Violence Against Children Surveys (VACS): towards a global surveillance system

Abstract

Objective

To describe the Violence Against Children Surveys (VACS). The survey is a national, household survey that systematically measures the prevalence, nature and consequences of sexual, physical and emotional violence against children.

Design

This report provides information about the history, implementation, ethical protections, utility, results, limitations, and future directions of the VACS work.

Results

The study has been implemented in 11 countries in Africa, Asia and the Caribbean, providing each of these countries with baseline data and momentum to address violence against children as a public health and human rights priority. These data are novel in each country, and VACS is well poised to contribute to an existing surveillance system or be used as the basis of a periodic surveillance system.

Conclusions

Without ongoing surveillance to assess prevalence and the impact of policy, prevention and response programming, violence will likely continue to be overlooked as the linchpin public health crisis that it is, globally and in individual countries.

INTRODUCTION

Violence against children globally

Violence against children is of pandemic proportion and is increasingly a global public health, human rights and development priority. Violence against children cuts across class, socioeconomic status and culture and is highly prevalent across the globe.¹ Besides the trauma and injury associated with the violence at the time of incident, a robust body of literature has documented a host of longterm health sequelae and risk-taking behaviours, including mental health outcomes,^2–4 sexually transmitted infections such as HIV,^5–8 sexual risktaking behaviours such as multiple concurrent sex partners or inconsistent condom-use with casual sex partners,^9–12 and alcohol and drug use and abuse.¹³ Further, the Adverse Childhood Experiences Study and the National Survey of Children’s Exposure to Violence, as well as other studies, demonstrate a dose-response effect, such that recurrent exposure to one type of violence and/or polyvictimisation have greater association with poor health outcomes.^14–19 Child victims of violence are also at an increased risk of revictimisation and later violence perpetration.^10,20–23

Violence against children does not merely affect the victim, but has consequences across the social ecology,²⁴ also impacting the community, and the larger society. The economic toll of violence is considerable. For example, aggregate lifetime costs of child maltreatment alone in the USA are estimated to be 121.6 billion dollars.²⁵ Childfund estimates the global economic impact of violence against children to be as much as $7 trillion, much higher than any estimated investment to effectively prevent violence against children.²⁶ Given the health, human rights and economic impacts, global bodies as well as national governments can clearly benefit from the implementation of prevention strategies.

Globally, there are several surveys that document violence against children including Multiple Indicator Surveys, WorldSAFE, Health Behavior in School-Aged Children Surveys and Global School Health Surveys; however, these studies tend to focus on a single type of perpetrator or location of the event (eg, parental punishment and schools, respectively). The Violence Against Children Surveys (VACS) is a national, household survey designed to capture the magnitude and prevalence of violence against children in a given country. The survey collects information on the prevalence of sexual, emotional and physical violence, multiple perpetrator types, event location, health outcomes, risk behaviours and protective factors. This paper describes the contribution VACS has made to global understanding of violence against children, examples of actions taken by governments in response to VACS data and how the survey can be positioned to become a surveillance system.

The Violence Against Children Surveys

Despite evidence of the pervasiveness of violence against children, the detriment to the individual child victim’s health and well-being, as well as the heavy cost to society, few countries have population-level data on the prevalence of violence against children.^27,28 There is a need for rigorous data collection to inform data-driven response and prevention programmes and policies focused on violence against children at local, national and global levels.

Anecdotal evidence that violence may be a driver of the HIV epidemic in Swaziland, particularly among adolescent girls, prompted key stakeholders in Swaziland to explore the feasibility of a national survey to collect rigorous, systematic, population data on the magnitude and context of violence against children. In 2007, the Government of Swaziland conducted the first VACS with technical and logistical support from Unicef-Swaziland and from the Division of Violence Prevention at the US Centers for Disease Control and Prevention (CDC). Survey data indicated high rates of violence, and subanalyses demonstrated an association between violence and a host of negative health outcomes.²⁹ Although the Swaziland VACS only collected data on girls and young women (age 13–24 years), the study’s success demonstrated the feasibility of conducting a national survey on violence against children as well as the leadership and vision to respond to the data.

The Swaziland results garnered considerable attention, inspiring two additional countries to undertake VACS (Tanzania and Kenya) and leading to the 2009 establishment of Together for Girls (TfG), a global public-private partnership with the mission to end violence against children. The partnership includes CDC, the US Agency for International Development; the President’s Emergency Plan for AIDS Relief, US Department of State’s Office of Global Women’s Issues; Foreign Affairs, Trade and Development Canada; Unicef; the Joint United Nations Programme on HIV/AIDS; WHO, the United Nations Population Fund, UN Women, Becton, Dickinson and Company, Nduna Foundation, CDC Foundation, and Grupo ABC. Under the partnership, countries receive support to conduct VACS which mobilises country-led and multisectoral prevention and response, and informs evidence-based solutions.

