Newcastle-Ottawa Scale risk of bias scores
| Reference | Selection | Comparability | Outcome | Total |
| Markou et al 12 | 3 | 2 | 3 | 8 |
| Nery et al 13 | 3 | 0 | 2 | 5 |
| Markou et al 14 | 3 | 2 | 2 | 7 |
| Martini and Barbisan15 | 3 | 0 | 2 | 5 |
| Nery and Barbisan16 | 3 | 2 | 2 | 7 |
| Rengo et al 17 | 4 | 2 | 3 | 9 |
| Giaccardi et al 18 | 3 | 2 | 2 | 7 |
| Cacciatore et al 19 | 3 | 2 | 2 | 7 |
| Noyez et al 20 | 3 | 2 | 3 | 8 |
| Min et al 21 | 4 | 2 | 1 | 7 |
| van Laar et al 22 | 3 | 0 | 3 | 6 |
| Average scores±SD | 3.18±0.40 | 1.45±0.93 | 2.27±0.65 | 6.91±1.22 |
Maximum scores are 4, 2 and 3 for selection, comparability and outcome, respectively. Maximum total score is 9. A lower score within each category and for a total score indicates a higher risk of bias.