If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author: Xiu Hua Wan, MD, Beijing Institute of Ophthalmology, Beijing Tongren Eye Center, Beijing Tongren Hospital of Capital Medical University, Beijing Key Laboratory of Ophthalmology and Visual Sciences, No. 17, Hougou Ally, Dongcheng District, Beijing 100005, China.
We compare multifocal intraocular lenses (MFIOLs) to monofocal IOLs for visual acuity (VA), contrast sensitivity, and adverse events using data from 21 randomized controlled trials with 2951 subjects. There was no statistical difference between uncorrected distance VA and corrected distance VA. Compared with monofocal IOLs, MFIOLs showed a better performance on uncorrected intermediate VA measured at 60 cm and uncorrected near VA; the mean differences were -0.06 (95% confidence interval [CI]: -0.10, -0.03) and -0.13 (95% CI: -0.20, -0.07). Distance-corrected intermediate VA and distance-corrected near VA were measured wearing distance correction. MFIOLs performed better than monofocal IOLs on distance-corrected intermediate VA at 60 cm and distance-corrected near VA; the mean differences were -0.09 (95% CI: -0.12, -0.06) and -0.31 (95% CI: -0.43, -0.19). The contrast sensitivity of the MFIOL group was lower than that of the monofocal IOL group; mean difference was -0.06 (95% CI: -0.11, -0.02). More patients were spectacle free in the MFIOL group; the risk ratio was 2.86 (95% CI: 1.73, 4.73). More patients were troubled by glare and halos in the MFIOL group; the risk ratios were 1.91 (95% CI: 1.24, 2.95) and 3.08 (95% CI: 2.11, 4.49). We conclude that, compared with monofocal IOLs, MFIOLs give patients better near vision and intermediate vision at 60 cm, both corrected and uncorrected. Patients undergoing MFIOLs implantation are more likely to be spectacle free but have a higher risk of glare, halos, and lower contrast sensitivity.
many kinds of IOLs, such as monofocal, bifocal, trifocal, and so forth, have been developed. Monofocal IOLs provide patients with good vision, but only at 1 focal distance. Many patients receiving monofocal IOLs desire clearer near vision. Multifocal intraocular lenses (MFIOLs) with multiple focal distances have been FDA-approved since 1997.
MFIOLs, however, are not perfect, with reports of reduced contrast and increased risk of glare and halos. Many researchers had tried to evaluate the advantages and disadvantages of different IOLs. In 2008, Li and coworkers
but neither evaluated distance-corrected intermediate VA outcome. Besides, they included 2 kinds of MFIOLs that were no longer in use.
It remained controversial whether MFIOLs were superior to monofocal IOLs. Our systematic review and meta-analysis provides an updated assessment of the VA, contrast sensitivity (CS), and adverse outcomes of MFIOLs and monofocal IOLs implantation in elderly cataract patients based on randomized controlled trials.
2. Materials and methods
2.1 Inclusion criteria
To assess only high-quality evidence, we only included randomized controlled trials. The intervention measures were monofocal IOLs implantation and MFIOLs implantation. The target population was cataract patients aged 50 years or older.
2.2 Databases and search strategy
We searched databases of Pubmed, Science Direct, the Chinese National Knowledge Infrastructure, and the Wanfang Database. We used the following terms or their combinations to conduct the literature search: randomization, randomized, randomly, trial, comparison, monofocal, multifocal, IOLs, intraocular lenses. The full search strategy was shown in Method of Literature search part. We restricted the publication date to be between January 1, 1998 and December 31, 2017.
We compared both corrected and uncorrected binocular VA, including distance VA, intermediate VA, and near VA. We only extracted logMAR VA.
For the measurement of distance VA, 5 studies used Early Treatment for Diabetic Retinopathy Study (ETDRS) acuity charts
letter charts, and 1 study did not report the measurement tool. For the measurement of intermediate VA, 2 studies used EDTRS, 1 study used Snellen acuity with Sloan optotypes (ACP-8 Auto Chart Projector, Topcon Europe BV, Milano, Italy), and 1 study did not report the measurement tool. For the measurement of near VA, 1 study used Rosenbaum near acuity card, 2 studies used Bailey-Lovie logMAR word reading acuity charts, 2 studies used EDTRS, and another 2 studies did not mention the measurement tool.
Spectacle independence (glasses free after IOLs implantation)
For the measurement of contrast sensitivity, 4 studies used Pelli-Robson contrast sensitivity chart to perform measurement, 2 studies used CSV-1000 (which presents sine wave gratings at 3, 6, 12, and 18 cycles/degree frequencies), and 1 study used the Vision Contrast Test System.
We only extracted data measured at the last follow-up time. For categorical data, we extracted the number of events of each treatment group, such as the number of spectacle independence and the number of glare and halos. For continuous data like VA, the mean values and standard deviations were extracted.
