AIM: To compare the efficacy of and without small incision lenticule extraction(SMILE)with cyclotorsion compensation for astigmatism correction.

METHODS: PubMed, Web of Science, EMBASE, Cochrane and CNKI, VIP, CBM, and Wan Fang Data were searched for clinically controlled studies from January 2010 to August 2022, including an experimental group with cyclotorsion compensation during SMILE and a control group without cyclotorsion compensation during SMILE. After literature screening, quality evaluation, and data extraction by two researchers independently, the Meta-analysis of uncorrected distance visual acuity(UDVA), residual astigmatism, vector analysis indicators for measuring the astigmatism correction including absolute value of angle of error(|AE|)and magnitude error(ME), and post-operative total higher order aberrations, spherical aberration and coma was carried out with Stata 16.0 software.

RESULTS: Seven studies with a total of 846 eyes(442 in the experimental group, 404 in the control group)were finally included. The Meta-analysis showed that there were significant differences in the percentage of eyes with residual astigmatism ≥1.00D(OR=0.17, 95%CI: 0.06～0.49, P<0.01), |AE|(WMD=-1.56, 95%CI: -2.68～-0.45, P<0.01), the coma(WMD=0.06, 95%CI: -0.08～-0.04, P<0.01), and the total higher order aberrations(WMD=-0.04, 95%CI: -0.06～-0.02, P<0.01). However, there were no differences in the postoperative UDVA(WMD=0.00, 95%CI: -0.02～0.01, P=0.54), residual astigmatism(WMD=0.08, 95%CI: -0.02～0.18, P=0.10), ME(WMD=-0.01, 95%CI: -0.14～0.12, P=0.85), and the spherical aberration(WMD=0.03, 95%CI: -0.07～0.13, P=0.52).

CONCLUSION: Cyclotorsion compensation in SMILE can reduce the angular error caused by eye rotation during astigmatism correction. It also decreases postoperative residual astigmatism. Overall, the SMILE with cyclotorsion compensation is superior in clinical efficacy of the precise correction of astigmatism.