule (ATR) astigmatism and at higher levels of baseline corneal astigmatism, but did not improve uncorrected or corrected visual acuity overall. In eyes with with-the-rule (WTR) astigmatism, non-toric IOLs achieved better uncorrected vision, a difference that appears to reflect the axis rather than the magnitude of the residual astigmatism. Overall, these findings demonstrate a statistically significant refractive benefit of low-power toric IOLs that is most evident in ATR eyes and at higher corneal astigmatism, while the functional visual benefit in this population remains uncertain and axis-dependent, and warrants further prospective study. Supplementary Information Below is the link to the electronic supplementary material.Supplementary file1 (PDF 162 kb) Supplementary file1 (PDF 162 kb)