Domestic abuse and sexual violence are highly prevalent and have profound consequences for mental health and wellbeing. Yet despite substantial investment in services and interventions, our understanding of what improves safety, wellbeing, and ultimately prevents abuse remains fragmented and difficult to interpret.
Part of the challenge is methodological. Interventions in this field are often highly complex, while evaluation methods tend to be comparatively simple. Services are multifaceted, relational, and frequently tailored to individuals’ needs in real time. Outcomes are equally diverse, encompassing safety, wellbeing, empowerment, attitudes, and behaviour change, with little consistency in how these outcomes are measured across studies.
The systematic review and meta-analysis by Carlisle et al. (2025) tackles this challenge directly. Focusing on UK-based services and interventions, the authors attempt something ambitious: synthesising evidence across the full spectrum of responses to domestic abuse and sexual violence, from advocacy and outreach services to psychological support and perpetrator programmes. By bringing together this diverse evidence base, the review offers an opportunity to better understand what works, for whom, and where important gaps remain.
Domestic abuse and sexual violence are common, but evidence about what works remains surprisingly difficult to navigate.
Methods
The authors drew on both peer-reviewed and grey literature, which is important in a field where much evaluation sits outside academic journals. The authors included randomised trials, non-randomised comparative studies, pre–post evaluations, and service-level reports. Appropriate quality appraisal tools were used for each method.
The review included UK studies evaluating secondary or tertiary prevention interventions for adults affected by domestic or sexual violence and abuse. To assess effectiveness, studies had to report outcomes over time or compare outcomes between groups.
Included interventions were grouped into four broad categories:
- Advocacy (e.g. independent domestic violence advocates, IDVAs)
- Outreach services
- Psychological support
- Perpetrator programmes
Results
A total of 28 studies were included in the review, representing 42,850 participants who had experienced domestic violence and sexual abuse (DVSA) and a further 246 participants who had been perpetrators of DVSA. The majority of included studies had a high risk of bias, owing to issues such as a lack of control groups, non-randomised designs and inconsistent outcome measurement.
Advocacy and outreach: encouraging signals on safety
The most striking quantitative findings come from advocacy and outreach services. Meta-analysis suggested that 58.7% of advocacy service users, and 46.2% for outreach services, reported cessation of abuse at case closure.
At face value, these are impressive figures. They suggest that nearly half of those engaging with services experience an end to abuse during the intervention period. But interpretation requires caution. These figures come largely from uncontrolled pre–post studies. As the authors explain, participants enter services because abuse is ongoing, so any cessation at follow-up looks like improvement. But without a comparison group, we cannot know:
- What would have happened without the intervention?
- Whether reporting bias influences outcomes?
- Whether “case closure” reflects genuine safety or disengagement?
Still, it would be overly dismissive to ignore these findings. Advocacy services, particularly IDVAs, have long been considered a cornerstone of the UK response to domestic abuse.
Psychological support: improvements in self-esteem
For psychological interventions, meta-analysis was not possible due to heterogeneity and limited data. Instead, the authors used “vote counting” based on direction of effect. They found that most studies reported improvements in self-esteem. This aligns with theoretical expectations. Psychological support aims not just to reduce symptoms but to restore agency, confidence, and identity. These factors are closely tied to recovery from abuse. However, self-esteem is a distal outcome. It matters, but it is not the same as safety. One could plausibly see improvements in self-esteem without reductions in abuse, or vice versa. This raises a key question: are we measuring what matters most to survivors?
Perpetrator programmes: tentative shifts in attitudes
Evidence for perpetrator programmes was similarly limited and heterogeneous. Findings suggested positive changes in attitudes towards sexual offending and some improvement on measures like the Balanced Inventory of Desirable Responding. Again, these are indirect indicators. Changing attitudes is important, but it is not the same as reducing violence. This area of the literature suffered from small sample sizes (only 246 participants across studies), a measured high risk of bias, and a heavy reliance on self-report. So, the field continues to struggle with a central tension: how to rigorously evaluate interventions aimed at those who use violence, without relying solely on their own accounts.
This review found that advocacy and outreach services led to cessation of domestic and sexual violence abuse at case closure in around half of recipients.
Conclusion
Domestic and sexual violence and abuse can have devastating and long-lasting consequences, making effective support services essential. This review provides some encouraging signals. Advocacy and outreach services were associated with abuse cessation for many service users, psychological interventions appeared to improve self-esteem, and perpetrator programmes showed tentative shifts in attitudes linked to offending behaviour.
Yet perhaps the most important finding is not what we know, but what we still do not know. Much of the evidence base rests on uncontrolled studies, small samples, and inconsistent outcome measurement. As a result, interventions that are widely delivered and highly valued remain surprisingly difficult to evaluate with confidence.
This review provides some encouraging signals that support services for people who have experienced domestic violence and sexual abuse are effective.
