Preoperative Higher-Order Aberrations as Predictors of Visual Acuity Changes Following Presbyopia-Correcting (PCIOL) and Aspheric Monofocal Intraocular Lens Implantation

Abstract

Purpose

To determine if preoperative ocular higher order aberrations (HOAs) were associated with changes in uncorrected distance (UCDVA), intermediate (UCIVA), near (UCNVA), corrected distance at reduced contrast (CDVA 10% and 25%) logMAR acuities after PCIOL or aspheric IOL implantation.

Patients and Methods

Records of patients implanted with AcrySof IQ PanOptix (group I, n = 30), TECNIS Symfony (group II, n = 35), TECNIS bifocal (group III, n = 34), or TECNIS monofocal (group IV, n = 32) were reviewed where UCDVA, reduced contrast CDVA (logMAR chart at 6 m), UCIVA and UCNVA (at 80 cm and 40 cm) assessed under photopic conditions (∼85 cd/m²), ocular aberrations (under mesopic conditions, pupil sizes 3mm and 5mm), and refraction had been monitored.

Results

Data from one eye/case were included for analysis. The chief findings were: (i) Median UCDVA, UCIVA and UCNVA values improved but CDVA (10% and 25%) remained unchanged in groups I–III. Median UCDVA, CDVA (10% and 25%) improved but UCIVA and UCNVA values remained unchanged in group IV. (ii) Mean spherical aberration (5mm pupil) and total root mean square HOA values (for 3 and 5mm pupil sizes) reduced in all 4 groups (ANOVA, p < 0.001). Mean trefoil and coma values did not change significantly. (iii) There were significant correlations between changes (Δ) in uncorrected acuities, CDVA (10% and 25%) and specific preop HOA values. The limits of agreement in the various models describing these associations were ≤0.2 units of logMAR acuity in groups II for ΔUCIVA and ΔCDVA (10% and 25%), III for ΔUCNVA and ΔCDVA (10% and 25%), IV for ΔUCIVA and ΔCDVA (10% and 25%).

Conclusion

Preoperative HOA measurements showed statistically significant though modest associations with changes in uncorrected visual acuities. In some instances, the estimated change improved by including information regarding preoperative refractive error. Validation in larger cohorts is required before clinical implementation.

Introduction

The original designs of intraocular lenses (IOLs) featured spherical optical surfaces, and this led to high levels of optical aberrations. However, the higher-order aberrations (HOAs) of the pseudophakic eye can be modified by aspherizing the IOL optical surfaces, resulting in an overall improvement in the quality of postoperative distance visual acuity.¹ Such IOLs have the potential to improve the quality of vision at reduced contrast, but only for viewing objects within a limited range from the eye. In short, the benefit gained when viewing distant objects is, to some extent, counter-balanced by poor near acuity. However, the design of the IOL can be further manipulated to mitigate this circumstance, potentially achieving good visual acuity at various distances. Today, several designs aimed at improving distance and near vision have been approved for implantation, and these can be labelled as presbyopia-correcting intraocular lenses (PCIOLs). The ideal PCIOL should deliver superior visual quality irrespective of the distance between the object of regard and the patient. This perfect goal has still not been achieved. Nevertheless, the patient can experience a compromised, though acceptable, level of vision over a wide range. The key factor motivating a patient to request PCIOLs is the desire to be spectacle-free, and patient satisfaction rates are highly dependent on the postoperative visual acuity achieved with PCIOLs.² Various formulas are available to calculate the PCIOL that should eliminate dependence on spectacles after implantation; however, none of them guarantee that the postoperative uncorrected distance visual acuity (UCDVA) will be 20/20 or better. Thus, it would be useful to predict, with reasonable certainty, the postoperative UCDVA a patient is likely to achieve following complication-free PCIOL implantation. The factors reported to correlate with postoperative visual outcomes after PCIOL implantation include axial length and pupil size,^3,4 age and preoperative visual acuity,⁵ age and PCIOL tilt,⁶ angles alpha⁷ and kappa,⁸ and PCIOL design.⁹ On the other hand, patients who do not wish to compromise their postoperative UCDVA may request aspheric IOLs. The preoperative values of axial length, pupil size, angles alpha and kappa can be considered as objective factors associated with postoperative UCDVA, but could other objective optical factors, such as the HOAs evaluated at the preoperative stage, be associated with the changes in uncorrected visual acuity? Despite extensive research, no validated predictive models incorporate automated preoperative measurements of HOAs for the various types of IOLs.

There is no consensus on the preoperative factors universally accepted as useful for estimating uncorrected distance (UCDVA), intermediate (UCIVA), and near (UCNVA) acuities, as well as corrected distance visual acuities (CDVA) at reduced contrast after PCIOL implantation.

PCIOLs are available in a wide range of optical designs, yet the desired endpoints are, in essence, the same. When there are no significant differences between groups before and after implantation across different PCIOL designs, the data from the groups can be regarded as drawn from a common population. If preoperative values of UCDVA, UCIVA, UCNVA, and low-contrast CDVA show no significant differences, and the lack of significance remains consistent after implanting different PCIOLs, then the data can be pooled as emanating from a common source. This may yield preoperative factors useful for estimating the likely change in UCDVA, UCIVA, UCNVA, and low-contrast CDVA, not just for one PCIOL but across a range of PCIOLs.

This study aimed to determine whether changes in UCDVA, UCIVA, UCNVA, and CDVA at low contrast (10% and 25%) following routine uncomplicated phacoemulsification and implantation with PCIOLs were associated with preoperative HOA values. Should significant associations be uncovered then the analysis would be extended to determine whether models could be developed for estimating the change in uncorrected visual acuity, and/or low-contrast CDVA after PCIOL implantation, based on objective HOAs and subjective refraction assessment before surgery.

Materials and Methods

Study Design

This was a retrospective post hoc analysis of data obtained from a completed, and published,¹⁰ partially masked, interventional trial that was undertaken at the University Eye Hospital “Svjetlost” in Zagreb. The subjects had satisfied the criteria for surgical implantation and monitoring following surgery. The trial had been approved by the Institutional Review Board of the University Eye Hospital “Svjetlost” (Zagreb, Croatia), conducted in accordance with the tenets of the Declaration of Helsinki [ClinicalTrials.gov identifier: NCT07008768]. When the trial was approved, the Ethics Committee at the University Eye Hospital “Svjetlost” in Zagreb, Croatia was not required by law to provide an ethical approval number. All subjects had previously signed consent forms after being fully informed of the trial.

Inclusion Criteria

The preoperative refractive errors of subjects ranged from +0.50 D to +5.00 D, with refractive astigmatism ≤+1.00 D. Patients suitable for binocular implantation, aged ≥45 years, requesting to become spectacle-free for distance, and ideally at 80cm and 40cm, had been enrolled. The limit of CDVA was 0.7 logMAR for those with signs of cataract and 0.05 logMAR in cases of refractive lens exchange. All cases attended follow-up appointments at 1, 6 and 12 months after surgery.

Exclusion Criteria

Patients with a history of eye injuries, ocular surgery, glaucoma, endocrine disorders, immunological or rheumatological systemic disease, corneal abnormalities, or amblyopia. Patients that failed to attend follow-up sessions at 1, 6 and 12 months after surgery.

