Risk of bias assessment based on Downs and Black’s criteria16
| Author | Year | Reporting | Internal validity | Confounding and selection bias | External validity | Power | Total score |
| Banerjee et al*9 | 2020 | 18 | |||||
| Campbell et al23 | 2016 | 12 | |||||
| Chigbu et al*18 | 2020 | 17 | |||||
| Duan et al*28 | 2021 | 21 | |||||
| Eakin et al24 | 2018 | 17 | |||||
| Fokom Domgue et al25 | 2020 | 14 | |||||
| Joshi et al26 | 2015 | 20 | |||||
| Joshi et al*10 | 2013 | 17 | |||||
| Mungo et al*19 | 2020 | 19 | |||||
| Naud et al*22 | 2016 | 14 | |||||
| Pinder et al*11 | 2020 | 24 | |||||
| Sandoval et al*20 | 2019 | 20 | |||||
| Viviano et al*21 | 2017 | 21 | |||||
| Zhao et al*12 | 2021 | 19 | |||||
| Zhao et al27 | 2020 | 20 |
Subsections were scored green if over 80% of indicators met, orange if 40%–79% of indicators met and red if <40% of indicators met.
*Included in the meta-synthesis.