They count for nothing: Poor child protection statistics are a barrier to a child - centred national framework.

Transcription

1 They count for nothing: Poor child protection statistics are a barrier to a child - centred national framework. Karen Broadley Child Abuse Prevention Research Australia, Monash Injury Research Institute, Monash University Chris Goddard Child Abuse Prevention Research Australia, Monash Injury Research Institute, Monash University Joe Tucci Australian Childhood Foundation February 2014

2 Contents Introduction Page 5 Part 1: Surveillance, an essential component of public health approach Page 8 Epidemiology one of the basic sciences of public health Page 9 Examples of public health surveillance Page 10 The source of the surveillance data Page 10 The importance of high quality child protection surveillance data Page 11 Child protection surveillance data what do we need to know? Page 11 Further benefits of a high quality data surveillance system Page 13 A high quality data surveillance system allows interventions to be evaluated Page 14 Public reporting of surveillance data Page 15 Part 2: The Australian context Page 16 There are no reliable data in relation to child protection notifications Page 17 There are no reliable data in relation to child protection substantiations Page 18 There are no reliable data in relation to the ethnicity of children, or the Page 19 characteristics of the parents coming to the attention of child protection authorities There are no useful data in relation to the source of the notification Page 20 The lack of reliable surveillance data limits the ability of the Australian Page 20 Government and public health community to effectively respond to the problem of child abuse and neglect The lack of reliable surveillance data makes it impossible to know whether Page 22 early intervention and prevention efforts are effective The states and territories Page 24 Conclusion Page 25 Recommendations Page 27 Appendix One: Child Protection Australia reports stating that much of the Page 28 data they report on is worthless Appendix Two: Status of National Framework indicators of change by Page 31 supporting outcome References Page 33 Acknowledgements The Australian Childhood Foundation and Child Abuse Prevention Research Australia would like to thank Kirsty Hamilton for her financial and moral support. All the children who appear in this publication are models. Page 2

3 About the Authors Karen Broadley works part-time at Child Abuse Prevention Research Australia, in the Monash Injury Research Institute, Monash University. She has more than twenty years experience working in the field of child and family welfare. Throughout the 1990 s Karen worked as a youth worker, particularly in the area of youth justice, also including youth housing, and family reconciliation. From 2000 to 2010 she worked within the Victorian statutory child protection system, as a child protection practitioner, then a team leader. She held a number of roles in the intake, response, and long term case management programs. She also worked as a high risk adolescent consultant, and specialist infant protective worker. Karen has a BA (Youth Affairs), and Master of Social Science (Family Studies). She completed a minor thesis in relation to the emotional abuse of children. Adjunct Professor Chris Goddard is the Director of Child Abuse Prevention Research Australia in the Monash Injury Research Institute, Monash University. He worked in social services in the UK and child protection in Australia before undertaking extensive research into child abuse and family violence. He has published widely in academic journals and writes regularly for newspapers. His most recent books include: In the Firing Line (Wiley) with Janet Stanley; The Truth is Longer than a Lie (Jessica Kingsley) with Neerosh Mudaly, Human Rights Overboard (Scribe) with Linda Briskman and Susie Latham (awarded an Australian Human Rights Commission award); and, Physical Punishment in Childhood (Wiley-Blackwell) with Bernadette Saunders. Dr Joe Tucci is the Chief Executive Officer of the Australian Childhood Foundation. He is a social worker and registered psychologist. He has over 20 years experience in child protection, child and family therapy and child welfare research. He has completed a doctoral thesis exploring the issues of child psychological abuse. He has been a consultant to state child protection systems nationally on a number of child abuse and child welfare evaluative projects. He has demonstrated experience in developing and implementing child focused therapeutic programs and child abuse prevention campaigns. His writing has been published in both Australian and international academic journals and the broader media. He is an Adjunct Research Fellow with Child Abuse Prevention Research Australia. Page 3