Description of Intraocular Lenses (IOLs) and Subjects

Group I, AcrySof IQ PanOptix (TFNT00) a hydrophobic aspheric diffractive trifocal intraocular lens with a non-apodised diffractive zone to provide clear vision from distance to near with ENLIGHTEN^TM technology. This combines a trifocal (near focus at 40 cm) and a quadrifocal (intermediate focus at 60 cm) design.

Group II, TECNIS Symfony (ZXR00), is a hydrophobic, aspheric, diffractive intraocular lens with an extended depth of focus. The lens has a wavefront-designed anterior surface, and a posterior achromatic diffractive surface with an echelette design.

Group III, TECNIS bifocal (ZKB00) is a hydrophobic aspheric diffractive bifocal intraocular lens with a minimum addition power of +2.75D. The lens has an anterior aspheric and posterior diffractive surface, where diffractive rings are set more towards the optic periphery, acting as more pupil-independent, providing near add power of 2.75D.

Group IV, TECNIS monofocal (ZCB00) is a hydrophobic aspheric with an anterior aspheric surface. This is a non-toric monofocal IOL designed to provide good distance visual acuity and a low incidence of dysphotopsia, but it does not correct near vision.

Clinical records of 150 PCIOL subjects (50/group) and 50 monofocal cases (group IV) that had been previously treated in the trial were assessed. The risks and benefits of each IOL type had been fully discussed with each subject prior to implantation.

IOL Power Selection

The Barrett Universal II formula was applied to calculate the IOL power using the IOL Master (IOL Master 700, Carl Zeiss Meditec AG, Jena, Germany). For each patient, the goal was to reach emmetropia in both eyes at 6 m.

Pre- and Postoperative Assessments

All patients had undergone examinations including tests for monocular UCDVA (logMAR chart at 6 m), UCIVA, UNDVA (tabletop adapted ETDRS charts at 80 cm and 40 cm), the respective corrected acuities, subjective refraction and autorefractometry (back vertex distance fixed at 12mm), corrected distance visual acuity (CDVA) at 10% and 25% contrast (computerized screen of Sloan letters with sizes decreasing in 0.1 logMAR steps at the appropriate contrast settings), applanation tonometry, slit-lamp assessment, fundoscopy, corneal tomography (Pentacam, Oculus, Wetzler, Germany), biometry (IOL Master700, Carl Zeiss Meditec AG, Jena, Germany), aberrometry (Luneau Wavefront L40 abberometer, Luneau Visionix, Prunay-le-Gillon, France), endothelial microscopy (CSO Perseus, Costruzione Strimenti Oftalmici, Florence, Italy), OCT of macula and optic nerve. Contact lens wearers had discontinued lens wear for up to four weeks, depending on the type of contact lenses, before the preoperative examination. Patients were examined 1 day, 1 week, 1 month, 6 months, and 12 months after surgery. UCDVA, UCIVA, UNDVA, UCDVA, refraction, distance corrected acuities at intermediate and near, CDVA at 10% and 25% contrast, tonometry, biomicroscopy and dilated fundus assessments were performed during each follow-up visit. Acuities were checked under ambient photopic conditions (∼85 cd/m²). The acuity was recorded as the line where the patient identified all letters correctly but was unable to do so on the subsequent line. All subjective tests were performed in the same room, using the same uniformly illuminated test charts, under the same photopic conditions and protocol. HOA values for primary spherical aberration, trefoil, coma, and RMS of the total (RMSHOA) were measured under mesopic conditions for pupil sizes of 3mm and 5mm.

Surgical Procedure

All cases had undergone standard phacoemulsification using a microcoaxial surgical technique (White Star Signature ^TM Phacoemulsification System, Johnson and Johnson) and were performed by a single experienced surgeon. After a superior clear corneal incision, anterior capsulorexis, and lens phacoemulsification, in each case, the IOL was injected into the capsular bag using an injector in all cases following a superior clear corneal incision, anterior capsulorexis, and phacoemulsification. The viscoelastic was removed by irrigation and aspiration. All incisions are sutureless and self-sealing. At the conclusion of surgery, an antibiotic drop was introduced to the eye, and an antibiotic ointment was spread into the conjunctival sac. Combined neomicine-dexamethasone drops were administered and gradually tapered off over 2 weeks after surgery.

Data and Statistical Analysis

Pre-operative refractive data were transformed using the procedure proposed by Thibos et al¹¹ This step-by-step process requires calculating the M, J₀, and J₄₅ vectors, and finally B. For a sphero-astigmatic refractive error where S and C are the powers of the sphere and astigmatism, θ is the axis of the astigmatism,

B is a single figure that eloquently describes the sphero-astigmatic refractive error, making it amenable to statistical scrutiny.

In all instances, a change (Δ) in acuity was defined as preoperative minus postoperative value in LogMAR units.

Data were stored in Excel spreadsheets (Microsoft, Redmond, WA) and, for each group, analyzed to determine the significance of:

• i)Changes in the uncorrected LogMAR visual acuities (UCDVA, UCIVA, UCNVA) over the period and association with preoperative values of HOAs (Pearson correlation).
• ii)Changes in the corrected distance LogMAR visual acuities (CDVA), at 10% and 25%, over the period and association with preoperative values of HOAs (Pearson correlation).
• iii)Changes in the uncorrected LogMAR visual acuities (UCDVA, UCIVA, UCNVA) over the period and association with preoperative B values (Pearson correlation).
• iv)Changes in the corrected distance LogMAR visual acuities, at 10% and 25%, over the period and association with preoperative B values (Pearson correlation).

Depending on these analyses, data were further subjected to multilinear regression to determine whether useful models could be developed to estimate likely changes in uncorrected acuities and CDVA (10% and 25%) using the preoperative values obtained in each group.

The results before surgery and at 1, 6, and 12 months after PCIOL implantation were assessed to determine if there were significant differences between groups I to III (1-way ANOVA or Kruskal– Wallis tests). If significant differences were not revealed, then the results for the PCIOL groups could be interpreted as being drawn from a single common pool. In that case, the pooled data would be analyzed to determine whether models (using single- and multiple-regression) could be developed to estimate changes in uncorrected acuities using preoperative values, irrespective of PCIOL design.

If models could be developed, then these would be tested to determine the limits of agreement (LoA) between predicted and actual changes in uncorrected acuities and CDVA (10% and 25%).

Appropriate non-parametric tests were applied when data were not normally distributed (Kolmogorov–Smirnov test of normality). Changes and differences were considered statistically significant at p <0.05. Where appropriate, p-values were adjusted using the Holm-Bonferroni procedure.¹² method Significant correlations between right and left data are expected in bilateral cases aimed at reaching the same endpoint. Thus, data from the treated eye in unilateral surgery and the right eye in bilateral cases were included as means to minimize the chances of making a statistical error due to any inter-eye dependency.

Results

Thirty group I, thirty-five group II, thirty-four group III, and thirty-two group IV subjects attended follow-up sessions at one, six and twelve months after surgery without complications. Baseline and postoperative values for UCDVA, UCIVA, UCNVA, CDVA (at 10% and 25%), refractive errors, M, J₀, J₄₅, and B vectors, and HOAs are noted in Tables 1–3.

Table 1.