Strengths and limitations
Perhaps the most thought-provoking aspect of this review is its engagement with what might be called the “outcome problem”. The authors’ earlier scoping review identified 426 outcomes across 80 studies, with fewer than half used in more than one evaluation. This level of fragmentation makes evidence synthesis extraordinarily difficult and reflects a broader challenge in the field: different services pursue different goals, funders require different metrics, and survivors may prioritise outcomes that researchers rarely measure.
A major strength of the review is its attempt to navigate this complexity through stakeholder co-production. Rather than imposing a single outcome across all interventions, the authors selected the most commonly reported outcome for each intervention type. This pragmatic approach acknowledges that advocacy services, psychological interventions and perpetrator programmes are trying to achieve different things. At the same time, it highlights how far the field remains from consensus on what meaningful success looks like.
The review has several additional strengths. The inclusion of grey literature captures evidence that would otherwise be missed, particularly given that many domestic abuse and sexual violence services are evaluated outside academic settings. The breadth of interventions considered also provides a more system-level perspective than previous reviews focused on a single service type. Finally, the authors are commendably transparent about the limitations of both the included studies and their own synthesis.
Those limitations are important. Most included studies were at high risk of bias, often relying on uncontrolled pre-post designs that cannot establish whether observed improvements were caused by the intervention itself. Interventions varied substantially in their intensity, duration, target populations and theoretical underpinnings, making direct comparisons difficult. Furthermore, while focusing on commonly reported outcomes enabled synthesis, it may have overlooked outcomes that matter deeply to survivors but are measured less frequently.
Overall, the review offers encouraging evidence that advocacy, outreach, psychological support and perpetrator programmes may be beneficial. However, the certainty of that evidence remains limited. The findings therefore support cautious optimism rather than firm conclusions, while reinforcing the need for more robust evaluation and greater agreement about what outcomes matter most.
The authors’ identified 426 outcomes across 80 studies exploring services for domestic violence and sexual abuse, with fewer than half used in more than one evaluation. This level of fragmentation makes evidence synthesis extraordinarily difficult.
Implications for practice
Implications for policy and practice
- In a field where harm is immediate and severe, waiting for gold-standard trials may not be either realistic or ethical. Services must continue to operate and be funded sustainably based on the best available evidence.
- There must be investment in improving evidence quality, comprising routine outcome measurement, standardised tools, and increased use of data linkage (e.g. health, policing/justice, social care).
- The development of a core outcome set for domestic abuse and sexual violence services is perhaps the most urgent priority. Without shared metrics, cumulative knowledge remains out of reach. It may be important to broaden outcomes under study beyond abuse cessation, as it risks missing the importance of increased safety planning, reduced severity of abuse, improved autonomy experienced by victim/survivors of domestic abuse and sexual violence and improved mental health. These are meaningful outcomes, even if abuse does not immediately stop.
Implications for research
- The authors recommend that future studies should move beyond pre–post designs, instead incorporating comparison groups, quasi-experimental methods, and longitudinal follow-up data collection.
- Where possible studies should incorporate qualitative methods, towards understanding how and why interventions work.
- Addressing perpetrator evidence gaps remains one of the key areas for improving the field.
- Outcome selection and interpretation should be grounded in what survivors themselves identify as meaningful change.
Final thoughts
The evaluation of domestic abuse and sexual violence interventions is methodologically challenging, ethically constrained, and complex. The paper from Carlisle and colleagues is helpful in mapping this research terrain, highlighting gaps that are relevant to service delivery, and setting a direction for future work.
There is a hopeful message: according to published research, services appear to help. People engaging with them often report improved safety and wellbeing. But the evidence base remains fragile and uncertain.
Better measurement, more methodologically rigorous designs, and deeper engagement with lived experience could help to move the field from plausible effectiveness to demonstrable impact. Until then, we are left navigating uncertainty, guided by imperfect evidence, but anchored in the urgent need to respond to violence.
At present, we are guided by imperfect evidence, but anchored in the urgent need to respond to violence. We need better measurement, more rigour and deeper engagement with lived experience to move from plausible effectiveness to demonstrable impact.
Statement of interests
Vishal Bhavsar is a fellow member of the VISION consortium, though did not contribute to the creation of this systematic review.
Laura Hemming has no interests to declare.
Editor
Edited by Laura Hemming.
Links
Primary paper
Carlisle, S., Bunce, A., Prina, M., McManus, S., Barbosa, E., Feder, G., & Lewis, N. V. (2025). Effectiveness of UK-based support interventions and services aimed at adults who have experienced or used domestic and sexual violence and abuse: a systematic review and meta-analysis. BMC Public Health, 25(1), 1003.
Photo credits
- Photo by Kelly Sikkema on Unsplash
- Photo by Volodymyr Hryshchenko on Unsplash
- Photo by Vitaly Gariev on Unsplash
- Photo by Compagnons on Unsplash
- Photo by Ning Shi on Unsplash
- Photo by William Warby on Unsplash