VACS has now been implemented in 11 countries: Swaziland (2007), Tanzania (2009), Kenya (2010), Zimbabwe (2011), Haiti (2012), Cambodia (2012), Indonesia (2013), Malawi (2013), Nigeria (2014), Zambia (2014) and Laos (2014). Of these, seven countries have published final reports (Swaziland, Tanzania, Kenya, Zimbabwe, Haiti, Cambodia and Malawi). Additional countries are in the planning or implementation phase and several countries have expressed interest in repeating VACS.

Due to the well-established methods and systematic approach to conducting VACS and responding to the data, VACS is wellpoised to result in the establishment of broader violence surveillance systems in countries where it is conducted.³⁰ In addition to the methodological strengths, there are multiple unique and innovative contributions of VACS that position it to fill an existing gap in violence surveillance: (1) in each country VACS has provided national baseline data on violence against children; (2) due to the cross-cutting nature of violence against children, each country forms a multisectoral steering committee or task force to oversee and guide the process from survey planning through the design of data-driven policy and programming response; (3) VACS is only implemented in countries where high-level governmental leadership commits to responding to data prior to implementation; (4) VACS is the only national household survey that systematically collects population-level data on the past 12 months and lifetime exposures to physical, sexual and emotional violence from girls and boys aged 13–24 years.

VACS METHODS

Procedures

VACS is a national, cross-sectional household survey of children and young adults aged 13–24 years, using a multistage geographically clustered sample design. The three-stage cluster and split-sample design ensure that only one individual is selected per household and that boys and girls are interviewed in different enumeration areas (EAs). The male and female samples are drawn separately and produce separate estimates. VACS samples an average of 200 EAs with an average cluster size of 25 households per EA in the 11 countries.

The VACS implementing partner in each country—often the national statistics agency or a local university—is responsible for recruiting interviewers. Preference is given to those with prior experience in a sensitive subject matter such as gender-based violence or HIV and those with prior experience on household surveys. Interviewers are also hired to ensure appropriate recruitment by language skills and cultural context, depending on the demographic profile of each country. All staff receive extensive training on methods, ethics and study procedures.

In 2012 the survey moved to electronic data collection. The questionnaires are programmed in CSPro (a free software programme developed by the US Census Bureau) into netbooks. Given the complexity of the skip patterns and logic sequencing, the switch to electronic data collection eliminated interviewer routing errors; eliminated the need for data transcription and entry; reduced the amount of time spent on data entry, data cleaning and data analysis; and reduced training on these areas. The results are significantly improved data quality, greater efficiency in producing a final country report and increased capacity for fielding household surveys using electronic data collection.

Prior to implementation, the survey is pilot-tested and findings from the pilot study are used to refine the survey instrument, field operations and follow-up procedures for respondent support. Every respondent provides verbal consent prior to enrolment in the survey. However, the study is presented to households and parents/guardians in general terms to avoid any unintentional consequences of retribution against the respondents by perpetrators of violence who may be household or community members. In addition, all interviews are conducted in private spaces to ensure confidentiality and reduce the potential of disclosure. The study follows documented ethical and safety procedures to ensure the rights and safety of each respondent, many of which were adapted from WHO guidelines on ethics and safety in studies on violence against women.³¹

Measures

VACS assess the burden of childhood sexual, physical and emotional violence through multiple measures. Across completed VACS the definition of sex and the precise wording for physical and emotional violence questions and probing for forms of violence and types of perpetrators varied slightly from country to country and are available in country technical documentation. Therefore, although each survey measures the burden of childhood sexual, physical and emotional violence, cross-country comparisons should be made with caution. Measures of sexual violence include unwanted touching, unwanted attempted sex, pressured sex and physically forced sex. In analyses, these measures are used as individual variables as well as combined for a composite measure of ‘any sexual violence.’ Physical violence measures generally include being punched, kicked, whipped, beat with an object; choked, smothered, attempted to drown, burned; and/or threatened/victimised with a weapon. Emotional violence measures generally include an individual being told s/ he was not loved or did not deserve to be loved; someone wished that s/he had never been born or was dead; and being ridiculed, humiliated, threatened with abandonment or forced to leave home. The survey includes additional measures on the context of violence such as time of day, setting and perpetrator characteristics. More information about the VACS measures, definitions and questionnaires can be obtained from the CDC VACS website (http://www.cdc.gov/violenceprevention/vacs/index.html).