4 Australian Childhood Foundation Child Abuse Prevention Research Australia at Monash University The Australian Childhood Foundation is an independent children s charity working in a number of ways to prevent child abuse and reduce the harm it causes to children, families and the community. Specialist Trauma Counselling. We provide a range of specialist counselling services for children and young people affected by abuse and for their families. Therapeutic care programs. We provide a range of therapeutic care programs within residential and foster care settings. Advocacy for children. We speak out for effective protective and support services for children and young people. All our programs affirm the importance of children. Education. We provide community and professional education, consultancy and debriefing programs. These programs aim to improve responses to children and young people who have experienced or are at risk of abuse, family violence and neglect. Child abuse prevention programs. We run nationally recognised child abuse prevention programs that seek to decrease the incidence of child abuse and raise awareness about how to stop it even before it starts. Inspiring and supporting parents. We provide ongoing parenting education seminars and easily accessible resources to strengthen the ability of parents to raise happy and confident children. Research. In partnership with Monash University, we have established Child Abuse Prevention Research Australia to research the problem of child abuse and identify constructive solutions. The Australian Childhood Foundation won the 1998 National and State Violence Prevention Awards for its efforts to prevent child abuse. In 2005, it was awarded the National Child Protection Award by the Australian Government. The Australian Childhood Foundation relies on the support of the community to enable it to continue its programs and services. Child Abuse Prevention Research Australia (CAPRA) was established as a joint initiative between the Australian Childhood Foundation and Monash University. The aim of CAPRA is to provide evidence-based advice to governments and community-based organisations to inform the development of effective policies and practices aimed at dramatically reducing the incidence of child abuse, neglect and murder. In recent years, CAPRA has received support for major research projects including: Child abuse deaths; Therapeutic responses to children who have been abused; Violence against child protection workers; Examining community attitudes towards the prevention of child abuse and family violence; and, Physical punishment of children. CAPRA has a number of PhD students in Australia and overseas, and Adjunct Professor Chris Goddard is regularly invited to speak at major conferences around the world. For further information please contact: Adjunct Professor Chris Goddard Director Child Abuse Prevention Research Australia Monash Injury Research Institute Monash University Building 70 Wellington Road Clayton Vic 3800 PO BOX 525 Ringwood VIC 3134 Phone: (03) info@childhood.org.au Website: chris.goddard@monash.edu Website: Page 4

5 Introduction In 2008, the Australian Childhood Foundation and Child Abuse Prevention Research Australia collaborated with Access Economics to estimate the cost to the Australian community of child abuse and neglect. They conservatively estimated that 177,000 children under the age of 18 years were abused or neglected in Australia in 2007, and that the cost incurred by the Australian community was $10.7 billion (Taylor et al., 2008). Further, they conservatively estimated that of these 177,000 children, 130,237 were abused and neglected for the first time in Based on this figure they calculated the projected monetary cost, over the life time of these children, to be $13.7 billion (Taylor et al., 2008). This is only part of the story. The less measureable cost is in relation to the pain and suffering experienced by children in the short and long term. Research confirms that the devastating long term outcomes for children who experience abuse and neglect are manifold. Perry and Hambrick for example, state: In extreme cases of developmental challenge such as maltreatment threat, neglect, humiliation, degradation, deprivation, chaos, and violence children express a range of serious emotional, behavioral, cognitive, and physiological problems (2008, p. 38). Lamont (2010), in his review of the literature, found that the consequences of child abuse and neglect include difficulties in forming relationships, poor physical health, learning and developmental difficulties, behavioural problems, and mental illness. He noted that other lifestyle consequences include an increased risk of developing drug and alcohol addictions, engaging in violence, criminal behaviour, long term unemployment, and homelessness (Lamont, 2010). Frederick and Goddard (2007) in their research also found that children who experience abuse and neglect are more likely, as adults, to experience poverty, disadvantage and exclusion. The short term impacts of child abuse and neglect are also important to acknowledge. As Browne and Finkelhor have suggested however, there is an unfortunate tendency towards focusing on long term effects and to consider effects less serious if their impact is transient and short term. They assert that: Page 5

