Table 2.
Subgroups and sensitivity analysis based on ORs of SYMPTOMS outcomes.
| Group | No. of studies | No. of participants | OR | Wilcoxon test | 95% CI | Z | P (effect) | I 2 | χ 2 | P (het) | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| M score | ≤3 | 5 | 432 | 4.76 | W = 7 | 2.85, 7.95 | 5.96 | <0.00001 | 0% | 2.21 | 0.70 |
| >3 | 2 | 118 | 2.24 | P = 0.4386 | 0.41, 12.17 | 0.94 | 0.35 | 67% | 3.06 | 0.08 | |
| Sample size | <78 | 4 | 214 | 2.49 | W = 4 | 1.09, 5.68 | 2.17 | 0.03 | 3% | 3.09 | 0.38 |
| ≥78 | 3 | 336 | 5.22 | P = 0.4795 | 3.05, 8.93 | 6.03 | <0.00001 | 0% | 1.07 | 0.59 | |
| No. of authors | 1 | 4 | 396 | 4.85 | W = 6 | 2.90, 8.11 | 6.01 | <0.00001 | 0% | 1.94 | 0.58 |
| >1 | 3 | 154 | 2.61 | P = 1 | 0.82, 8.36 | 1.62 | 0.11 | 35% | 3.06 | 0.22 | |
| Publication year | <2005 | 4 | 346 | 3.84 | W = 7 | 1.65, 8.93 | 3.12 | 0.002 | 47% | 5.71 | 0.13 |
| ≥2005 | 3 | 204 | 3.82 | P = 0.7237 | 1.78, 8.21 | 3.44 | 0.0006 | 0% | 0.58 | 0.75 | |
| Trial date | Reported | 5 | 432 | 4.76 | W = 7 | 2.85, 7.95 | 5.96 | <0.000 01 | 0% | 2.21 | 0.70 |
| Not reported | 2 | 118 | 2.24 | P = 0.4386 | 0.41, 12.17 | 0.94 | 0.35 | 67% | 3.06 | 0.08 | |
| Comparable | Reported | 4 | 238 | 2.83 | W = 2 | 1.36, 5.89 | 2.78 | 0.005 | 10% | 3.35 | 0.34 |
| Not reported | 3 | 312 | 5.45 | P = 0.1573 | 3.02, 9.83 | 5.64 | <0.00001 | 0% | 1.12 | 0.57 | |
| AEs | Reported | 4 | 310 | 3.30 | W = 6 | 1.58, 6.88 | 3.18 | 0.0001 | 21% | 3.78 | 0.29 |
| Not reported | 3 | 240 | 5.17 | P = 1 | 2.74, 9.78 | 5.06 | <0.00001 | 0% | 1.71 | 0.43 | |
| Improvement | >50% | 7 | 550 | 4.11 | W = 7 | 2.57, 6.55 | 5.93 | <0.00001 | 6% | 6.36 | 0.38 |
| >80% | 5 | 370 | 1.83 | P = 0.2506 | 1.20, 2.82 | 2.77 | 0.006 | 0% | 1.10 | 0.89 |
CI is confidence interval. Mann-Whitney-Wilcoxon test was performed on ORs of SYMPTOMS. Z and P (effect) evaluated the statistics of overall effect; I 2, χ 2, and P (het) were computed to assess heterogeneity.