Table 2.
Reporting on quality of included interventions (25 interventions reported in 45 papers included in the review)
| Selection bias (random sequence generation) | Selection bias (allocation concealment) | Performance bias | Detection bias | Attrition bias (incomplete outcome data) | Reporting bias (selective reporting) | Other sources of bias | % risk of bias | Comments | |
|---|---|---|---|---|---|---|---|---|---|
| Abdool Karim et al. (2015) [68]; Humphries et al., (2017) [69], South Africa | Low risk | Low risk | Low risk | Low risk | Unclear | Low risk | Low risk | 14.3% | Study design: quantitative (comparison of treatment and control groups) |
| Adoho et al. (2014) [70], Liberia | Low risk | Unclear | Low risk | Low risk | Low risk | Low risk | Unclear | 28.6% | Study design: quantitative (comparison of two treatment groups to a control group) |
| Austrian and Muthengi (2014) [71]; Muthengi (2014) [72], Uganda | High risk | Unclear | Low risk | Low risk | Low risk | Low risk | Unclear | 42.9% | Study design: mixed methods (comparison of two treatment groups to a control group) |
| Baird et al. (2012) [73]; Baird et al. (2013) [74], Malawi | Low risk | Low risk | Low risk | Low risk | Unclear | Low risk | Low risk | 14.3% | Study design: mixed methods (pre- and post-test comparison for intervention and control groups) |
| Bandiera et al. (2012) [75]; Bandiera et al. (2018) [76], Uganda | Low risk | Unclear | Low risk | Low risk | Low risk | Low risk | Low risk | 14.3% | Study design: quantitative (pre- and post-test comparison for intervention and control groups) |
| Bazika (2007) [77], Congo | Unclear | Unclear | Low risk | Low risk | Unclear | Unclear | Unclear | 71.4% | Study design: quantitative (pre- and post-test assessment of intervention participants) |
| Cho et al. (2018) [78], Kenya | Low risk | Unclear | Low risk | Low risk | Low risk | Low risk | Low risk | 14.3% |
Study design: quantitative (comparison between intervention and control groups) Longitudinal study with annual repeated measures over 4 years |
| Cluver et al. (2016) [79], South Africa | Low risk | Unclear | Low risk | Low risk | Low risk | Low risk | Low risk | 14.3% |
Study design: quantitative (comparison between cash alone and integrated cash plus care intervention for HIV-risk reduction) Prospective longitudinal study |
| de Walque et al. (2012) [80]; de Walque et al. (2014) [81], Tanzania | Low risk | High risk | Low risk | Low risk | Low risk | Low risk | Low risk | 14.3% | Study design: quantitative (pre- and post-test comparison for intervention and control groups) |
| Dunbar et al. (2010) [82]; Dunbar et al. (2014) [83], Zimbabwe | Low risk | High risk | Low risk | Low risk | Low risk | Low risk | Low risk | 14.3% | Study design: quantitative (pre- and post-test comparison for intervention and control groups) |
| Erulkar and Chong (2005) [84]; Hall et al. (2006) [85], Kenya | High risk | High risk | Low risk | Low risk | Low risk | Low risk | Unclear | 42.9% |
Study design: quantitative (pre- and post-assessment of intervention participants) Longitudinal study |
| Goodman et al. (2014) [86], Kenya | Low risk | Unclear | Low risk | Low risk | Unclear | Low risk | Low risk | 28.6% | Study design: quantitative (cross-sectional comparison among 3 cohorts) |
| Hallfors et al. (2011) [22]; Hallfors et al. (2015) [23]; Luseno et al. (2015) [87], Zimbabwe | Low risk | Unclear | Low risk | Low risk | Low risk | Low risk | Low risk | 14.3% |
Study design: quantitative (comparison between intervention and control groups) Longitudinal study with annual repeated measures over 3 years |
| Handa et al. (2014) [88]; Rosenberg et al. (2014) [89], Kenya | Low risk | Unclear | Low risk | Low risk | Low risk | Low risk | Low risk | 14.3% |
Study design: quantitative (comparison between intervention and control groups) Longitudinal study with repeated measures |
| Jewkes et al. (2014) [90],South Africa | High risk | High risk | Low risk | Low risk | Unclear | Low risk | Unclear | 57.1% | Study design: mixed methods (interviews and quantitative time series design for pre- and post-intervention assessment) |
| Khoza et al. (2018) [91], South Africa | Low risk | Unclear | Unclear | Unclear | Unclear | Low risk | Unclear | 71.4% | Study design: qualitative (using interviews) |
| Kim et al. (2009) [92]; Pronyk et al. (2006) [93]; Kim et al. (2007) [94]; Pronyk et al. (2008) [95], South Africa | Low risk | Low risk | Low risk | Low risk | Low risk | Low risk | Low risk | 0.0% | Study design: mixed methods pre- and post-test comparison of intervention and control group) |
| Kohler and Thornton (2012) [96], Malawi | Low risk | Unclear | Low risk | Low risk | Low risk | Low risk | Low risk | 14.3% | Study design: quantitative (pre- and post-comparison between intervention and control groups) (longitudinal study) |
| Nyqvist et al. (2015) [97]; Nyqvist et al.(2018) [98], Lesotho | Low risk | Unclear | Low risk | Low risk | Low risk | Low risk | Low risk | 14.3% | Study design: quantitative (pre- and post-comparison of intervention and control groups) |
| O’Neill Berry et al., (2013) [99], Lesotho | Unclear | High risk | High risk | Low risk | High risk | Low risk | High risk | 71.4% | Study design: mixed method (pre- and post-test comparison of intervention and control group, as well as follow-up observations of intervention group) |
| Pettifor et al. (2016) [18]; Pettifor et al. (2016b) [100], South Africa | Low risk | Low risk | Low risk | Low risk | Low risk | Low risk | Low risk | 0.0% | Study design: quantitative (pre- and post-comparison of intervention and control groups) |
| Rotheram- Borus et al. (2012) [101], Uganda | Unclear | High risk | Low risk | Low risk | Low risk | Low risk | Unclear | 42.9% |
Study design: quantitative Had a delayed intervention group. Pre- and post-test comparison between immediate intervention and delayed intervention group |
| Ssewamala et al., (2009) [39]; Ssewamala et al. (2010) [102]; Ssewamala et al. (2010b) [26]; Ismayilova et al. (2012), Uganda | Low risk | Unclear | Low risk | Low risk | Low risk | Low risk | Low risk | 14.3% | Study design: quantitative (pre- and post-comparison of intervention and control groups) |
| Stark et al. (2018) [41]; Falb et al. (2016) [104], Ethiopia | Low risk | Unclear | Low risk | Unclear | High risk | Low risk | Unclear | 57.1% |
Study design: quantitative (pre- and posttest comparison for intervention and control groups). Reported null findings that the intervention did not seem to keep the participants in school, nor influence out-of-school girls to return to school |
| Visser et al. (2015) [105]; Visser et al. (2018) [106], South Africa | High risk | Unclear | Low risk | High risk | High risk | Low risk | Unclear | 71.4% | Study design: mixed method (quasi-experimental post-intervention assessment between intervention and control group and focus group discussions). Utilized focus group discussions to generate information on strategies to sustain ISIBINDI intervention |