6 Adult traumas such as rape are not assessed ultimately in terms of whether they will have an impact on old age: They are acknowledged to be painful and alarming events, whether their impact lasts 1 year or 10. Similarly childhood traumas should not be dismissed because no long term effects can be demonstrated (1986, p. 76). So what can be done to deal effectively with this complex and intractable problem? Many researchers and policy makers in Australia have called for a public health approach to child protection. For example, in 2008 a report commissioned and released by the Australian Research Alliance for Children and Youth (ARACY), claimed: There is consensus by the Australian, state and territory governments and community organisations in Australia of the need to reduce the burden on the tertiary end of the system and to enhance primary and secondary supports and services in line with the public health model (Allen Consulting Group, 2008, p. 5). O Donnell, Scott and Stanley have also argued that protecting children requires a: public health approach with a platform of universal preventative services, secondary prevention with targeted services for at-risk families, and interventions at various levels depending on the risk of harm to the child and the needs of the family (2008, p. 329). Clearly this call has been heard, with clear evidence that it is shaping the policy paradigm of child protection in Victoria and other jurisdictions (Hunter, 2011; Scott, 2006). The Council of Australian Governments used a public health model to develop what has become a key driver of reform in the integration of statutory child protection systems within a broader framework underpinned by child and family support services (Babbington, 2011; Council of Australian Governments, 2009). Similar arguments in favour of public health models have been made internationally. For example, the 2002 World report on violence and health, and the 2003 World Health Assembly resolution on implementing the report s recommendations, stressed the role of public health in preventing and responding to child maltreatment (World Health Organization, 2006, p. 1). In 2006 the World Health Organization advocated a public health model of child protection, one that focuses on prevention and early intervention, and reduces the necessity for tertiary services (World Health Organization, 2006). The World Health Organization s (2007) Preventing child maltreatment in Europe: a public health approach also recommended a public health approach to protecting children. Finally, the Centers for Disease Control and Prevention (CDC) in Georgia, USA, in Child Maltreatment Surveillance: Uniform Definitions for Public Health and Recommended Data Elements similarly recommended a public health model of child protection (Leeb, Paulozzi, Melanson, Simon, & Arias, 2008). In contrast to the Australian experience however, the international literature has placed greater emphasis on identifying the central role of surveillance in models of public health aimed at protecting children from abuse and neglect (Jack, 2010; Leeb et al., 2008; World Health Organization, 2006). Indeed interdisciplinary public health experts suggest that surveillance is a cornerstone and the foundation of public health practice (Lee, Teutsch, Thacker, & St. Louis, 2010, p. 17). So what is surveillance? According to Webb and Bain: The word surveillance, meaning the constant watching of subversives, came into use during the time of the Napoleonic wars. The modern epidemiological meaning is consistent with the idea of constant watching, but usually of diseases rather than suspects (2011, p. 308). However it is important to point out at the beginning that the word surveillance, as it is used within the public health context, is quite different from the way it is most commonly used within the child protection literature, where vulnerable children and families are seen to be the victims of a surveillance approach that is seen as preventing help and care being given (Lonne, Parton, Thomson, & Harries, 2008; Scott, 2006). Within this literature, it is clear that the constant watching is over individual children and families who are considered to be particularly vulnerable or at risk. For example, Fluke, Shusterman, Hollinshead and Yuan referred to the surveillance of children and families in their research into whether families are subject to greater surveillance in the community when they receive services (2008, p. 78). Scott used the word surveillance in this context when she suggested that children and families who are referred inappropriately to an overloaded child protection system can suffer because: Page 6

