Table 5.
The certainty of evidence.
| Author(s), year | Interventions | Outcomes | Trials (subjects) | Limitations | Inconsistency | Indirectness | Imprecision | Publication bias | Relative effect (95% CI) | Quality |
|---|---|---|---|---|---|---|---|---|---|---|
| Du et al., 2014 [21] | WXKL + CM vs. CM | Effective rate | 13 (1050) | −1① | 0 | 0 | 0 | −1⑤ | OR: 3.40 (2.41, 4.80) | Low |
| WXKL + CM vs. CM | Ventricular rate | 4 (319) | −1① | 0 | 0 | 0 | −1④ | MD: −5.86 (−6.73, −4.99) | Low | |
| WXKL + CM vs. CM | MRSR | 3 (263) | −1① | 0 | 0 | 0 | −1④ | OR: 2.76 (1.29, 5.92) | Low | |
| Huang et al., 2018 [22] | WXKL + metoprolol vs. metoprolol | Effective rate | 10 (651) | −1① | 0 | 0 | −1③ | −1⑤ | OR: 4.06 (2.68, 6.15) | Very low |
| WXKL + metoprolol vs. metoprolol | Ventricular rate | 4 (263) | −1① | 0 | 0 | −1③ | −1④ | MD: −9.86 (−17.88, −1.84) | Very low | |
| WXKL + metoprolol vs. metoprolol | LVEF | 4 (339) | −1① | 0 | 0 | 0 | −1④ | MD: 5.17 (3.08, 7.26) | Low | |
| Li et al., 2018 [32] | WXKL + CM vs. CM | Ventricular rate | 10 (870) | −1① | −1② | 0 | 0 | 0 | MD: −7.14 (−8.42, −5.87) | Low |
| WXKL + CM vs. CM | MRSR | 6 (648) | −1① | 0 | 0 | 0 | −1④ | RR: 1.19 (1.09, 1.29) | Low | |
| WXKL + CM vs. CM | Recurrence rate | 5 (346) | −1① | 0 | 0 | 0 | −1④ | RR: 0.28 (0.13, 0.59) | Low | |
| WXKL + CM vs. CM | LVEF | 4 (402) | −1① | −1② | 0 | −1③ | −1④ | MD: 3.44 (0.87, 6.01) | Very low | |
| WXKL + amiodarone vs. amiodarone | Pmax | 4 (388) | −1① | −1② | 0 | 0 | −1④ | MD: −10.75 (−14.05, −7.45) | Very low | |
| WXKL + CM vs. CM | Pd | 6 (603) | −1① | 0 | 0 | 0 | −1④ | MD: −4.04 (−4.15, −3.93) | Low | |
| Wang et al., 2015 [28] | WXKL + amiodarone vs. amiodarone | Effective rate | 11 (854) | −1① | 0 | 0 | 0 | 0 | RR: 1.22 (1.14, 1.31) | Moderate |
| Wang et al., 2019 [31] | WXKL + CM vs. CM | Effective rate | 22 (2328) | −1① | 0 | 0 | 0 | 0 | OR: 3.37 (2.69, 4.22) | Moderate |
| WXKL + CM vs. CM | MRSR | 7 (856) | −1① | 0 | 0 | 0 | −1④ | OR: 2.32 (1.67, 3.22) | Low | |
| WXKL + CM vs. CM | Pmax | 4 (319) | −1① | 0 | 0 | 0 | −1④ | MD: −9.91 (−12.86, −6.95) | Low | |
| WXKL + CM vs. CM | Pd | 9 (732) | −1① | 0 | 0 | 0 | −1④ | MD: −5.48 (−7.32, −3.64) | Low | |
| Xin et al., 2019 [30] | WXKL + CM vs. CM | Ventricular rate | 9 (632) | −1① | −1② | 0 | 0 | −1④ | MD: −11.66 (−15.79, 7.54) | Very low |
| WXKL + CM vs. CM | Recurrence rate | 2 (184) | −1① | 0 | 0 | 0 | −1④ | RR: 0.34 (0.15, 0.76) | Low | |
| WXKL + CM vs. CM | LVEF | 9 (694) | −1① | −1② | 0 | 0 | −1④ | MD: 6.72 (4.61, 8.84) | Very low | |
| Yang et al., 2019 [27] | WXKL + CM vs. CM | Effective rate | 17 (1735) | −1① | 0 | 0 | 0 | −1⑤ | RR: 1.22 (1.17, 1.27) | Low |
| WXKL + CM vs. CM | Recurrence rate | 4 (353) | −1① | 0 | 0 | 0 | −1④ | RR: 0.18 (0.08, 0.41) | Low | |
| Zhao et al., 2014 [29] | WXKL + metoprolol vs. metoprolol | Effective rate | 4 (269) | −1① | 0 | 0 | 0 | −1④ | RR: 1.34 (1.17, 1.54) | Low |
GRADE: Grading of Recommendations Assessment, Development, and Evaluation; OR: odds ratio; RR: relative risk; MD: mean difference; VL: very low; L: low; H: high; MRSR: maintenance rate of sinus rhythm; CM: conventional medication. ①: the experimental design had a large bias in random, distributive findings or was blind. ②: the confidence intervals overlapped less, the P value for the heterogeneity test was very small, and the I2 was larger. ③: the confidence interval was not narrow enough. ④: fewer studies were included, and there may have been greater publication bias. ⑤: funnel graph asymmetry.