Table 4. Subgroup analysis estimating the effect of trial design on percentage IOP decline.
| Subgroup | No. of studies | WMD (fixed) (95%CI) | Heterogeneity |
Overall effect |
||||||
| Chi2 | P | I2 (%) | Z | P | ||||||
| Deep sclerectomy vs combined deep sclerectomy with phacoemulsification | ||||||||||
| All trials | 5 | 2.85 (-0.69, 6.39) | 7.35 | 0.12 | 46 | 1.58 | 0.12 | |||
| RCTs | 3 | -2.25 (-6.44, 1.95) | 2.30 | 0.32 | 13 | 1.05 | 0.29 | |||
| Pro | 1 | 2.97 (-6.35, 12.29) | - | - | - | 0.62 | 0.53 | |||
| Retro | 1 | -11.80 (-21.21, -2.39) | - | - | - | 2.46 | 0.01 | |||
| Canaloplasty vs phaco-canaloplasty | ||||||||||
| All trials | 5 | -3.78 (-8.38,0.81) | 5.66 | 0.23 | 29 | 1.61 | 0.11 | |||
| RCTs | 1 | -9.78 (-20.99,1.43) | - | - | - | 1.71 | 0.09 | |||
| Pro | 2 | 2.82 (-4.55, 10.19) | 0.19 | 0.67 | 0 | 0.75 | 0.45 | |||
| Retro | 2 | -7.31 (-14.23, -0.40) | 0.29 | 0.59 | 0 | 2.07 | 0.04 | |||
| Trabeculotomy vs combined trabeculotomy with phacoemulsification | ||||||||||
| All trials | 6 | 12.65 (10.56, 14.74) | 53.56 | <0.00001 | 91 | 11.87 | <0.00001 | |||
| RCTs | 1 | 10.74 (5.87, 15.61) | - | - | - | 4.33 | <0.00001 | |||
| Pro | 1 | 12.51 (8.73, 16.29) | - | - | - | 6.48 | <0.00001 | |||
| Retro | 4 | 13.42 (10.50, 16.35) | 52.70 | <0.00001 | 94 | 9 | <0.00001 | |||