Uncorrected logMAR Visual Acuities, and Corrected logMAR Visual Acuities for 10% and 25% Contrast, at Pre- and Postoperative Stages

	Preop	1 Month	6 Months	12 Months
Group I: AcrySof IQ PanOptix (TFNT00), n = 30
UCDVA	0.40, 0.40 (−0.12 to 0.92)	0.02, 0.00 (−0.07 to 0.05)	0.02, 0.00 (−0.03 to 0.07)	0.01, 0.00 (−0.04 to 0.06)	<0.001
UCIVA	0.45, 0.35 (−0.10 to 1.00)	0.00, 0.00 (0.00 to 0.00)	0.00, 0.00 (0.00 to 0.00)	0.00, 0.00 (0.00 to 0.00)	<0.001
UCNVA	0.60, 0.60 (−0.05 to 1.15)	0.00, 0.00 (−0.03 to 0.03)	0.00, 0.00 (0.00 to 0.00)	0.00, 0.00 (0.00 to 0.00)	<0.001
A	0.10, 0.10 (−0.03 to 0.23)	0.10, 0.10 (−0.02 to 0.22)	0.10, 0.10 (−0.05 to 0.25)	0.10, 0.10 (0.00 to 0.20)	0.296
B	0.05, 0.05 (−0.01 to 0.11)	0.02, 0.00 (−0.03 to 0.07)	0.02, 0.00 (−0.07 to 0.11)	0.02, 0.02 (−0.03 to 0.07)	0.243
Group II: TECNIS Symfony (ZXR00), n = 35
UCDVA	0.51, 0.80 (−0.19 to 1.21)	0.00, 0.00 (−0.04 to 0.04)	0.00, 0.00 (−0.03 to 0.03)	0.00, 0.00 (−0.03 to 0.03)	<0.001
UCIVA	0.45, 0.40 (−0.15 to 1.05)	0.00, 0.00 (0.00 to 0.00)	0.00, 0.00 (0.00 to 0.00)	0.00, 0.00 (0.00 to 0.00)	<0.001
UCNVA	0.55, 0.70 (−0.15 to 1.25)	0.00, 0.00 (0.00 to 0.00)	0.00, 0.00 (−0.05 to 0.05)	0.00, 0.00 (0.00 to 0.00)	<0.001
A	0.10, 0.10 (−0.10 to 0.30)	0.10, 0.10 (0.00 to 0.20)	0.09, 0.10 (−0.01 to 0.19)	0.10, 0.10 (0.00 to 0.20)	0.932
B	0.05, 0.02 (−0.05 to 0.15)	0.05, 0.05 (0.00 to 0.10)	0.04, 0.05 (−0.01 to 0.09)	0.05, 0.05 (0.00 to 0.10)	0.962
Group III: TECNIS bifocal (ZKB00), n = 34
UCDVA	0.52, 0.52 (−0.18 to 1.22)	0.00, 0.00 (−0.02 to 0.02)	0.00, 0.00 (−0.05 to 0.05)	0.00, 0.00 (−0.05 to 0.05)	<0.001
UCIVA	0.40, 0.40 (−0.20 to 1.00)	0.00, 0.00 (0.00 to 0.00)	0.00, 0.00 (0.00 to 0.00)	0.00, 0.00 (0.00 to 0.00)	<0.001
UCNVA	0.55, 0.70 (−0.15 to 1.25)	0.00, 0.00 (0.00 to 0.00)	0.00, 0.00 (0.00 to 0.00)	0.00, 0.00 (0.00 to 0.00)	<0.001
A	0.10, 0.10 (−0.02 to 0.22)	0.10, 0.10 (−0.11 to 0.31)	0.10, 0.10 (0.00 to 0.20)	0.10, 0.10 (0.00 to 0.20)	0.956
B	0.05, 0.05 (0.00 to 0.10)	0.02, 0.02 (−0.03 to 0.07)	0.02, 0.02 (−0.05 to 0.09)	0.04, 0.02 (−0.04 to 0.12)	0.387
Group IV: TECNIS monofocal (ZCB00), n = 32
UCDVA	0.40, 0.40 (−0.12 to 0.92)	0.07, 0.10 (−0.03 to 0.17)	0.03, 0.00 (−0.04 to 0.10)	0.02, 0.00 (−0.05 to 0.09)	<0.001
UCIVA	0.45, 0.35 (−0.10 to 1.00)	0.40, 0.40 (−0.20 to 1.00)	0.45, 0.45 (−0.10 to 1.00)	0.45, 0.40 (−0.10 to 1.00)	0.678
UCNVA	0.60, 0.60 (−0.05 to 1.25)	0.60, 0.60 (−0.07 to 1.27)	0.60, 0.58 (−0.05 to 1.25)	0.60, 0.55 (−0.08 to 1.28)	0.055
A	0.10, 0.10 (−0.03 to 0.23)	0.08, 0.05 (−0.12 to 0.28)	0.07, 0.05 (−0.03 to 0.17)	0.07, 0.05 (−0.03 to 0.17)	<0.001
B	0.05, 0.05 (−0.01 to 0.11)	0.02, 0.00 (−0.08 to 0.12)	0.02, 0.00 (−0.03 to 0.07)	0.02, 0.00 (−0.03 to 0.07)	<0.001

Notes: In groups 1–3, differences between pre- and postoperative values of i) UCDVA, UCIVA, UCNVA were significant (Friedman test, p < 0.001); ii) CDVA at 10% and 25% contrast were not significant (Friedman test, p > 0.05). In group 4, differences between pre- and postoperative values of UCDVA, CDVA at 10% and 25% were significant (Friedman test, p < 0.001) but not for UCIVA and UCNVA (p > 0.05).

Abbreviations: Median, mode (interquartile range) for uncorrected distance visual acuity (UCDVA), uncorrected intermediate visual acuity (UCIVA) and uncorrected near visual acuity (UCNVA). A, CDVA at 10% contrast (median, mode, interquartile range); B, CDVA at 25% contrast (median, mode, interquartile range).

Table 2.

Refractive Errors at Pre- and Postoperative Stages, and Vectorial Components (M, J₀, J₄₅ and B Vectors) of the Preoperative Refractive Errors

	Preop	1 Month	6 Months	12 Months	p
Group I: AcrySof IQ PanOptix (TFNT00), n = 30
S	1.38 (±0.53, 1.19 to 1.57)	0.00 (±0.35, −0.13 to 0.13)	0.07 (±0.26, −0.03 to 0.17)	0.05 (±0.26, −0.05 to 0.15)	<0.001
C	0.26 (±0.27, 0.16 to 0.36)	0.16 (±0.27, 0.06 to 0.26)	0.07 (±0.16, 0.01 to 0.13)	0.08 (±0.16, 0.02 to 0.14)	0.007
M, 1.50 (±0.58, 1.29 to 1.71). J0, 0.01 (±0.16, −0.05 to 0.07). J45, 0.02 (±0.09, −0.01 to 0.05). B, 1.51 (±0.59, 1.30 to 1.72)
Group II: TECNIS Symfony (ZXR00), n = 35
S	1.64 (±0.74, 1.20 to 1.88)	−0.20 (±0.50, −0.37 to 0.03)	−0.24 (±0.51, −0.42 to −0.05)	−0.08 (±0.21, −0.29 to −0.01)	<0.001
C	0.22 (±0.32, 0.12 to 0.32)	0.09 (±0.22, 0.02 to 0.17)	0.13 (±0.27, 0.04 to 0.22)	0.08 (±0.21, 0.01 to 0.15)	0.003
M, 1.75 (±0.75, 1.50 to 2.00). J0, −0.07 (±0.15, −0.02 to 0.02). J45, 0.02 (±0.09, −0.01 to 0.05). B, 1.76 (±0.75, 1.51 to 2.01)
Group III: TECNIS bifocal (ZKB00), n = 34
S	1.70 (±1.07,1.34 to 2.06)	−0.01 (±0.42, −0.16 to 0.14)	−0.08 (±0.33, −0.20 to 0.04)	−0.04 (±0.29, −0.14 to 0.06)	<0.001
C	0.30 (±0.31, 0.20 to 0.40)	0.16 (±0.33, 0.04 to 0.28)	0.10 (±0.26, 0 to 0.20)	0.10 (±0.26, 0 to 0.20)	0.011
M, 1.85 (±1.09, 1.48 to 2.22). J0, −0.04 (±0.17, −0.10 to 0.02). J45, 0.00 (±0.13, −0.04 to 0.04). B, 1.86 (±1.09, 1.50 to 2.22)
Group IV: TECNIS monofocal (ZCB00), n = 32
S	1.90 (±0.80,1.62 to 2.18)	−0.22 (±0.34, −0.34 to −0.10)	−0.17 (±0.33, −0.29 to −0.05)	−0.13 (±0.25, −0.22 to −0.04)	<0.001
C	0.38 (±39, 0.24 to 0.52)	0.18 (±0.31, 0.07 to 0.29)	0.16 (±0.29, 0.06 to 0.27)	0.10 (±0.23, 0.02 to 0.18)	0.003
M, 2.09 (±0.72, 1.84 to 2.34). J0, −0.09 (±0.21, −0.16 to −0.02). J45, −0.04 (±0.14, −0.09 to 0.01). B, 2.11 (±0.71, 1.86 to 2.36).