7 Parents left suspicious and uncertain about who in their family, neighbourhood or local services may have notified them to the authorities are likely to become anxious and withdrawn. Given that parental stress and social isolation are strong correlates of child abuse and neglect, and given the scale on which child protection surveillance is now occurring in Australia, it is possible that child protection services are contributing to child abuse and neglect by reducing protective factors (2006, p. 11). Parton also used surveillance in this context when he contended that: England is witnessing the emergence of the preventivesurveillance state which aims to intervene earlier in order to ensure that all children develop to their full potential Policies and practices that emphasize such an approach to prevention and early intervention are intimately connected to the need to expand systems of surveillance and information sharing (2008, p. 185). In the context of public health, however, surveillance means something quite different. Whilst it still refers to constant watching, it is the watching of populations rather than individuals. It refers to the collection of data (de-identified data) so that a problem, child abuse and neglect for example, can be understood, effectively responded to, and monitored. A high quality data surveillance system in this context enables the government and community stakeholders to know the magnitude of the child abuse problem, to identify geographic areas or population types that are more affected by the problem than others, to track the incidence (the number of new incidents of child abuse and neglect) and prevalence (the proportion of the population that were abused or neglected as a child) over time, and to understand causes. It informs the setting of priorities, the facilitation of prevention and response programs, and enables the evaluation of those programs. It can also be used as a useful guide when deciding on research to be undertaken (Grimes & Schulz, 2002; Lee et al., 2010; Leeb et al., 2008; Webb & Bain, 2011; World Health Organization, 2006). This report offers a critique of the public health model as it is currently applied to child protection in Australia. It makes the argument that without a strong system for collecting and analysing population - based data associated with child abuse and neglect, the public health approach is doomed to failure. Indeed, without relevant surveillance, any long term changes in the number of children who experience abuse and neglect will not be identifiable. Moreover the factors which may contribute to a genuine reduction in child abuse in Australia will not be understood or replicable. The report is in two parts. In part one, we explain and consider the role of a high quality data surveillance system as an essential component of a public health model of child protection. In part two, we apply these discussions to the Australian context. We then conclude and make recommendations. Page 7

8 Part 1: Surveillance, an essential component of a public health approach Whilst international researchers agree that a public health model of child protection consists of three service platforms primary, secondary and tertiary, many of them also argue that surveillance is an essential component of such an approach (Jack, 2010; Lee et al., 2010; Leeb et al., 2008; World Health Organization, 2006). For example, Leeb and colleagues on behalf of the Centers for Disease Control and Prevention (CDC) in Georgia, USA, in their report Child Maltreatment Surveillance: Uniform Definitions for Public Health and Recommended Data Elements emphasise the importance of surveillance. They argue that: The ultimate goal of CDC s child maltreatment prevention activities is to prevent child maltreatment before it occurs. To do this, CDC uses the public health model in which surveillance is the first step (2008, p. 4). Other international researchers describe surveillance as being the first of four steps. For example Whitaker, Lutzker and Shelley in their Child Maltreatment Prevention Priorities at the Centers for Disease Control and Prevention state: the public health model includes four major activities that inform one another: (a) surveillance to determine the magnitude of the problem, (b) etiologic research to identify risk and protective factors, (c) development and empirical testing of prevention strategies and interventions, and (d) broad dissemination of empirically supported prevention strategies and interventions. The public health model is scientifically based and action oriented; all four activities utilize scientific methodology, with each having the ultimate goal of informing efforts to improve health outcomes (2005, p. 246). Page 8

