Aims To evaluate the long-term outcomes of small incision lenticule extraction (SMILE) in subjects with myopic astigmatism of ≥2.00 dioptres (D).

Methods Patients who underwent SMILE 4 years prior with astigmatism ≥2.00 D and ≤1.00 D were assigned to the high astigmatic group (HA group) or the low astigmatic group (LA group), respectively. The visual and refractive results as well as corneal wavefront aberrations were measured.

Results The preoperative cylinder was −2.47±0.54 D in 43 eyes in the HA group and −0.55±0.28 D in 31 eyes in the LA group. At 4 years, the residual cylinder was −0.31±0.29 D in the HA group and −0.20±0.28 D in the LA group (p=0.088). An uncorrected distance visual acuity of 20/20 was achieved in 88.4% of eyes in the HA group and 93.5% of eyes in the LA group. The efficacy index was 0.99±0.14 and 1.10±0.21 (p=0.025), and the safety index was 1.11±0.20 and 1.22±0.21 in the HA and LA groups, respectively (p=0.012). Eighty-six per cent and 90.3% of eyes were within ±0.50 D of the attempted cylindrical correction in the HA and LA groups, respectively. Vector analysis showed that the magnitude of error was −0.14±0.28 D and −0.05±0.16 D (p=0.085), the angle of error was −0.13±4.48 degrees and −2.57±29.42 degrees (p=0.592), the correction index was 0.94±0.13 and 0.94±0.35 (p=0.959), the index of success was 0.15±0.14 and 0.46±0.62 (p=0.517), and the flattening index was 0.93±0.13 and 0.71±0.59 (p=0.450) in the HA and LA groups, respectively.

Conclusions This study demonstrates that SMILE is effective and safe for correcting high astigmatism. Vector analysis shows a tendency for the undercorrection of astigmatism in subjects with high astigmatism.