Notes: Differences between pre- and postoperative values of “S” and “C” were significant (p values according to 1-way ANOVA test for repeat measures). Between 1 and 12 months, differences in S and C were not significant (p > 0.05 for each of the 4 groups).

Abbreviations: Mean (±sd, 95% CI) spherical (S) and astigmatic (C, in positive format) powers, M, J₀, J₄₅ and B vectors. As noted in main text, M = S + (C/2), J₀ = (C/2).cos(2θ), J₄₅= (C/2).sin(2θ) and B = √(M² + J₀² + J₄₅)² where θ is the axis of astigmatism.

Table 3.

Higher Order Aberrations (HOAs) at Pre- and Postoperative Stages

3 mm Pupil Size
	RMSHOA	Coma	Trefoil	SA
Group I: AcrySof IQ PanOptix (TFNT00), n = 30
Preop	0.50 (±0.31, 0.39 to 0.61)	0.07(±0.06, 0.05 to 0.09)	0.06(±0.06, 0.04 to 0.08)	0.01(±0.03, 0.00 to 0.02)
1m	0.28 (±0.09, 0.25 to 0.31)	0.06(±0.03, 0.05 to 0.07)	0.05(±0.03, 0.04 to 0.06)	−0.01(±0.02, −0.02 to 0.00)
6m	0.27 (±0.09, 0.24 to 0.30)	0.05(±0.03, 0.04 to 0.06)	0.05(±0.02, 0.04 to 0.06)	−0.01(±0.02, −0.02 to 0.00)
12m	0.26 (±0.07, 0.24 to 0.28)	0.05(±0.04, 0.04 to 0.06)	0.04(±0.02, 0.03 to 0.05)	−0.01(±0.02, −0.02 to 0.00)
F, p	12.16, <0.001	0.431, >0.05	1.082, >0.05	5.007, <0.001
Group II: TECNIS Symfony (ZXR00), n = 35
Preop	0.73 (±0.91, 0.43 to 1.03)	0.12(±0.35, 0.00 to 0.24)	0.12(±0.37, 0.00 to 0.24)	0.02(±0.14, −0.03 to 0.07)
1m	0.33 (±0.14, 0.28 to 0.38)	0.06(±0.04, 0.05 to 0.08)	0.05(±0.04, 0.04 to 0.06)	−0.01(±0.02, −0.02 to 0.00)
6m	0.33 (±0.14, 0.29 to 0.37)	0.05(±0.04, 0.04 to 0.06)	0.05(±0.04, 0.04 to 0.06)	−0.01(±0.02, −0.02 to 0.00)
12m	0.32 (±0.14, 0.27 to 0.37)	0.05(±0.04, 0.04 to 0.06)	0.05(±0.04, 0.04 to 0.06)	−0.01(±0.02, −0.02 to 0.00)
F, p	6.005, <0.001	1.122, >0.05	0.981, >0.05	1.449, >0.05
Group III: TECNIS Bifocal (ZKB00), n = 34
Preop	0.67 (±0.28, 0.57 to 0.77)	0.05(±0.04, 0.04 to 0.06)	0.04(±0.03, 0.03 to 0.05)	0.01(±0.02, 0.00 to 0.02)
1m	0.27 (±0.08, 0.24 to 0.30)	0.06(±0.03, 0.05 to 0.07)	0.04(±0.02, 0.03 to 0.05)	−0.01(±0.02, −0.02 to 0.00)
6m	0.30 (±0.08, 0.27 to 0.33)	0.07(±0.05, 0.05 to 0.09)	0.04(±0.03, 0.03 to 0.05)	0.00(±0.02, −0.01 to 0.01)
12m	0.30 (±0.11, 0.26 to 0.34)	0.06(±0.03, 0.05 to 0.07)	0.04(±0.03, 0.03 to 0.05)	0.00(±0.02, −0.01 to 0.01)
F, p	39.95, <0.001	2.172, >0.05	1.231, >0.05	2.93, 0.023
Group IV: TECNIS Monofocal (ZCB00), n = 32
Preop	0.81 (±0.32, 0.70 to 0.92)	0.08(±0.06, 0.06 to 0.10)	0.08(±0.07, 0.06 to 0.10)	0.02(±0.05, 0.00 to 0.024)
1m	0.30 (±0.09, 0.27 to 0.33)	0.07(±0.03, 0.06 to 0.08)	0.06(±0.03, 0.05 to 0.07)	−0.01(±0.03, −0.02 to 0.00)
6m	0.30 (±0.08, 0.26 to 0.32)	0.07(±0.03, 0.06 to 0.08)	0.06(±0.03, 0.05 to 0.07)	−0.02(±0.03, −0.03 to −0.01)
12m	0.30 (±0.08, 0.26 to 0.32)	0.10(±0.08, 0.07 to 0.13)	0.06(±0.03, 0.05 to 0.07)	−0.02(±0.03, −0.03 to −0.01)
F, p	57.917, <0.001	0.668, >0.05	0.801, >0.05	3.64, 0.007
5 mm Pupil Size
	RMSHOA	Coma	Trefoil	SA
Group I: AcrySof IQ PanOptix (TFNT00), n = 30
Preop	1.10 (±0.76, 0.83 to 1.37)	0.14(±0.16, 0.08 to 0.20)	0.16(±0.27, 0.06 to 0.26)	0.04(±0.05, 0.02 to 0.06)
1m	0.44 (±0.19, 0.37 to 0.51)	0.12(±0.08, 0.09 to 0.15)	0.10(±0.08, 0.06 to 0.12)	0.00(±0.03, −0.01 to 0.01)
6m	0.43 (±0.16, 0.37 to 0.49)	0.10(±0.06, 0.08 to 0.12)	0.09(±0.07, 0.06 to 0.12)	0.00(±0.03, −0.01 to 0.01)
12m	0.41 (±0.17, 0.35 to 0.47)	0.10(±0.06, 0.08 to 0.12)	0.10(±0.08, 0.07 to 0.13)	−0.01(±0.03, −0.02 to 0.00)
F, p	17.48, <0.001	1.129, >0.05	1.877, >0.05	7.411, <0.001
Group II: TECNIS Symfony (ZXR00), n = 35
Preop	1.51 (±1.07, 1.15 to 1.43)	0.19(±0.36, 0.07 to 0.30)	0.20(±0.38, 0.08 to 0.32)	0.05(±0.15, 0.00 to 0.10)
1m	0.52 (±0.21, 0.45 to 0.59)	0.12(±0.12, 0.08 to 0.16)	0.11(±0.09, 0.08 to 0.14)	−0.01(±0.04, −0.02 to 0.00)
6m	0.52 (±0.22, 0.45 to 0.59)	0.12(±0.12, 0.08 to 0.16)	0.11(±0.10, 0.08 to 0.14)	−0.02(±0.04, −0.03 to −0.01)
12m	0.51 (±0.21, 0.50 to 0.52)	0.12(±0.11, 0.08 to 0.16)	0.11(±0.10, 0.08 to 0.14)	−0.02(±0.03, −0.03 to −0.01)
F, p	24.034, <0.001	0.942, >0.05	1.308, >0.05	4.96, <0.001
Group 3: TECNIS Bifocal (ZKB00), n = 34
Preop	1.49 (±0.78, 1.23 to 1.75)	0.10(±0.06, 0.08 to 0.12)	0.12(±0.08, 0.09 to 0.15)	0.05(±0.05, 0.03 to 0.07)
1m	0.46 (±0.16, 0.41 to 0.51)	0.12(±0.06, 0.10 to 0.14)	0.08(±0.06, 0.06 to 0.10)	0.00(±0.03, −0.01 to 0.01)
6m	0.49 (±0.18, 0.44 to 0.54)	0.12(±0.07, 0.10 to 0.14)	0.07(±0.06, 0.05 to 0.09)	0.00(±0.04, −0.01 to 0.01)
12m	0.47 (±0.18, 0.41 to 0.53)	0.10(±0.06, 0.08 to 0.12)	0.08(±0.07, 0.05 to 0.10)	0.00(±0.04, −0.01 to 0.01)
F, p	44.43, <0.001	0.345, >0.05	1.934, >0.05	10.402, <0.001
Group IV: TECNIS Monofocal (ZCB00), n = 32
Preop	1.59 (±0.70, 1.35 to 1.83)	0.16(±0.10, 0.12 to 0.20)	0.14(±0.12, 0.10 to 0.18)	0.05(±0.05, 0.03 to 0.07)
1m	0.42 (±0.13, 0.38 to 0.46)	0.09(±0.04, 0.08 to 0.10)	0.07(±0.04, 0.06 to 0.08)	−0.01(±0.03, −0.02 to 0.00)
6m	0.42 (±0.13, 0.37 to 0.47)	0.10(±0.05, 0.08 to 0.12)	0.07(±0.03, 0.06 to 0.08)	−0.01(±0.04, −0.02 to 0.00)
12m	0.40 (±0.10, 0.36 to 0.44)	0.10(±0.06, 0.08 to 0.12)	0.06(±0.05, 0.04 to 0.08)	−0.02(±0.02, −0.03 to −0.01)
F, p	69.89, <0.001	0.777, >0.05	1.173, >0.05	7.169, <0.001