9 Peden and colleagues, in their World Report on Child Injury Prevention, agree that a public health model is made up of the same: four logical steps - all of which call for good evidence on which to base activities (2008, p. 12). The World Health Organization s (2007) Preventing child maltreatment in Europe: a public health approach also recommended the same four steps. Similarly over a decade ago Professor David Finkelhor wrote: First, we need good epidemiological data to see the location and source of the child abuse problem, and also to be able to track and monitor its response to our efforts. This is something we currently do not have, at least at the level that would satisfy any even generous public health epidemiologist. Second, we need experimental studies to evaluate new and existing practices, so we can agree on what works. Currently, we have practically none, outside of a couple in regard to home visitation and a couple in regard to sexual abuse treatment. There is more experimental science in the toilet paper we use every day than in what we have to offer abused children or families at risk of abuse (1999, p. 969). Scott, Tonmyr, Fraser, Walker and McKenzie have maintained that: The importance of applied research and surveillance in providing data to monitor and evaluate progress in diminishing child maltreatment related mortality and morbidity cannot be understated (sic) (Scott, Tonmyr, Fraser, Walker, & McKenzie, 2009, p. 792). Munro, in her review of the English child protection system also stated that: It is essential that high quality data are used intelligently at local and national levels to drive improvements in practice that benefit children and young people (2011, p. 139). Whilst surveillance has gained very little attention within Australia, O Donnell, Nassar, Jacoby and Stanley (2011) in their series of analyses using linked population data have argued the importance of child maltreatment surveillance. Guthridge, Ryan, Condon, Bromfield, Moss and Lynch (2012) acknowledge the importance of high quality child protection data. Cummins, Scott and Scales (2012) in the Report of the Protecting Victoria s Vulnerable Children Inquiry also recommended an investment of resources into information management systems in order to improve the quality of surveillance. Interdisciplinary public health experts also promote surveillance as being an essential component of public health practice. For example, Webb and Bain noted that surveillance is the eyes and ears of public health (2011, p. 308). Lee and colleagues suggested: Public health surveillance is a cornerstone of public health practice, providing accurate and timely data that are essential to informed decision making and action. Surveillance is the foundation of all public health practice, and we must continue to develop methodologically sound systems that yield highquality, useful data that inform policy and practice(2010, p. 17). Bill Gates, in the Bill and Melinda Gates Foundation annual letter, stated: in the past year I have been struck again and again by how important measurement is to improving the human condition setting clear goals, picking the right approach, then measuring results to get feedback and refine the approach continually helps us to deliver tools and services to everybody who will benefit. (2013, p. 1 & 17). Epidemiology one of the basic sciences of public health Whilst the development of a high quality data surveillance system is the first step toward building an effective public health model of child protection, sound epidemiological principles must be used to design such a system, and accurately interpret the data. According to Lee and colleagues the role of a high quality surveillance system is to guide: epidemiologic research and influence other aspects of the overall mission of public health (2010, p. 17). In fact: epidemiology is considered one of the basic sciences of public health, just as anatomy, physiology, biochemistry, and genetics constitute basic sciences for medicine and nursing (Valanis, 1999, p. 4). The importance of an epidemiological approach to child protection has been highlighted internationally. In 2008, the American Journal of Preventative Medicine devoted a special issue to inflicted traumatic brain injury (TBI), also known as shaken baby syndrome. Butchart in the opening commentary wrote: Page 9

10 By presenting concrete examples of public health approaches to measuring inflicted TBI, it highlights what is arguably the major element missing from the global response to child maltreatment, namely epidemiologically informed methods for monitoring its occurrence (2008, p. 103). Examples of public health surveillance Epidemiological surveillance has made important contributions to health, but none more impressive than smallpox eradication. Surveillance and containment were responsible for the elimination of smallpox from the world, an extraordinary public health achievement. Whereas mass immunization of the world s population had failed, the approach of identification of cases through surveillance and then immunization of susceptible persons in the surrounding communities stopped transmission (Grimes & Schulz, 2002, p. 146). Another example can be found in the identification of the AIDS (acquired immune deficiency) virus in the 1980 s. After investigation it was found that the virus was restricted to certain population groups, with common characteristics. Efforts to manage the spread of the virus were only made possible through effective surveillance systems (Valanis, 1999). By the end of the twentieth century the principles of public health practice were no longer limited to the problem of infectious diseases. The principles had been applied to problems such as chronic disease (such as cancer), behaviorally based health conditions (such as smoking), occupational health (such as the hazards of asbestos for miners and construction workers), road safety (such as traffic accidents), and other types of injury and violence (such as gun laws) (Haddon, 1972; Lee et al., 2010; Whitaker et al., 2005). The source of surveillance data The first task of a high quality and reliable surveillance data system is to provide accurate information about the scale of the problem. Accurately determining the number of children and young people who are the victims of abuse and neglect each year, however, is a challenge. While it is difficult for high income countries such as Australia, the UK, New Zealand, Canada and the USA, it is even more difficult for low and middleincome countries where the large majority of the world s children live, where poverty is common, where there are more wars and disease (for example HIV infection and AIDS), where there are fewer social services, and limited (or in some instances no) legal protections for children. In some parts of the world the problem of child abuse and neglect remains virtually invisible (Butchart, 2008; Lachman et al., 2002). In high income countries that have child protection systems in place, it is difficult to know the scale of the problem because child protection databases are known to underestimate the true incidence of child abuse and neglect, because not all child abuse and neglect is reported to child protection and so will not be represented in the data (Scott et al., 2009). Furthermore, if child protection systems do not have reliable data recording systems then this further limits capacity to know the scale of the problem. The International Statistical Classification of Diseases and Related Health Problems (ICD), the standard system used by hospitals to classify health conditions, also significantly underestimates the true incidence of child abuse and neglect (Scott et al., 2009). This is because many children suffering abuse and neglect do not present to hospital. Examples include children suffering from emotional abuse or neglect, children who are being sexually abused, and children who are physically injured but their injuries remain untreated. Indeed, hospital databases will be biased toward those incidents of serious physical abuse and even then only when the parent or carer takes the child to the hospital (Webb & Bain, 2011). Even when a child is admitted to hospital, a determination about the cause of the child s injuries may not be made at all, or not until after a child protection or police investigation has commenced. In this case the hospital records will be incomplete, and this will influence the way the information is classified, and result in an underrepresentation of the problem (Scott et al., 2009). The police also receive reports in relation to allegations of child abuse and neglect but again experts consider this to be a small proportion of the actual number, a drop in the ocean, because the vast majority of cases go unreported (Newton, Frederick, Wilson, Dibben, & Goddard, 2010, p. 989). Again police reports will not include many cases involving emotional abuse and neglect. Page 10