2.5 Assessment on risk of bias and paper quality
We applied the risk of bias tool recommended by the Cochrane Collaboration
to evaluate the risk of bias of included studies from 6 aspects: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessors, incomplete outcome data, selective outcome reporting. The risk of bias would be classified into high level, low level, or unclear according to each study's description.
2.6 Data synthesis and statistical analysis
We used mean deviation (MD—mean value of MFIOLs minus mean value of monofocal IOLs) with 95% confidence interval (CI) to estimate continuous outcomes, and risk ratio (RR) with 95% CI to estimate categorical outcomes. We conducted all the analyses using open source R program (version 3.4.4). The significance level was 0.05, two tailed.
For studies with more than 2 MFIOLs treatments, we combined mean values and standard deviations of each MFIOLs group according to Formula (1) and Formula (2), respectively. For categorical data, we added up the number of events of each MFIOLs group; all aforementioned statistical methods were recommended by the Cochrane handbook.
as a way to adjust the effect for funnel plot asymmetry.
2.7 Assessment on heterogeneity
Before calculating the pooled MD of pooled RR, we assessed the heterogeneity across studies firstly using I2 statistic. If I2 were above 50%, a random-effects model (DerSimonian-Laird method) would be applied to do the meta-analysis,
Analysis on risk of bias (Fig. 2 and Fig. 3) showed that all studies generated the random sequence in an appropriate way. Only a few studies had a high-level risk of bias on blinding, incomplete data, or selective reporting.
There was no statistical difference on uncorrected distance VA (UDVA) between MFIOLs and monofocal IOLs (Fig. 4); the MD was 0.02 (95% CI: 0.00, 0.03). The I2 was 21%, indicating that the heterogeneity across included studies was small. Two researchers measured uncorrected intermediate VA (UIVA) at the distance of 60 cm, and they both reported a better UIVA of patients undergoing MFIOLs implantation than those received monofocal IOLs implantation (Fig. 5); the MD was -0.06 (95% CI: -0.10, -0.03), and the I2 was 49%. There was no statistical difference on UIVA at the distance of 63 cm; the MD was 0.03 (95% CI: -0.02, 0.08). At a measured distance of 100 cm, 1 researcher reported a better UIVA of patients in the monofocal IOLs group than in the MFIOLs group; the MD was 0.07 (95% CI: 0.04, 0.10). Patients after MFIOLs implantation had a better uncorrected near VA (UNVA) than those accepting monofocal IOLs surgery (Fig. 6); the MD was -0.13 (95% CI: -0.20, -0.07). The I2 was 91%, indicating a large heterogeneity across included studies; thus, a subgroup analysis was done (Fig. 6). We split studies into 3 subgroups by the measurement distance of UNVA in each study; it turned out that no matter at the distance 30 cm, 33 cm, or 40 cm, patients undergoing MFIOLs implantation achieved a better UNVA than those accepting monofocal IOLs implantation; the MDs were -0.30 (95% CI: -0.36, -0.24), -0.08 (95% CI: -0.16, -0.01), and -0.12 (95% CI: -0.20, -0.04).
There was no statistical difference on corrected distance VA (CDVA) between MFIOLs and monofocal IOLs (Fig. 7). The I2 was 80%, indicating a large heterogeneity across included studies. We performed a subgroup analysis by the measurement distance of CDVA in each study (Fig. 7). We also performed a sensitivity analysis, and it turned out that no matter in which subgroup, no matter which study was omitted, there was no statistical difference on CDVA. Different studies measured distance-corrected intermediate VA at difference distances, including 60 cm, 63 cm, and 80 cm. Specifically, at the distance of 60 cm, patients' distance-corrected intermediate VA of the MFIOLs group was better than that of the monofocal IOLs group (Fig. 8); the MDs were -0.09 (95% CI: -0.12, -0.06). However, at the distances of 63 cm and 80 cm, there was no statistical difference; the MDs were 0.03 (95% CI: -0.02, 0.08) and 0.06 (95% CI: -0.05, 0.17). MFIOLs showed a better performance than monofocal IOLs on distance-corrected near VA (DCNVA) (Fig. 9); the MD was -0.31 (95% CI: -0.43, -0.19); however, the I2 was 96%, indicating a large heterogeneity across included studies. We also did a subgroup analysis (Fig. 9), splitting studies into 4 subgroups according to the measurement distance used in each study. It turned out that no matter at 30 cm, 33 cm, or 40 cm, MFIOLs showed a better performance than monofocal IOLs. The MDs were -0.38 (95% CI: -0.44, -0.32), -0.14 (95% CI: -0.18, -0.10), and -0.33 (95% CI: -0.56, -0.10). The I2 of the 40 cm subgroup was still above 50% because the CI of Peng and coworkers study did not overlap with the others. Actually, all 3 studies in the 40 cm subgroup reported a better DCNVA of MFIOLs.
