Table 6.
Results of evidence quality.
| Author, year (country) | Outcomes | Studies (participants) | Limitations | Inconsistency | Indirectness | Imprecision | Publication bias | Quality |
|---|---|---|---|---|---|---|---|---|
| A (CLE/C vs placebo) | ||||||||
| Raveendhara R. 2018 (USA) [26] | Pain | 5 (331) | 0 | 0 | 0 | 0 | −1④ | Moderate |
| Function | 3 (232) | 0 | 0 | 0 | 0 | −1④ | Moderate | |
| The use of rescue drugs | 3 (141) | −1① | 0 | 0 | −1③ | −1④ | Very low | |
| Incidence of withdrawal from treatment due to adverse events | 4 (288) | 0 | 0 | 0 | −1③ | −1④ | Low | |
| Adverse events | 3 (247) | 0 | 0 | 0 | −1③ | −1④ | Low | |
| James W. Daily 2016 (South Korea) [34] | VAS | 3 (104) | 0 | 0 | 0 | −1③ | 0 | Moderate |
| WOMAC scale | 3 (122) | 0 | −1② | 0 | −1③ | 0 | Low | |
| An-Fang Hsiao 2021 (China) [28] | VAS | 7 (501) | 0 | −1② | 0 | −1③ | −1④ | Very low |
| Adverse events | 6 (527) | 0 | 0② | 0 | −1③ | 0 | Moderate | |
| Igho J. ONAKPOYA 2017 (UK) [32] | VAS | 5 (366) | −1① | −1② | 0 | 0 | −1④ | Very low |
| WOMAC scale | 3 (167) | −1① | −1② | 0 | −1③ | −1④ | Very low | |
| LPFI | 2 (107) | −1① | 0 | 0 | −1③ | −1④ | Very low | |
| Zhiqiang Wang 2021 (Australia) [33] | Pain | 12 (1,071) | 0 | −1② | 0 | 0 | −1④ | Low |
| Function | 10 (973) | 0 | −1② | 0 | 0 | −1④ | Low | |
| Adverse events | 8 (791) | 0 | 0 | 0 | −1③ | −1④ | Low | |
| The use of rescue drugs | 7 (300) | 0 | 0 | 0 | 0 | −1④ | Moderate | |
| Analgesic discontinuation rate | 4 (154) | 0 | −1② | 0 | −1③ | −1④ | Very low | |
| Wenli Dai 2021 (China) [29] | VAS | 8 (569) | 0 | −1② | 0 | 0 | 0 | Moderate |
| WOMAC scale | 5 (377) | 0 | −1② | 0 | 0 | 0 | Moderate | |
| WOMAC (pain) scale | 5 (377) | 0 | −1② | 0 | 0 | 0 | Moderate | |
| WOMAC (physical) scale | 5 (377) | 0 | −1② | 0 | 0 | 0 | Moderate | |
| WOMAC (stiffness) scale | 5 (377) | 0 | −1② | 0 | 0 | 0 | Moderate | |
| Adverse events | 7 (623) | 0 | 0 | 0 | −1③ | 0 | Moderate | |
| Jian Wu 2019 (China) [30] | WOMAC scale | 3 (146) | 0 | 0 | 0 | −1③ | −1④ | Low |
| VAS | 2 (98) | 0 | 0 | 0 | −1③ | −1④ | Low | |
| Adverse events | 2 (113) | 0 | 0 | 0 | −1③ | −1④ | Low | |
| Liuting Zeng 2021 (China) [31] | VAS | 6 (381) | −1① | 0 | 0 | 0 | 0 | Moderate |
| WOMAC (pain) scale | 4 (315) | −1① | 0 | 0 | 0 | −1⑤ | Low | |
| WOMAC (physical) scale | 4 (315) | −1① | −1② | 0 | 0 | 0 | Low | |
| WOMAC (stiffness) scale | 4 (315) | −1① | 0 | 0 | 0 | 0 | Moderate | |
| Adverse events | 6 (629) | −1① | 0 | 0 | −1③ | 0 | Low | |
| Weiyan Gong 2017 (China) [27] | VAS | 2 (82) | −1① | 0 | 0 | −1③ | −1④ | Very low |
| WOMAC scale | 2 (82) | −1① | 0 | 0 | −1③ | −1④ | Very low | |
| Adverse events | 2 (152) | −1① | 0 | 0 | −1③ | −1④ | Very low | |