Limitations

Despite the extent of information that VACS collects, there remain some limitations important to highlight. First, as a household survey, we are not able to collect data on children living outside of family care who may be most vulnerable to violence victimisation. Second, the survey only collects contextual data on the first and most recent incidents of each type of violence. As such, for those who experienced more than one event, we do not have the complete context by each incident. Third, the survey relies on self-reports and might underestimate the true prevalence if the respondent does not feel comfortable disclosing his or her experiences. Finally, the survey may be more or less successful in different cultural contexts. For example, we find lower rates of violence in Asian countries, which could be representative of the violence epidemic or could mean lower disclosure in contexts where consequences for victims and stigma may be greater. As part of the VACS process, the survey is sensitised to local cultural and linguistic contexts; however, due to the national nature of these surveys, particular, local-level differences may not be detected. Further, given that many countries are multicultural, culturally specific forms of violence (eg, initiation rites, or specific harsh parenting practices) may not be picked up at sufficient rates in the current approach.

USING VACS FOR SURVEILLANCE

VACS presents an opportunity to develop a global surveillance system for childhood violence, as evidenced by the type of data and indicators collected, the high response and disclosure rates, the population studied and the global coverage attained through the survey. In addition, due to the survey’s widespread acceptability in various regions of the world, a surveillance system built on the implementation of VACS to date has the potential to provide regional, national and global tracking of violence against children.

Comprehensive indicators

The information collected on violence represents novel and comprehensive data for countries participating in VACS. Globally, there remains a dearth of representative population-level data to guide programmatic work on childhood violence prevention; VACS is providing nationally representative estimates of a wide range of indicators for the first time in many countries. Figure 1 presents the prevalence of sexual, physical and emotional violence experienced before age 18 years among respondents aged 18–24 years in the first seven VACS countries with available data. In the majority of these countries, more than 25% of girls and 10% of boys disclosed childhood sexual violence. Rates of physical violence are even higher than sexual violence in nearly all countries studied, often by a factor of 2 or more. For example, whereas 21.8% of women and 14.8% of men in Malawi reported sexual violence as a child, 42.4% of women and 64.5% of men reported physical violence. Among the six countries where male and female respondents were surveyed (Tanzania, Kenya, Zimbabwe, Haiti, Cambodia and Malawi), four countries had similar rates of physical violence among boys and girls; two had higher rates among boys (figure 1). Lastly, the prevalence of emotional violence was reported to be between approximately 20– 30% for boys and girls in the majority of countries studied. The indicators measured by VACS permit countries to monitor changes in childhood violence over time and better assess the impact of nationwide initiatives to prevent violence against children that have followed VACS.³²

Figure 1.

Lifetime prevalence of experiencing physical, emotional and sexual violencea: 7 countries, 2007–2014. aViolence occurring before age 18 years among respondents aged 18–24 years. PV, physical violence; EV, emotional violence (by a parent, adult caregiver or other adult relative), SV, sexual violence. Year of data collection: Swaziland, 2007; Tanzania, 2009; Kenya, 2010; Zimbabwe, 2011; Haiti, 2012; Cambodia, 2013; Malawi, 2014.

However, beyond essential surveillance data, VACS also provides a wide range of indicators that can be used to guide individual, community-level or policy solutions to violence. Some additional measures that VACS collects include information on characteristics of the violent event (eg, perpetrator characteristics, location, time of day), service seeking and service availability, risk and protective factors (eg, family and peer relationships, school attendance, economic deprivation), local norms about violence and health consequences of violence (eg, sexually transmitted illnesses, poor mental health, substance abuse). Thus, VACS can permit countries not just to monitor trends in childhood violence exposure, but also devise locally relevant solutions. The full range of indicators collected by VACS is available in country technical documentation (all reports available at: www.cdc.gov/violenceprevention/vacs/index.html)

Response rates

Notably high response rates have been achieved through VACS; overall respons rates are greater than 80% in all countries with published results and similarly high rates achieved for male and female participants (table 1). This demonstrates the ability of VACS to obtain a representative sample of household-dwelling children, which is an advance in monitoring the burden of sexual violence among children globally. Previous studies have been limited to smaller geographical areas within countries or focused on specific subpopulations, such as only children attending school.^33–35

Table 1.