The CS of the MFIOLs group was lower than that of the monofocal IOLs group (Fig. 10); the MD was -0.06 (95% CI: -0.11, -0.02). The heterogeneity was slightly large as the I2 was 59%, so a random-effects model was applied, and we also did a sensitivity analysis. It turned out that, after omitting study of Maxwell and coworkers, the I2 dropped to a mere 3.3%, and the CS of the MFIOLs group was still lower than monofocal IOLs group. The MD was -0.04 (-0.07, -0.01), and the conclusion remained unchanged.
More patients in the MFIOLs group were spectacle free than those in the monofocal IOLs group (Fig. 11); the RR was 2.86 (95% CI: 1.73, 4.73). The I2 was 94%, however, indicative of a large heterogeneity. A subgroup analysis was done (Fig. 11), which showed that MFIOLs performed better than monofocal IOLs at 3, 6, or 12 months after implantation. The RR was 1.85 (95% CI: 1.20, 2.85), 4.03 (95% CI: 2.21, 7.36), and 4.03 (95% CI: 1.56, 10.41). We also did a sensitivity analysis, and it turned out that, after removing studies of Y and coworker, Maxwell and coworkers, and Maxwell and coworkers, the I2 dropped to 0% and the conclusion remained the same, with an RR of 2.83 (95% CI: 2.37, 3.39). In addition, for spectacle independence, a publication bias was found (t = 3.702, P = 0.004) by Egger's test (Table 2). We used a trim-and-fill method to adjust the effect for funnel plot asymmetry, after which the conclusion remained unchanged with an RR of 1.68 (95% CI: 1.14, 2.49). Clearly, MFIOLs performed better than monofocal IOLs on spectacle independence.
More patients in the MFIOLs group suffered from glare than in the monofocal IOLs group (Fig. 12). The RR was 1.91 (95% CI: 1.24, 2.95), and there was no heterogeneity. Similarly, more patients were troubled with halos after MFIOLs implantation (Fig. 13). The RR was 3.08 (95% CI: 2.11, 4.49), and the I2 was 16%, indicative of a small heterogeneity.
For all evaluated outcomes, no publication bias was found by Egger's test (Table 2) except for spectacle independence.
For outcomes including UDVA, CDVA, UNVA, DCNVA, CS, and spectacle independence, we also did a subgroup analysis by MFIOLs types; MFIOLs were split into refractive and diffractive groups; the results were shown in the Appendix. It turned out that, for refractive MFIOLs or diffractive MFIOLs, there was no difference on UDVA and CDVA compared with monofocal IOLs. Both refractive MFIOLs and diffractive MFIOLs provided patients with a better UNVA, a better DCNVA, but a lower CS compared with monofocal IOLs.
Cataract patients are eager for a high visual quality after IOL implantation. They want clear vision at different focal points; thus, MFIOLs were designed. There had been evidence showing that the major advantage of MFIOLs is the restoration of near vision function for cataract patients. Back in the 1990s, El-Maghraby and coworkers
randomly allocated 77 cases to receive a 3M MFIOL or a conventional monofocal implantation. After a 4-month follow-up, 30% of the MFIOLs cases had near acuity J1, compared to only 4% of the monofocal IOLs cases. For the newly approved MFIOLs, like PanOptix from Alcon,
also revealed the superiority of PanOptix's near vision.
In this meta-analysis, we also found that MFIOLs provided better near vision than monofocal IOLs, especially at reading distance, which is usually 30 cm to 33 cm. Our meta-analysis showed the gap between monofocal IOLs and MFIOLs was up to 0.3 log units for UNVA and 0.38 log units for DCNVA. Given that loss of visual ability at the reading distance had a strong impact on patients' life quality, even on mobility orientation and the avoidance of falls,
MFIOLs are a better choice for cataract patients. Of course, MFIOL is not the only way to improve unaided reading ability and quality of life. Monovision by using monofocal lenses provides many advantages of the multifocal lenses without the photic adverse effects. Stock and coworkers
had already revealed MFIOLs' superiority over monofocal IOLs, and from this meta-analysis based on randomized controlled trials, we obtained a similar finding. Specifically, for UIVA, 4 studies were included, and only Wilkins and coworkers reported a worse UIVA of MFIOLs patients measured at 100 cm, which was reasonable because usually intermediate VA would be measured at a distance between 60 cm and 70 cm. Actually, some researchers tended to define VA measured at 100 cm as distance VA.
Only 1 study reported the measurement of UIVA at the distance of 63 cm, and no statistical difference was found between MFIOLs and monofocal IOLs, while at the distance of 60 cm, both Peng and coworkers and Maxwell and coworkers reported a better UIVA of the MFIOLs group. The mean differences of UIVA between MFIOLs and monofocal IOLs at 100 cm, 63 cm, and 60 cm were 0.07, -0.03, and -0.07, respectively, indicating that there might be a tendency that the nearer the distance was, the better the UIVA would be; distance-corrected intermediate VA showed a similar distribution pattern as UIVA.