| Walking distance | 1 (48) | −1① | −1② | 0 | −1③ | −1④ | Very low | |
| B (CLE/C vs CT) | ||||||||
| Raveendhara R. 2018 (USA) [26] | Pain (vs NSAIDs) | 2 (422) | 0 | 0 | 0 | −1③ | −1④ | Low |
| Function (vs NSAIDs) | 1 (331) | 0 | −1② | 0 | −1③ | −1④ | Very low | |
| The use of rescue drugs (vs NSAIDs) | 2 (422) | 0 | 0 | 0 | −1③ | −1④ | Low | |
| Incidence of withdrawal from treatment due to adverse events (vs NSAIDs) | 2 (474) | 0 | 0 | 0 | 0 | −1④ | Moderate | |
| Adverse events (vs NSAIDs) | 2 (467) | 0 | 0 | 0 | 0 | −1④ | Moderate | |
| James W. Daily 2016 (South Korea) [34] | WOMAC scale (vs painkillers) | 5 (625) | 0 | −1② | 0 | −1③ | 0 | Low |
| An-Fang Hsiao 2021 (China) [28] | VAS (vs NSAIDs) | 2 (256) | 0 | 0 | 0 | 0 | −1⑤ | Moderate |
| Adverse events (vs NSAIDs) | 3 (623) | 0 | 0 | 0 | −1③ | 0 | Moderate | |
| Igho J. ONAKPOYA 2017 (UK) [32] | WOMAC scale (vs NSAIDs) | 1 (331) | 0 | −1② | 0 | −1③ | −1④ | Very Low |
| Zhiqiang Wang 2021 (Australia) [33] | Pain (vs NSAIDs) | 5 (648) | 0 | 0 | 0 | −1③ | −1④ | Low |
| Function vs NSAIDs) | 3 (477) | 0 | 0 | 0 | −1③ | −1④ | Low | |
| Adverse events (vs NSAIDs) | 3 (571) | 0 | 0 | 0 | 0 | −1④ | Moderate | |
| The use of rescue drugs (vs NSAIDs) | 2 (443) | 0 | 0 | 0 | −1③ | −1④ | Low | |
| Jian Wu 2019 (China) [30] | WOMAC scale (vs NSAIDs) | 1 (331) | 0 | −1② | 0 | −1③ | −1④ | Very low |
| Adverse events (vs NSAIDs) | 2 (159) | 0 | 0 | 0 | −1③ | −1④ | Low | |
| Liuting Zeng 2021 (China) [31] | VAS (vs NSAIDs) | 2 (230) | −1① | 0 | 0 | −1③ | 0 | Low |
| WOMAC (pain) scale (vs NSAIDs) | 1 (331) | −1① | −1② | 0 | −1③ | −1④ | Very low | |
| WOMAC (physical) scale (vs NSAIDs) | 1 (331) | −1① | −1② | 0 | −1③ | −1④ | Very low | |
| WOMAC (stiffness) scale (vs NSAIDs) | 1 (331) | −1① | 0 | 0 | −1③ | −1④ | Very low | |
| Adverse events (vs NSAIDs) | 3 (561) | −1① | 0 | 0 | 0 | 0 | Moderate | |
| Weiyan Gong 2017 (China) [27] | VAS (vs NSAIDs) | 1 (112) | −1① | −1② | 0 | −1③ | −1④ | Very low |
| WOMAC scale (vs NSAIDs) | 1 (331) | −1① | −1② | 0 | −1③ | −1④ | Very low | |
| Walking distance (vs NSAIDs) | 2 (360) | −1① | 0 | 0 | −1③ | −1④ | Very low | |
| Adverse events (vs NSAIDs) | 3 (491) | −1① | −1② | 0 | 0 | −1④ | Very low | |
Note. ①The included studies have a large bias in methodology such as randomization, allocation concealment, and blinding. ②The confidence interval overlaps less, or the I2 value of the combined results was larger. ③The sample size from the included studies does not meet the optimal sample size or the 95% confidence interval crosses the invalid line. ④The funnel chart is asymmetry. ⑤Fewer studies were included, and their results were all positive, which may result in a large publication bias.