Notes: Root mean square, coma, trefoil and primary spherical aberration measured under mesopic conditions for pupil sizes of 3mm and 5mm. All values are in µm. F and p are the results of 1-way ANOVA test for repeat measures. Mean (±sd, 95% CI) values (μm).

Abbreviations: RMS_HOA is the root mean square of total ocular aberrations, SA is the spherical aberration. 1m = at 1 month, 6m = at 6 months, 12m = at 12 months.

Changes in the Visual Acuities, Refractive Errors, and Higher-Order Aberrations (HOAs) Within Each Group

Table 1 shows that, in groups I–III, UCDVA, UCIVA, and UCNVA improved significantly following implantation (Friedman test, p < 0.001 for all comparisons). In contrast, the apparent changes in CDVA (at 10% and 25%) were not significant (p > 0.05). In group IV, UCDVA and CDVA (at 10% and 25%) improved, but UCIVA and UCNVA did not change significantly (p > 0.05).

Table 2 shows that the changes in refractive error resulting from implantation were significant across all 4 groups.

Table 3 shows that the mean values for coma and trefoil did not change significantly (1-way ANOVA, p > 0.05). However, the changes in the mean values for spherical aberration (SA) and RMS total HOAs (RMSHOA) were significant in each group for 3mm and 5mm pupil sizes.

Correlations between changes (Δ) in the uncorrected visual acuities and baseline values of i) higher order aberrations (HOAs), ii) B vector descriptor of refractive errors in each group (I–IV).

A total of 540 correlations were assessed, revealing several significant associations between the changes in UCDVA, UCIVA, UCNVA, and CDVA (at both 10% and 25%) following implantation and the respective preoperative HOAs and B values. The significant associations (p < 0.05) are listed in Table 4 together with cases (marked *) interpreted as insignificant according to the Holm-Bonferroni correction.

Table 4.

Significant (P ˂ 0.05) Correlations [r²] for the Associations Between Changes in the Uncorrected logMAR Visual Acuities, Reduced Contrast Corrected Distance Visual Acuities at Stages After Surgery and Preoperative Values of Higher Order Aberrations and Refractive Error

1 Month	3 mm Pupil						5mm Pupil				B
	RMSHOA	Trefoil			Coma	SA	RMSHOA	Trefoil	Coma	SA
Group I: AcrySof IQ PanOptix (TFNT00), n = 30
ΔUCDVA	0.228						0.169*				0.561
ΔUCIVA
ΔUCNVA
ΔCDVA (10%)						0.139*
ΔCDVA (25%)
Group II: TECNIS Symfony (ZXR00), n = 35
ΔUCDVA											0.210
ΔUCIVA
ΔUCNVA
ΔCDVA (10%)
ΔCDVA (25%)
Group III: TECNIS bifocal (ZKB00), n = 34
ΔUCDVA										0.133*	0.405
ΔUCIVA						0.139*
ΔUCNVA						0.212
ΔCDVA (10%)						0.237
ΔCDVA (25%)		0.125*				0.252
Group IV: TECNIS monofocal (ZCB00), n = 32
ΔUCDVA							0.392			0.229	0.161*
ΔUCIVA
ΔUCNVA								0.158*
ΔCDVA (10%)											0.168*
ΔCDVA (25%)											0.161*
6 months	3 mm pupil						5mm pupil				B
	RMSHOA		Trefoil	Coma		SA	RMSHOA	Trefoil	Coma	SA
Group I: AcrySof IQ PanOptix (TFNT00), n = 30
ΔUCDVA											0.558
ΔUCIVA
ΔUCNVA			0.157*	0.153*
ΔCDVA (10%)
ΔCDVA (25%)
Group II: TECNIS Symfony (ZXR00), n = 35
ΔUCDVA											0.458
ΔUCIVA
ΔUCNVA						0.117*				0.122*
ΔCDVA (10%)
ΔCDVA (25%)
Group III: TECNIS bifocal (ZKB00), n = 34
ΔUCDVA	0.181									0.178	0.568
ΔUCIVA
ΔUCNVA						0.137*
ΔCDVA (10%)						0.151*
ΔCDVA (25%)						0.158*					0.119*
Group IV: TECNIS monofocal (ZCB00), n = 32
ΔUCDVA											0.558
ΔUCIVA
ΔUCNVA			0.152*	0.153*						0.206
ΔCDVA (10%)
ΔCDVA (25%)
12 Months	3 mm Pupil						5mm Pupil				B
	RMSHOA		Trefoil	Coma		SA	RMSHOA	Trefoil	Coma	SA
Group I: AcrySof IQ PanOptix (TFNT00), n = 30
ΔUCDVA	0.200*										0.567
ΔUCIVA
ΔUCNVA
ΔCDVA (10%)
ΔCDVA (25%)
Group II: TECNIS Symfony (ZXR00), n = 35
ΔUCDVA											0.438
ΔUCIVA							0.145*
ΔUCNVA											0.138*
ΔCDVA (10%)
ΔCDVA (25%)
Group III: TECNIS Bifocal (ZKB00), n = 34
ΔUCDVA	0.193									0.247	0.462
ΔUCIVA											0.124*
ΔUCNVA						0.152*
ΔCDVA (10%)			0.156*
ΔCDVA (25%)			0.134*			0.118*
Group IV: TECNIS Monofocal (ZCB00), n = 32
ΔUCDVA							0.393			0.201	0.219
ΔUCIVA	0.137*
ΔUCNVA
ΔCDVA (10%)	0.129*						0.188				0.313
ΔCDVA (25%)	0.204						0.199				0.379