Theoretically, multifocal lenses would split light among multiple focal points and lead to a reduced contrast on retinal image, which may arise functional loss of contrast sensitivity. In the 1990s, Lehmann and coworkers
did another comparison on CS between MFIOLs and monofocal IOLs, and it turned out that only near vision showed a 0.19 log units decrease, while for distance vision, there was no decrease. In this meta-analysis, we did observe a lower contrast sensitivity in MFIOLs patients; however, the gap between the MFIOLs group and monofocal IOLs group was only 0.06 unit; thus, the disadvantage of MFIOLs on CS was not large.
In this meta-analysis, 12 researchers recorded the outcome of spectacle independence, and in all these studies, more MFIOL patients were free from glasses than those receiving monofocal IOLs. The pooled RR was up to 2.86, which was a large effect size. MFIOLs performed far better than monofocal IOLs on spectacle independence; however, by this meta-analysis, the disadvantage of MFIOLs on glare and halos was clear as, in every single study that recorded outcomes of glare or halos, there were more cases in the MFIOLs group. There was no heterogeneity across included studies for glare; the heterogeneity for halos was small (I2 = 16%). Unfortunately, glare, as well as halos, is a natural flaw of MFIOLs. Theoretically, multiple focal points will induce more light scatter than single focal point,
and forward scattered light out of a glare source will generate a veil of luminance on the retina. In addition, the out-of-focus image tends to show a larger diameter than the sharp image, which forms halos.
We need to acknowledge that there are many types of MFIOLs. The difference among refractive, diffractive, and segmental lenses and the difference among bifocal, trifocal, and extended depth of focus lenses are not negligible, but because of the limited number of published papers with extractable data, for some outcomes, we have to include all MFIOLs types. For outcomes such as UDVA, CDVA, UNVA, DCNVA, CS, and spectacle independence, there were enough studies, and we did a subgroup analysis by MFIOLs types; MFIOLs were split into refractive and diffractive groups; however, it seemed that, although refractive lenses were different from diffractive lenses in optical design, their performance was much alike. Both refractive MFIOLs and diffractive MFIOLs helped patients achieve a better UNVA and DCNVA compared with monofocal IOLs, while a lower CS was found in both subgroups. As a comparison between refractive MFIOLs and diffractive MFIOLs, Yildirim and coworkers
conducted a prospective, nonrandomized study where 20 patients were implanted with refractive ReZoom NXG1 IOLs in their dominant eye, while diffractive Tecnis ZMA00 IOLs were implanted in the nondominant eye. After a 6-month follow-up, there was no statistical difference between near VA, CS, reading speed, halos, or glare.
In conclusion, compared with monofocal IOLs aimed for emmetropia, MFIOLs have advantages in regards to near vision and intermediate vision at 60 cm, and patients having MFIOLs implanted are more likely to be spectacle free. There are drawbacks as patients undergoing MFIOLs implantation have an increased risk of glare and halos compared to those accepting monofocal IOLs implantation, as well as a slightly lower contrast sensitivity.
5. Method of literature search
The search strategy was (“Lenses, Intraocular”[Mesh] And comparison [ALL]); (“Multifocal Intraocular Lenses”[Mesh] AND “randomized” [ALL]); (“monofocal Intraocular lenses” and “randomized” [ALL]); (“Multifocal Intraocular Lenses”[Mesh] AND “randomly” [ALL]); (“monofocal Intraocular lenses” and “randomly” [ALL]) (“Multifocal Intraocular Lenses”[Mesh] AND “randomization” [ALL]); (“monofocal Intraocular lenses” and “randomization” [ALL]); (“Intraocular lenses” and “trial” [ALL]).
There is no conflict of interest.
Study design and concept of the article were carried out by Xiu Hua Wan, Jinda Wang, Jingshang Zhang, and Kai Cao; database search was carried out by Kai Cao, Shanshan Jin, Simeng Hou, and Guyu Zhu; data extraction was performed by Kai Cao and Shanshan Jin; data analysis was carried out by Kai Cao; manuscript writing was carried out by Kai Cao, David S. Friedman, Xiu Hua Wan, Ying Xiong, Xiaoxia Li, Jing Li, Hailong He, and Lijing Chai; English polishing was done by Mayinuer Yusufu; manuscript revision was done by David S. Friedman.
This study was funded by Beijing New Star of Science and Technology ( H020821380190 and Z131102000413025 ), Fund of Work Committee for Women and Children of China State Department (2014108), National Natural Science Fund ( 30471861 ), Beijing Institute of Ophthalmology Leading Program (201515).