Twelve RCTs (727 eyes) were included in the review. Sample sizes ranged from 22 to 180 eyes. Quality assessment scores ranged from 4 to 9 out of a possible maximum score of 10 points. Five trials scored 6 or more out of 10 and were considered to be of adequate quality. Most trials did not report the method of allocation concealment, methods of blinding, or an acceptable rate of drop-outs. None of the trials an intention-to-treat analysis. The duration of follow-up ranged from four weeks to two years; six RCTs had follow-up of at least one year.
Accommodating intraocular lens significantly improved distance-corrected near visual acuity compared with monofocal intraocular lens (SMD -1.36, 95% CI -2.22 to -0.49; 10 RCTs). However, there was significant statistical heterogeneity (I2=94%), which could not be explained by any population or methodology characteristic. When the six homogeneous RCTs were pooled, there was no significant difference in distance-corrected near visual acuity between accommodating intraocular lens and monofocal intraocular lens.
Accommodating intraocular lens were associated with significantly greater anterior lens shift than monofocal intraocular lens (WMD -0.36, 95% CI -0.47 to -0.24; four RCTs). There was also significant statistical heterogeneity (I2=58%), which could not be explained by any population or methodology characteristic.
There were no statistically significant differences in glare (two RCTs) or contrast sensitivity (two RCTs) between accommodating intraocular lens and monofocal intraocular lens. Two trials reported better reading speed in the accommodating intraocular lens group. One trial reported better spectacle independence in the accommodating intraocular lens group.
Regarding adverse events, five RCTs reported rates of posterior capsule opacification, with four of the RCTs reporting increased rates in the accommodating intraocular lens group several months after surgery. One RCT reported a patient suffering anterior chamber haemorrhage. One RCT reported a patient suffering posterior capsule shrinkage.
The funnel plot did not indicate that publication bias was present.