Key findings about what the evidence covers (i.e. from the Evidence and Gap Map)

Many of the studies are recently-published:

  • There has been a significant uptick in the number of studies in this area (see graph below): now 108 completed primary studies, up from 58 in the first iteration of this EGM, which used a search in 2019. {Though this needs caution because there can be giant delays in publishing the study. For example, one study published in 2021 examines an intervention in 2013.}
  • There are more treatment studies now: in the first EGM, there were studies of only two interventions (though one had generated many papers), and both had started before 2000; now there are studies of more treatment interventions.
  • There are eight ongoing studies on the map, up from three in our first EGM.

The evidence is concentrated:

  • Though much less so than in our first iteration three years ago
  • Most studies are about prevention. Prevention was examined in 120 papers (some studies generate more than one paper).
  • Treatment was studied in only a handful of primary studies. On disclosure, we found three primary studies of interventions aiming to facilitate disclosure; last time, we found none.
  • Most studies assess education-based prevention programmes, in early education and in school settings. Fully 60 of the 108 completed primary studies look at that.
  • More than 70% (n=87) of the completed primary studies assess curriculum-based prevention programmes educating children about sexual abuse and teaching safety skills, e.g., how to handle ‘secrets’ and whom to tell in school settings.
  • Sexual abuse is more studied than other types of abuse. Of the 108 completed primary studies, 75 look at sexual abuse, whereas 49 look at other types of abuse. (The numbers do not add up because some studies look at multiple types of abuse.)

The outcomes studied:

  • Few studies look at actual incidences of abuse (n=21)..
  • Most studies assess intermediate outcomes, such as children’s acquisition and retention of knowledge (n=84), but not actual disclosure of incidence.
  • The outcomes were usually self-reported by children or young people, which are proxies for incidence. They may or may not be reliable proxies.
  • Only one study looks at educational attainment.
  • Only seven completed studies assessed interventions to stop adults from offending in organisations.
  • Studies addressed only three of the seven INSPIRE strategies of the World Health Organization. Four of the INSPIRE strategies have no rigorous impact evaluations.

On geography:

  • Most studies are from high-income countries. The USA dominates, with 40% of completed primary studies (n=44).
  • There are very few studies from Africa (n=10), and they cover only four interventions, all in prevention. In other words, there are zero studies from Africa about encouraging disclosure, response or treatment.
  • There are no studies from India, despite its billion people. There is only one study from South Asia (Pakistan) and only one from South East Asia (Indonesia), despite its huge populations.

On reliability:

  • Few studies have low risk of bias. Most of the completed primary studies have a medium (n=60) to high (n=28) risk of bias, i.e., the ‘answers’ they report may be wrong.

On types of children served:

  • Only a few studies focus on children particularly at-risk: programmes targeting them were studied in only six completed primary studies.
  • No studies examine programmes related to clergy, churches, synagogues or other religious communities. (One study includes clergy, but only as a type of ‘child-serving professional. The clergy were in groups with other professionals, and the data about the clergy were not broken out.)

Notable gaps in the evidence are around:

  • Interventions encouraging disclosure of maltreatment (n=6), organisational responses (n=23) and treatment (n=17).
  • Africa (few studies, and none outside prevention)
  • China (over a billion people; only three completed primary studies)
  • India (over a billion people; no studies at all, nor protocols for new ones)
  • Studies which look at actual incidences of abuse, or educational attainment, to which a lot of funding is attached.

Key findings about what the evidence says (i.e. from the Guidebook)

  • Almost all the interventions succeed: No intervention studied seems to cause harm! However, most results attenuate over time: most of the interventions studied are training, and people forget things over time.
  • Most effects are small! – that is true not just of child protection but of most social interventions: An effective programme may improve knowledge by 20-30 percent and reduce abuse by 10-20%. There is no vaccine for child abuse.
  • Some interventions have no effect, or at least no effect on some outcomes. For example, a programme run in the Netherlands with at-risk boys living in residential care aimed to reduce sexual harassment by them but found no effect.
  • There are some mixed results. Some interventions found a positive effect on some outcomes but no effect on others. For example, Edwards (2019) examined a bystander programme in US high schools: it found no statistically significant effects on participants stopping harassment, speaking against blame or excuses, or talking to an upset person, but did find improvement in victim empathy and denying that rape is possible or had occurred.
  • Only three studies provided information on the programme’s costs (*listed below). This is really a shame given that anybody considering running a programme needs to know its costs.

Specific findings include:

  • School-based programmes have succeeded in raising children’s awareness about physical and sexual abuse, training them on what to do, and increasing disclosure. This finding is consistent across over 60 studies in many countries.
  • School-based prevention interventions that involve parents improved knowledge regarding abuse and attitudes toward violence against children.
  • School-based interventions showed significant and positive improvements in children’s knowledge and awareness about sexual abuse and empower children to recognise, prevent, avoid, and report sexual abuse (if it happens).
  • School-based programmes have been successful in raising awareness about physical and sexual abuse in children, training them on what to do, and increasing disclosure.
  • Staff training programmes include the IRIE classroom toolbox and Second Step Child Protection Unit to improve teachers’ knowledge, attitudes, and behaviours when working with children. These programs have also shown promise in identifying maltreatment and avoiding violent behaviours by teachers.
  • School-based sexual abuse prevention programmes have not been found to increase children’s anxiety.
  • Though prevention programmes appear to not increase anxiety among children, their impact on other mental health issues is unclear.
  • The effects of school-based interventions to prevent peer violence and gender-based violence are unclear: the studies are inconclusive. Further research is necessary to identify the most effective strategies and interventions to create safe and inclusive learning environments for all students.
  • The Bucharest Early Intervention Project (BEIP) in Romania showed that moving orphaned children out of awful orphanages and into foster families significant improved most developmental, physical, mental, emotional, and cognitive health outcomes. This was fairly early intervention: the children were moved before their third birthday.
  • The evidence on institutional responses to violence and abuse against children and the impact of these interventions, such as the Multi-Agency Investigation & Support Team (MIST) and the Clinical Decision Tool for Paediatric Intensive Care Unit (PICU), is weak. MIST improves co-ordination of statutory agencies (e.g., the police, social services, the fire service), to improve response to child abuse cases. PICU is a specialized unit in a hospital which provides care to children and young people with severe trauma.

* Programmes for which cost information was provided were:

  • In the Irie Classroom Toolbox intervention in Jamaica (Baker-Henningham 2021), teachers received lunch and a small stipend to cover transportation (USD 4 per workshop).
  • In Skhokho (South Africa), a whole-school approach to reduce gender-based violence among eighth graders (Jewkes 2019), caregivers received Rand 50 and teenage participants received Rand 20 for transportation costs for each session.
  • Similar compensation for travel (USD 2.17) and free food and beverages were available to institutional caregivers in Tanzania in the ICC-T intervention (Hecker 2021).