Table 2 Mean differences in the main outcomes and adverse events in randomized controlled trials after corneal collagen cross-linking (epi-on vs. epi-off)

Kmax (D)

 Lombardo et al. 2019^a [48]

− 1.5 ± 4.9

–

12

− 1.0 ± 3.7

–

22

 Iqbal et al. 2019^b [50]

− 1.2 ± 1.0

–

91

0.9 ± 1.1

–

88

 Rossi et al. 2015^b [60]

− 1.1 ± 2.1

–

10

− 1.1 ± 1.0

–

10

 Soeters et al. 2015^b [61]

− 1.5 ± 2.0

–

26

0.3 ± 1.8

–

35

UDVA (logMAR)

 Lombardo et al. 2019 [48]

− 0.3 ± 0.3

–

12

− 0.3 ± 0.4

–

22

 Iqbal et al. 2019 [50]

− 0.3 ± 0.1

–

91

0.2 ± 0.1

–

88

 Rossi et al. 2015 [60]

− 0.1 ± 0.1

–

10

− 0.1 ± 0.1

–

10

 Soeters et al. 2015 [61]

− 0.1 ± 0.4

–

26

− 0.1 ± 0.4

–

35

CDVA (logMAR)

 Lombardo et al. 2019 [48]

− 0.0 ± 0.1

–

12

− 0.1 ± 0.1

–

22

 Al Zubi et al. 2019 [49]

− 0.1 ± 0.1

–

40

− 0.1 ± 0.1

–

40

 Iqbal et al. 2019 [50]

− 0.2 ± 0.2

–

91

0.1 ± 0.1

–

88

 Rush et al. 2017 [55]

− 0.2

–

64

− 0.1

–

80

 Bikbova et al. 2016 [57]

− 0.0 ± 0.3

–

73

− 0.1 ± 0.5

–

76

 Rossi et al. 2015 [60]

− 0.1 ± 0.0

–

10

− 0.2 ± 0.1

–

10

 Soeters et al. 2015 [61]

− 0.1 ± 0.2

–

26

− 0.1 ± 0.21

–

35

Higher-order aberrations (µm)

 Lombardo et al. 2019 [48]

− 0.1 ± 0.9

–

12

1.0 ± 1.1

–

22

 Godefrooij et al. 2017 [54]

− 0.2 ± 0.4

–

26

0.0 ± 0.5

–

35

Comatic aberrations (μm)

 Rossi et al. 2015 [60]

− 0.4 ± 1.2

–

10

− 0.7 ± 1.3

–

10

Demarcation Line (depth in µm at 1 month)

 Bikbova et al. 2016 [57]

287.0 ± 15.0

–

73

179.0 ± 18.0

–

76

 Soeters et al. 2015 [61]

266.0 ± 64.0

–

26

0.0

–

35

Central corneal thickness (µm)

 Lombardo et al. 2019 [48]

5.0 ± 21.0

–

12

7.0 ± 27.0

–

22

 Al Zubi et al. 2019 [49]

5.4 ± 12.5

–

40

6.7 ± 15.5

–

40

 Bikbova et al. 2016 [57]

− 13.0 ± 37.2

–

73

− 6.7 ± 38.6

–

76

 Rossi et al. 2015 [60]

1.8 ± 14.6

–

10

− 2.7 ± 37.3

–

10

Corneal thinnest point (µm)

 Iqbal et al. 2019 [50]

− 8.9 ± 14.9

–

91

− 6.7 ± 9.0

–

88

 Soeters et al. 2015 [61]

− 4.0 ± 8.0

–

26

0.0 ± 12.0

–

35

ECD (cells/mm²)

 Lombardo et al. 2019 [48]

− 30.0 ± 368.0

–

12

− 33.0 ± 309.0

–

22

 Bikbova et al. 2016 [57]

20.0 ± 91.0

–

73

17.0 ± 69.0

–

76

 Rossi et al. 2015 [60]

− 31.4 ± 66.6

–

10

− 53.0 ± 202.4

–

10

 Soeters et al. 2015 [61]

− 59.0 ± 284.2

–

26

11.0 ± 338.9

–

35

KC progression

 Lombardo et al. 2019 [48]

–

0 (0.0%)

12

–

2 (9.1%)

22

 Iqbal et al. 2019 [50]

–

0 (0.0%)

91

–

25 (28.4%)

88

 Bikbova et al. 2016 [57]

–

0 (0.0%)

73

–

1 (1.3%)

76

 Al Fayez et al. 2015 [59]

–

0 (0.0%)

36

–

19 (55.9%)

34

 Soeters et al. 2015 [61]

–

0 (0.0%)

26

–

8 (22.9%)

35

Delay in epithelial healing

 Iqbal et al. 2019 [50]

–

17 (18.7%)

91

–

0 (0.0%)

88

 Rush et al. 2017 [55]

–

1 (1.6%)

64

–

0 (0.0%)

80

 Bikbova et al. 2016 [57]

–

4 (5.5%)

73

–

0 (0.0%)

76

 Soeters et al. 2015 [61]

–

2 (7.7%)

26

–

0 (0.0%)

35

Persistent stromal haze

 Lombardo et al. 2019 [48]

–

2 (16.7%)

12

–

0 (0.0%)

22

 Al Zubi et al. 2019 [49]

–

4 (10.0%)

40

–

0 (0.0%)

40

 Iqbal et al. 2019 [50]

–

2 (2.2%)

91

–

0(0.0%)

88

 Bikbova et al. 2016 [57]

–

4 (5.5%)

73

–

0 (0.0%)

76

 Soeters et al. 2015 [61]

–

1 (3.8%)

26

–

0 (0.0%)

35

Sterile infiltrates

 Soeters et al. 2015 [61]

–

1 (3.8%)

26

–

0 (0.0%)

35

Infection

 Rush et al. 2017 [55]

–

1 (1.6%)

64

–

0 (0.0%)

80

 Soeters et al. 2015 [61]

–

1 (3.8%)

26

–

0 (0.0%)

35