11 It has been suggested that one way to obtain a more accurate understanding of the magnitude of the child abuse and neglect problem is to use a combination of data from multiple sources rather than from just one source (Schnitzer, Covington, Wirtz, Verhoek-Oftedahl, & Palusci, 2008; Schnitzer, Slusher, & Van Tuinen, 2004; Scott et al., 2009). For example, a data linkage project has been developed in Western Australia which links de-identified data from the Department of Health, Department of Child Protection, and many other departments. To date, the data bases have been used to investigate issues including, the specific child and parental factors associated with increased vulnerability to substantiated child maltreatment (O Donnell, Nassar, Leonard, Jacoby, et al., 2010, p. 921), and whether children presenting to hospitals with suspected maltreatment are being notified to child protection agencies (O Donnell et al., 2011; O Donnell, Nassar, Leonard, Mathews, et al., 2010). In relation to the latter, it was found that only 21% of maltreatment related Emergency Departments presentations had a related notification to Child Protection (O Donnell et al., 2011), and the authors noted that during the study period Western Australia did not have mandatory reporting legislation (O Donnell et al., 2011). Studies such as these are important because they shine light on parts of the system that are not functioning as they should. Statistics also underestimate the number of children who die as a result of abuse and neglect in the USA, the UK, in Australia, and elsewhere (Lamont, 2010a). For example, the determination that a child died as a result of child abuse or neglect is a coronial issue in many jurisdictions and therefore, may or may not appear on a death certificate as the cause of death before the case has been fully investigated (Scott et al., 2009, p. 792). Even when national statistics are collected, there is general consensus that the figures underestimate the size of the problem because some child deaths that are labelled as accidents may actually be attributable to child abuse and neglect (Lamont, 2010a; Newton et al., 2010). Furthermore, in most Australian jurisdictions, child death reviews only concentrate on children who were known to child protection at the time of death or within a limited time frame (Newton et al., 2010). The importance of high quality child protection surveillance data Although there are many obstacles, the challenges of protecting children from abuse and neglect and reducing the incidence of fatal child abuse and neglect still confront us. While a public health approach is one possible way to approach the problem, it must be recognized that integral to this approach is the availability of high quality, reliable surveillance data. We contend that one of the key sources of information that must be enhanced and improved is the child protection database. This is the essential first step toward a public health model of child protection, and reducing the incidence of child abuse and neglect. Whilst child protection databases admittedly underestimate the magnitude of the problem, out of all the surveillance data sources (when used in isolation) they will provide a more accurate estimation of the problem than hospital ICD data or complaints to the police. Indeed, under mandatory reporting, child protection systems should be aware of all child abuse victims who present to hospitals and the police, and many other children who do not. Child protection surveillance data what do we need to know? In order to develop a high quality child protection surveillance system we need to start with the question what do we want to know? (Lee et al., 2010, p. 19; Webb & Bain, 2011, p. 72). As a starting point, accurate data on basic child protection activity is required. For example, information about the number of children notified to statutory child protection systems across Australia each year is required, the number of re-notifications, and the number of notifications that are substantiated. Before proceeding it is important to make some brief comments about definitions. Very generally, any person within Australia, who has concerns about the safety of a child, can make a notification to the statutory child protection authority in his/her State or Territory. However in some jurisdictions this is referred to as a report. In this paper we have chosen to use the term notification so as not to confuse the term with a written report. In relation to the term substantiation, Bromfield and Higgins say that it: refers to notifications (or reports or allegations ) of maltreatment or harm that are found on investigation by a statutory child protection service to have substance (that is, are true) (2005, p. 3). This is a very loose definition. In reality each of the states and territories define and understand the terms report and substantiation very differently. We explore this in further detail in section two of this report. Page 11