Notes: Changes in logMAR acuities (Δ) defined as value at preoperative stage minus value at postoperative stage. Empty cells indicate associations that were not significant (p > 0.05). *Correlation (ie, r² value) initially accepted as significant and subsequently deemed insignificant after applying the Holm-Bonferroni correction.

Abbreviations: SA is the spherical aberration. RMS_HOA is the root mean square of total ocular aberrations (trefoil, coma and spherical aberrations). B is the single figure vectorial description of the sphero-astigmatic refractive error at preoperative stage.

For groups I–III, the significant factors indicate that up to 25 models could be formulated to estimate changes in UCDVA, UCIVA, UCNVA, and CDVA (at both 10% and 25%). However, there is no single, overriding combination of HOAs that could be used to develop useful models to estimate changes in UCDVA, UCIVA, UCNVA, and CDVA (at both 10% and 25%) for each PCIOL. Turning to group IV, ΔUCDVA at both 1 and 12 months after surgery significantly correlated with preoperative SA (5mm), RMSHOA (3mm), and B. ΔUCIVA at 12 months was associated with RMSHOA (3mm). ΔUCNVA at 1 month was linked with trefoil (5mm), and by 6 months, ΔUCNVA was also linked with coma (3mm) and SA (5mm). ΔCDVA (10% and 25%) at 1 month correlated with B, while ΔCDVA (25%) at 12 months correlated with RMSHOA (3mm, 5mm) and B. The results were subjected to various transformations and linear regression (where appropriate, single- and multiple) to formulate models suitable for estimating changes in uncorrected acuities and CDVA (10% and 25%) using preoperative HOA and B values.

In theory, the factors indicated as no longer significant, according to the Holm-Bonferroni correction, should be excluded from any modelling. However, exclusion may prevent us from realizing the potential of a model that might emerge by inclusion of all initially detected as significant factors.

Comparing the Visual Acuities, Preop Refractive Errors, and Higher Order Aberrations (HOAs) Between Groups I, II and III

There were no statistically significant differences (p > 0.05 at preoperative, 1, 6 and 12 months after surgery) between groups I, II and III for:

[a] the B vector descriptor of refractive errors (1-way ANOVA),

[b] each of the eight HOAs for both 3mm and 5mm pupil sizes (1-way ANOVA),

[c] UCDVA (Kruskal–Wallis test),

[d] UCIVA (Kruskal–Wallis test),

[e] UCNVA (Kruskal–Wallis test),

[f] CDVA at 10% and 25% (Kruskal–Wallis test).

Therefore, the individual data sets from groups I, II, and III could be regarded as having been drawn from a single common pool. Hence, as an additional exercise, the datasets were merged and the analysis extended to determine whether any preoperative HOA values were associated with changes in UCDVA, UCIVA, UCNVA, or CDVA (10% and 25%), regardless of PCIOL design. The significant associations at 12 months after surgery are noted in Table 5 with cases considered as no longer remaining significant according to the Holm-Bonferroni correction.

Table 5.

Significant (P ˂ 0.05) Correlations [r²] for the Associations Between Changes in the Uncorrected logMAR Visual Acuities, Reduced Contrast Corrected Distance Visual Acuities at 1, 6 and 12 Months After Implantation and Preoperative Values of Higher Order Aberrations and Refractive Error for the Pooled Results from Groups I, II and III (n = 99)

	3 mm Pupil				5mm Pupil				B
	RMSHOA	Trefoil	Coma	SA	RMSHOA	Trefoil	Coma	SA
1 Month
ΔUCDVA					0.040*				0.348
ΔUCIVA
ΔUCNVA		0.037*		0.107				0.069
ΔCDVA (10%)				0.054*
ΔCDVA (25%)								0.080	0.044*
6 Months
ΔUCDVA									0.621
ΔUCIVA				0.045*
ΔUCNVA								0.053*
ΔCDVA (10%)
ΔCDVA (25%)
12 Months
ΔUCDVA			0.050*	0.052*					0.664
ΔUCIVA
ΔUCNVA		0.047*		0.080
ΔCDVA (10%)				0.054*
ΔCDVA (25%)				0.050*

Notes: Changes in logMAR acuities (Δ) defined as value at preoperative stage minus value at postoperative stage. Empty cells indicate associations that were not significant (p > 0.05). * Correlation (ie, r² value) initially accepted as significant and subsequently deemed insignificant after applying the Holm-Bonferroni correction.

Abbreviations: SA is the spherical aberration. RMS_HOA is the root mean square of total ocular aberrations (trefoil, coma and spherical aberrations). B is the single figure vectorial description of the sphero-astigmatic refractive error at preoperative stage.

The results were subjected to various transformations and linear regression (where appropriate, single- and multiple) to formulate models suitable for estimating changes in uncorrected acuities and CDVA (10% and 25%) using preoperative HOA and B values. The factors marked * in Table 5 were included for the reasons expressed earlier.

Models for Predicting UCDVA, UCIVA, UCNVA, and CDVA (10% and 25%) Resulting from Single and Multiple Regression

Regression analysis revealed significant associations (p < 0.05) between changes in UCDVA, UCIVA, UCNVA, and CDVA (10% and 25%) and certain covariates. The significant correlations are shown in Tables 4 and 5. Descriptions of the least squares lines describing these associations at 12 months after surgery are listed as equations 1–21 in Table 6, together with the predicted mean (±sd, 95% CI limits) changes in acuity values and the limits of agreement (±1.96sd) between the predicted and actual changes in acuity at twelve months.

Table 6.

