Report / Paper metadata
Population-level effects of a community-based program on intimate partner violence: A stepped-wedge trial in South African informal settlements.
A community programme studied in eighteen South African informal settlements was associated with less reported intimate partner violence among women. Outcomes differed by type of violence and by respondent group; the findings should not be read as a uniform reduction in every form of harm.
What programme was studied?
The Asibonisane Community Responses programme combined a ten-session Stepping Stones curriculum with community-wide education about HIV and intimate partner violence. Researchers evaluated it in eighteen settlements in KwaZulu-Natal using a stepped-wedge cluster-randomised trial. This design studies a programme introduced to groups at different stages, rather than assigning each person separately.
Who contributed to the analysis?
The study used four survey rounds, sampling men aged eighteen to thirty-five and women aged eighteen to twenty-four. The analysis included 549 women and 438 men in primary partnerships who had been interviewed at least twice. These age groups and inclusion rules limit which populations the findings describe.
What did participants report?
The intention-to-treat analysis found lower reported emotional, physical and sexual violence among women. The adjusted odds ratios were 0.42, 0.70 and 0.27, respectively. Men reported lower emotional and physical perpetration, but not lower sexual perpetration. An odds ratio is not a direct percentage-point reduction in the number of people affected.
What needs a cautious reading?
The abstract concerns self-reported experiences and behaviour, not independently observed incidents. It does not provide all the absolute follow-up rates or uncertainty ranges needed to judge the size of the change. Full methods are needed to assess survey follow-up, time trends and how the clustered design was analysed. These results support further evaluation of a specific programme; they do not establish that any community education programme will have the same effect.
- Journal
- Global public health
- Kind
- Paper metadata
- Identifier
- PMID 42826221