Number of respondents and overall response rates by country, Violence against Children Surveys, 2007–2014^*

	Girls		Boys
Country	Number of respondents	Overall response rate (%)†	Number of respondents	Overall response rate (%)
Swaziland, 2007	1244	96.3	-	-
Tanzania, 2009	1968	93.3	1771	92.5
Kenya, 2010	1227	84.8	1456	80.4
Zimbabwe, 2011	1062	80.4	1348	82.0
Haiti, 2012	1457	85.6	1459	82.0
Cambodia, 2013	1121	91.0	1255	89.9
Malawi, 2014	1029	84.4	1133	83.4

^*Number of respondents and response rates are calculated for the entire sample only (ages 13–24 years).

^†Only individual response rate available for Swaziland.

Disclosure rates

Beyond excellent response rates, VACS also achieves high levels of disclosure of violence. This is critical given the sensitive nature of the questions. For example, in six of the seven countries with published reports, more than one in five women disclosed childhood sexual violence. Interviews are administered by highly trained interviewers of the same sex as the respondent, which may contribute to high levels of disclosure. Evidence indicates that disclosing such experiences is generally not distressing for participants and may even be therapeutic.^36,37 Participants who disclose violence during the interview are offered a referral to support services, ensuring that VACS can also be a link to support and safety services. Furthermore, all participants are offered a list of counselling, medical, social and legal support services in their geographical area, regardless of whether they reported any violence.

Novel data from a unique study population

Additionally, VACS data permit novel analyses of childhood violence. For example, to date, information on violence against boys is particularly scarce globally. Beyond the importance of preventing violence exposure among male children, better understanding male victimisation is essential given the strong associations between male victimisation, particularly sexual violence, and subsequent perpetration of violence documented in previous studies.^21–23 As mentioned above, unique to VACS is the inclusion of respondents as young as 13 years old; most other national surveys, such as the Demographic and Health Surveys, interview respondents were 15 years and older. Including younger respondents provides an understanding of violence at a younger range of the age spectrum, which is critical to designing early interventions needed to minimise the sequelae of these adverse experiences.¹⁴

Global coverage

Lastly, the coverage of VACS is considerable, particularly in countries where youth experience a high burden of illness and injury. Nationally representative estimates of childhood violence have been generated in each of the seven countries with published final reports, where over 17 500 children and youth have completed the survey; thus, VACS has currently sampled 11% of the world’s 13–24-year-old population. These data are useful to governments as they implement and evaluate intervention programmes. Particularly at the global level, VACS nations now have specific data to monitor progress towards the Sustainable Development Goal target 16.2: End abuse, exploitation, trafficking and all forms of violence against and torture of children.

RESPONDING TO VACS DATA

National leadership by governments and in-country stakeholders

Government commitment to respond to VACS data is a key component of the VACS planning process. VACS findings are the basis for national governments and civil society to make evidence-informed coordinated policy and programme changes, including legal and policy reform, improved services for children who have experienced violence, and prevention programmes. To initiate this, a governmental ministry leads a multisectoral task force or steering committee to plan and implement VACS. This body is comprised of multiple government ministries such as health, education, justice, finance and social welfare; as well as multilateral agencies; donor countries; and global non-governmental organisations and other civil society representatives. Following the completion of VACS, these diverse agencies and organisations work together to respond to the survey findings and secure resources for those responses. This method of securing broad buy-in for the results prior to the survey is a critical strategy in generating national commitment to respond to VACS data. In line with this commitment, the governments of Swaziland, Tanzania, Kenya and Zimbabwe have launched multisectoral supported plans in response to the findings of VACS in their countries; Haiti, Malawi and Cambodia are in the process of doing the same. As part of this process, VACS countries often initiate ‘data-to-action’ workshops when results are available. Participating stakeholders review and interpret the VACS data, and develop sector-specific plans on how to address violence against children. Unicef, CDC and other TfG partners provide technical support for these workshops.

Guidance on preventing and responding to violence against children

There is an emerging body of research that demonstrates what works to prevent and respond to violence. To support countries and provide technical guidance, TfG partners have created technical packages on best practices and intervention areas. Across packages, these approaches consistently include multisectoral strategies; a focus on social norms change, parenting programmes and early childhood interventions; economic strengthening and empowerment; legal and policy reform; education; improvements in service access and quality; and a focus on strengthening monitoring, evaluation and surveillance. One example is CDC’s global technical package, THRIVES (Training in parenting; Household economic strengthening; Reduced violence through legislative protection; Improved services; Values and norms that protect children; Education and life skills; Surveillance and evaluation).³⁸ This package represents the best available evidence for preventing and responding to violence against children, based on the current literature of evidence-based interventions. The strategies are applicable across health, social services, education and justice sectors. Though much of the evidence presented in THRIVES is based on research and practice in high-income countries, emerging research from low-income and middle-income countries is expanding rapidly. The problem of violence against children is too large and has too many urgent consequences to wait for contextually specific solutions for each country. Rather, promising evidence from one context should be assessed for adaptation to other contexts. Another example of technical guidance for responding to violence against children is Unicef’s Ending Violence Against Children: Six Strategies for Action.³⁹