Least Squares Expressions for the Key Significant (P < 0.05) Associations Between Predicted Changes in Uncorrected Visual Acuities at 12 Months After Implantation and Preoperative Values of Higher Order Aberrations And/Or Refractive Error

Least Squares Expressions		Mean (Predicted)	Mean (Actual)	LoA
I	y1= 0.362x1 − 0.186 (r2= 0.200). [eq.1]	0.37 (±0.25, 0.28–0.46)	0.37 (±0.22, 0.29–0.46)	±0.43
	y1= 0.287B + 0.137x1 − 0.136 (r2= 0.591). [eq.2]	0.37 (±0.19, 0.30–0.43)	0.37 (±0.22, 0.29–0.46)	±0.31
II	y1= 0.242B + 0.057 (r2= 0.438). [eq.3]	0.48 (±0.18, 0.42–0.54)	0.48 (±0.27, 0.39–0.57)	±0.40
	*y2= 0.044x2 + 0.382 (r2= 0.145). [eq.4]	0.45 (±0.05, 0.43–0.47)	0.45 (±0.12, 0.41–0.49)	±0.23*
	y3= 0.075B + 0.416 (r2= 0.138). [eq.5]	0.55 (±0.06, 0.53–0.57)	0.55 (±0.15, 0.50–0.60)	±0.28
III	y1= 0.589–2.635x8 (r2= 0.247). [eq.6]	0.46 (±0.13, 0.41–0.51)	0.46 (±0.27, 0.36–0.56)	±0.46
	y1= 0.184 + 0.414x1 (r2= 0.193). [eq.7]	0.20 (±0.02, 0.19–0.21)	0.46 (±0.27, 0.36–0.56)	±0.53
	y1= 0.188B + 0.24x1 − 1.452x8 + 0.179 (r2= 0.463). [eq.8]	0.62(±0.28, 0.53–0.71)	0.46 (±0.27, 0.36–0.56)	±0.39
	y2= 2.240x7 + 0.397 (r2= 0.124). [eq.9]	0.42 (±0.05, 0.40–0.44)	0.42 (±0.14, 0.37–0.47)	±0.25
	*y3= 2.106x7 + 0.533 (r2= 0.152). [eq.10]	0.55 (±0.05, 0.53–0.57)	0.55 (±0.11, 0.51–0.59)	±0.21*
	*y4 = 0.001 + 1.729x7 (r2= 0.156). [eq.11]	0.02 (±0.04, 0.01–0.03)	0.02 (±0.09, −0.01–0.05)	±0.17*
	*y5= 0.782x5 + 1.211x7 − 0.039 (r2= 0.226). [eq12]	0.01 (±0.03, 0.00–0.02)	0.01 (±0.08, 0.00–0.02)	±0.15*
IV	y1= 0.272 + 0.219x2 − 2.215x8 (r2= 0.617). [eq.13]	0.55 (±0.19, 0.48–0.62)	0.55 (±0.24, 0.47–0.63)	±0.29
	y1= 0.370 + 0.303x2 − 2.494x8 − 0.106B (r2= 0.651). [eq.14]	0.55 (±0.19, 0.48–0.62)	0.55 (±0.24, 0.47–0.63)	±0.28
	*y2= 0.175 − 0.119x1 (r2= 0.137). [eq.15]	0.08 (±0.04,0.07−0.09)	0.00 (±0.01, −0.03–0.03)	±0.19*
	*y4 = 0.220 − 0.070x2 (r2= 0.188). [eq.16]	0.09 (±0.09,0.06−0.12)	0.11 (±0.11, 0.07–0.15)	±0.20*
	*y5= 0.183 − 0.060x2 (r2= 0.199). [eq.17]	0.09 (±0.04, 0.07–0.11)	0.09 (±0.09, 0.06–0.12)	±0.16*
	*y5= 0.261 − 0.091B + 0.012x2 (r2= 0.383). [eq.18]	0.09 (±0.06, 0.07–0.11)	0.09 (±0.09, 0.06–0.12)	±0.14*
^	y1= 0.437 − 2.885x32 (r2= 0.490). [eq.19]	0.42 (±0.06, 0.41–0.43)	0.44 (±0.27, 0.39–0.49)	±0.51
	y1= 0.116 − 1.577x32 + 0.176 x9 (r2= 0.365). [eq.20]	0.41 (±0.16, 0.38–0.44)	0.44 (±0.27, 0.39–0.49)	±0.39
	y3= 0.537 + 0.700x7 (r2= 0.042) [eq.21]	0.54 (±0.03, 0.53–0.55)	0.55 (±0.13, 0.52–0.58)	±0.24*

Notes: Mean predicted and actual changes in logMAR acuities, HOA values in µm. Mean (±sd, 95% CI limits) predicted and actual changes of acuity, and respective limits of agreement (LoA = ±1.9sd) between these values are shown. Significant expressions featuring preoperative descriptions of subjective refractive error (B) are shown where the corresponding LoA values were either minimal or reduced by the inclusion of B. *Models where the range of LoA values were about ±0.20. ^ after pooling results from all 3 PCIOL groups.

Abbreviations: HOA (higher order aberrations) values in µm; change in uncorrected distance visual acuity (y₁); change in uncorrected intermediate visual acuity (y₂); change in uncorrected near visual acuity (y₃); change in corrected distance visual acuity at 10% (y₄) or 25% (y₅) contrast. RMS_HOA value of root mean square higher order aberrations for 3mm (x₁) and 5mm (x₂) pupil sizes. Coma values for 3mm (x₃) and 5mm (x₄) pupil sizes. Trefoil values for 3mm (x₅) and 5mm (x₆) pupil sizes. Spherical aberration (SA) values for 3mm (x₇) and 5mm (x₈) pupil sizes. B is the single figure vectorial description of the sphero-astigmatic refractive error at preoperative stage. I = AcrySof IQ PanOptix (TFNT00) n = 30, II = Symfony (ZXR00) n = 35, III = TECNIS bifocal (ZKB00) n = 34, IV = TECNIS monofocal (ZCB00) n = 32.

The variance inflation factors (VIFs) for the covariates in equations 1–21 were consistently below 3, indicating that multicollinearity is unlikely to affect the estimates when the covariates are treated as independent variables.^13–15

Discussion

Tables 1 and 2 show, as expected, the visual acuities improved and refractive errors reduced following implantation of the IOLs. On closer inspection, the mean values of the refractive errors were both small and comparable with the typical test-retest, or repeatability, of subjective refraction.^16–19 Furthermore, the associated s.d. values and ranges (as represented by the 95% confidence limits) of the postoperative residual refractive errors were comparable with the repeatability values. Thus, the average postoperative residual refractive error in the 4 groups was near zero. The typical eye after surgery was quasi-emmetropic.

Several studies report and compare postoperative visual acuities across different PCIOL designs.^20–36 The overarching impression is that subtle differences in the optical design of a PCIOL may contribute to differences in the magnitude of postoperative visual acuity. However, as shown in Table 1, even though the PCIOLs had different designs, the outcomes for unaided visual acuities were the same. No single PCIOL demonstrated a significant advantage over the others.

Reports indicate that changes in the profile of ocular higher-order aberrations (HOAs) occur after PCIOL implantation.^37–45 Increasing the HOAs have deleterious effects on retinal image quality, leading to symptoms such as difficulty with night vision, glare, halos, blurring, starburst patterns, and diplopia.^37–39 And, eyes implanted with PCIOLs have reduced contrast sensitivity compared with those implanted with monofocal IOLs.^46–52 The reason probably lies in the shift in the mode of light distribution from a single to two or more foci required for simultaneous vision with PCIOLs.^53–55 However, Table 1 shows that there were neither significant changes in CDVA at 10% and 25% contrast in groups I–III, nor significant differences in these CDVA values between the groups over the postoperative period. These results support other studies reporting no statistically significant differences in the outcomes between different PCIOL designs.^56–58

In contrast to these findings, in the monofocal group (IV), UCIVA and UCNVA remained unchanged, but CDVA at 10% and 25% had improved. In all four groups, there was a convergence towards emmetropia, as indicated by the improved UCDVA and ≈ zero levels of residual refractive errors, but the UCIVA and UCNVA, typically associated with presbyopia, remained unchanged in group IV. Table 3 shows that the mean values of the higher-order aberrations (HOAs), coma, and trefoil remained stable across all 3 PCIOL groups. Spherical aberration (SA) tended to shift from positive to negative, and overall RMSHOA decreased. For the 5mm pupil, the mean changes in SA ranged from 0.05µm to 0.07µm while the corresponding changes in RMSHOA ranged from 0.69µm to 1.19µm. Thus, it is reasonable to consider changes in RMSHOA and/or SA as likely factors associated with changes in uncorrected acuities following implantation. Table 4 shows that, other than pre-operative refractive error represented by vector “B”, no single descriptor of the ocular higher order aberrations stands out as being associated with the changes in uncorrected acuities recorded at 1-, 6-, or 12-months following IOL implantation. Various correlations were significant, but none emerged as particularly prominent.

The key results of multiple linear regression are shown in Table 6 and listed in Eqs. 1–21. In the three PCIOL groups, the limits of agreement (LoA) range ±0.15 (Group III, eq 12) to ±0.53 (Group II, eq 7). The repeatability of logMAR acuity testing ranges from ±0.01 to ±0.18 units.^59,60 Thus, nine of the 21 models in Table 6, with LoA values close to ±0.2 (marked *), could be viewed as suitable for estimating changes in the uncorrected logMAR acuities. Eq.4 could be used to estimate changes in UCIVA in group II (Tecnis Symfony) and eqs 10–12 may be used to estimate changes in UCNVA, CDVA at 10% and 25% in group III (Tecnis Bifocal). Turning to Figures 1 and 2, the bulk of data points are located between the hatched lines representing the limits ±0.10 units of logMAR acuity. So, the preop value of the root mean square higher order aberrations (RMS_HOA) for 5mm pupil size in conjunction with eq.4 could be used to provide a reasonable estimate for the likely change in UCIVA one year after implanting with Tecnis Symfony PCIOL. And the preop value of spherical aberration (SA) for 3mm pupil size in conjunction with eq.10 could be used to provide a reasonable estimate for the likely change in UCNVA one year after implanting with Tecnis bifocal PCIOL.

Figure 1.

Difference between estimated and actual change in uncorrected intermediate visual acuity (ΔUCIVA) in group II (TECNIS Symfony PCIOL).

Notes: Estimated changes were calculated using eq. 4 in Table 6. Visual acuity values are expressed in logMAR units. Hatched lines highlight ±0.10 range in ΔUCIVA.

Figure 2.

Difference between estimated and actual change in uncorrected near visual acuity (ΔUCNVA) in group III (TECNIS bifocal PCIOL).

Notes: Estimated changes were calculated using eq. 10 in Table 6. Visual acuity values are expressed in LogMAR units. Hatched lines highlight ±0.10 range in ΔUCNVA.

However, there were no significant differences in the clinical results between the three PCIOLs. On an amalgamated group basis, eq. 21 could be considered to estimate changes in UCNVA. The estimated and actual changes in UCNVA are shown in Figure 3 where the bulk of data points fall within limits of ±0.10 units. This suggests the preop value of spherical aberration (SA) for 3mm pupil size in conjunction with eq.21 could be used to provide a reasonable estimate for the likely change in UCNVA one year in over 67% of cases regardless of PCIOL design.

Figure 3.

Difference between estimated and actual change in uncorrected near visual acuity (ΔUCNVA) for the data pooled from three PCIOL groups.

Notes: Estimated changes were calculated using eq. 21 in Table 6. Visual acuity values are expressed in LogMAR units. Hatched lines highlight ±0.10 range in ΔUCNVA.

Biarnés et al⁶¹ identified the patient’s age and pupil size as factors that influenced the distance-corrected near visual acuity in cases implanted with partial-range-of-focus (often referred to as extended depth of focus) IOLs. In contrast to the current study, they did not report any significant association between changes in uncorrected corrected near visual acuity and HOAs such as coma or spherical aberration. Their follow-up was no more than 3 months, and they only investigated one type of IOL without any comparison groups.

Turning to group IV (Tecnis aspheric monofocal), the changes in UCDVA and UCIVA can be estimated using eqs 13 to 15, and the corresponding LoA values are equivalent to ±3 and ±2 lines on a standard logMAR chart. The design of this IOL is aimed at maximizing UCDVA, though predicting the change in UCIVA is more precise. For the CDVA at 10% and 25%, Figure 4 shows the difference between estimated and actual changes in these acuities also tend to fall between the ±0.10 units of logMAR acuity. The preop RMS_HOA for the 5mm pupil size and refractive error in conjunction with eqs. 17 and 18 could be used to reliably estimate the reduced contrast CDVA that can be achieved by one year after implantation.

Figure 4.

Difference between estimated and actual change in the corrected distance visual acuity (ΔCDVA) at contrast levels of 10% and 25% in group IV (TECNIS monofocal).

Notes: Estimated changes were calculated using eqs 16 (for 10%) and 18 (for 25%) in Table 6. Unfilled (filled) circles show results for changes in CDVA at 10% (25%) contrast. Visual acuity values are expressed in LogMAR units. Hatched lines highlight ±0.10 range in ΔCDVA.

The value of these models would be compromised should any complications occur or the development of PCO. The development of PCO is expected to affect visual acuity in all cases. Yet the evidence from the literature suggests that the prevalence of PCO is greater in PCIOL than in IOL cases.^62–67 However, signs of PCO were not evident in the cohort of cases monitored during this study.

Limitations of the Study

The results are wholly dependent on the inclusion criteria, patient demographics, lens designs, the number of cases that completed the study, and the data included in the analysis. Future investigators could extend the work by including additional objective ocular measures.

Conclusion

Changes in the uncorrected visual acuities following implantation of PCIOLs could be estimated using preoperative measurements of specific HOAs. However, the limits of agreement between estimated and actual changes of these acuities tend to rest outside ±0.3 units, equivalent to ±3 lines, on a LogMAR chart. Estimating the likely change in UCNVA and CDVA (at 10% and 25%) falls within ±2 lines and can be interpreted as being more reliable in cases planned for implantation with the Tecnis Bifocal PCIOL in comparison with those intended for implanting with the other two PCIOLs. Changes in CDVA (at 10% and 25%) could be estimated within ±2 lines using preoperative measurements of selected HOAs for cases planned for implantation with the Tecnis aspheric monofocal IOL.

Data Sharing Statement

Data supporting the findings of this investigation are available from the corresponding author upon reasonable request.

Ethics Approval and Informed Consent

The initial clinical study had been approved by the Ethics Committees of the University Eye Hospital “Svjetlost” in Zagreb, and the tenets of the Helsinki agreement were followed throughout. There was no legal requirement to issue an ethical approval number at the time the study was approved. Signed informed consent had been obtained from all patients after fully explaining all the procedures, risks, and benefits.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors report no conflicts of interest in this work. None of the authors have any financial interest in any of the products or procedures mentioned in this paper. The study was self-financed.