CONCLUSION

VACS is a population-based survey of girls and boys aged 13–24 years that uses well established methods for sampling and implementation and a standardised protocol. VACS is increasingly emerging as an innovative tool to provide global and national data and a platform for policy reform and evidence-based programming. Critics may argue that VACS is too complex and costly to repeat. However, we should look to other public health topics and remember that nationally representative surveys are often the only source of data for these topics (eg, HIV) and are regularly repeated despite the high costs. The questionnaire has a core set of questions that are implemented in all countries, allowing for adaptation to the cultural context. There is standard, rigorous training for field staff to assure high quality data, and electronic data collection allows for real-time quality assurance monitoring during data collection. Although VACS in itself is not a surveillance system, it is well poised for use in surveillance by repeating the survey at regular intervals.³⁰

Due to the cost of implementing VACS as well as the complexity associated with preventing and responding to violence, longer intervals may be prudent. Changing social and cultural norms that actively or passively condone and/or perpetuate interpersonal violence is a long-term process, and a successful violence surveillance system must take this into consideration. However, working with other surveillance systems to monitor a subset of the VACS indicators more frequently is another complementary surveillance option and can provide additional critical data on associations between violence against children and other health issues.

A surveillance system that routinely measures violence against children nationally, regionally and/or globally can show trends in violence over time and can inform and help monitor the effects of changes in policies or programmes. For example, this type of surveillance can collect data on rates of violence, identifying changes in specific types of violence, thus helping to monitor whether implemented interventions change norms that support violence or if newly implemented policies prohibiting or criminalising violence is contributing to a decrease in violence. The monitoring of national efforts through VACS surveillance could fill a critical need in the global effort to prevent violence against children.

The data from each VACS completed continue to demonstrate the urgency for a global response to the violence inflicted upon children worldwide. The global investment in the surveillance of violence against children through VACS and the commitment to responding to these data with evidence-based programmes and policies is necessary to begin addressing this critical public health issue. A surveillance system to facilitate this effort is overdue; expanding VACS from a ‘single point in time’ survey to one that is used as the basis of periodic surveillance or contributes to an existing surveillance system can bridge this gap.

Tanzania: data to action.

The Government of Tanzania launched the findings of VACS in 2011. Based on the data, ministerial commitments were developed into the Multi-Sector National Plan of Action to Prevent and Respond to Violence against Children 2013–2016, with clear objectives, costed activities and targets. The plan focuses on an integrated approach to strengthening national and district child protection systems, with a strong monitoring and evaluation component.

Achievements include the establishment of 36 multisectoral District Child Protection Systems in various stages of development, and 12 One Stop Centers with trained staff; an action plan and training for establishing and managing Gender and Children’s Desks in police stations; development of a teachers’ code of conduct; protection guidelines for safe schools; child protection budgeting guidelines; and violence against children communication strategies to address harmful social norms.

Sustained advocacy and work is resulting in high-level support for child protection work with increased financial allocation for child protection and scale up of prevention and response work with a focus on gender equality and social norms change. Much remains to be done to ensure sustainability, and Tanzania is already discussing next steps when the current plan concludes in 2016.

What is already known on this subject.

• ► There is a dearth of rigorous epidemiological data on the magnitude and impact of violence against children globally, but available data indicate that violence against children is of pandemic proportion.
• ►The Violence Against Children Survey (VACS) has published results in seven countries in Africa, Asia and the Caribbean, demonstrating high rates of physical, emotional and sexual violence among girls and boys. Other US studies (the Adverse

• Childhood Experiences Study and the National Survey of Children’s Exposure to Violence) have demonstrated high rates of violence against children in high-income settings as well as long-term negative health implications associated with experiences of violence in childhood.

What this study adds.

• ►This study describes the origin and development of VACS. Furthermore it highlights the unique contributions of VACS in childhood violence surveillance beyond current capabilities, such as the collection of a diverse set of indicators ranging from physical, emotional and sexual violence prevalence; violent event characteristics; local violence norms; risk and protective factors; and health consequences of violence.
• ►The utility of VACS as a tool for host-nation response to violence is discussed. VACS can and has been used for legal and policy reform, improving services for children who have experienced violence, and the development of novel prevention programmes.
• ►Conducted at repeated intervals, VACS holds promise for establishing an improved global surveillance system for violence against children that leads to concrete violence prevention activities by countries.