12 The important point to be made at this stage is that information about the number of children who are notified to child protection authorities each year is essential. This is the only way that the magnitude of the child abuse problem can be understood. We also need to know the number of re-notifications and substantiations. Further, it is important to know whether some geographical areas or population groups are more affected by the problem than others. It is also important to be able to track the incidence and prevalence of child abuse and neglect over time. All of this information is important because it can provide some clues about the nature and cause of the child abuse problem. It can point the way to some more targeted research, which in turn can be used to inform the design and approach of intervention and prevention strategies, and enable the evaluation of these strategies and efforts. In order to gather the type of detailed information that will be helpful it is useful to turn to the field of epidemiology. According to Valanis the epidemiologist is like a medical detective concerned with the who, what, where, when and how of disease causation (1999, p. 3), the ultimate goal being to prevent or control the occurrence of the disease. According to Grimes and Schulz: Good descriptive research, like good newspaper reporting, should answer five basic W questions who, what, why, when and where (2002, p. 145). The primary author first learnt about the five W s when she worked as an intake worker within Victoria s child protection program, responding to notifications about child abuse and neglect. The five basic W questions were used to guide the initial stages of investigations who, when, what, where and why? Who was involved? This involves questions about the age, sex and ethnicity of the victim and perpetrator, about any particular vulnerabilities or risk factors, and also any witnesses. This also involves questions about the child s and family s support network, both informal and formal. When did it happen? This involves questions in relation to the date and time of the incident, whether anything similar had previously occurred, if so, how many times, and when? What exactly happened? This involves questions about the incident. For example a description of the incident, what was said and done. Was it a punch, a slap, or kick? How many times? Are there or were there injuries, where on the body, what do they look like? Has the Page 12

13 child been medically examined? The same type of questions in relation to previous incidents should also be asked. Where did it happen? This involves questions about the location - for example the home, the street, the shops? Finally, the question about why it happened will be informed by well considered who, what, when and where questions. The five W questions can be used to guide the types of surveillance data that are required to implement a public health model of child protection. With better surveillance data more can be known about the magnitude of the child abuse problem, the location and source of the problem, changes in incidence and prevalence over time, and the effectiveness of prevention and intervention strategies. For example, surveillance data can be established in relation to notifications using the five W questions as follows. Who refers to the children who are the subject of the notification for example their sex, age and ethnicity. Who refers also to the parents - whether they are a single parent, same sex parents, or blended family/ parents, whether there are parental drug or alcohol problems or parental mental illness. Who also refers to type of notifier for example the police, school employee, or parent/guardian. These details are important because they can help answer the final why question. That is, if the final analysis does reveal an increase in notifications over time, then it can be very useful to know the characteristics of the children being notified (are some ethnic groups over represented?), the characteristics of the parents (is there a growing problem of drug and alcohol use?), and the characteristics of the notifiers (is there a rise or change in the number of notifications from one or more professional or community group?). What refers to what the notification is about physical abuse, sexual abuse, family violence, emotional and psychological abuse, or neglect. What also refers to the number of re-notifications. Again, these details are important because they can help answer the final why question. When refers to time. Information that is gathered regularly, over a period of years can reveal trends (whether numbers of notifications are stable, increasing or decreasing). Clearly this information can only be generated by a surveillance system that continuously uses the same operational definitions of child abuse and neglect and the same data reporting methods year in and year out. Where refers to place. This information tells us whether there are more notifications made to child protection about alleged abuse or neglect in some geographical areas than in others. Again, this information can only be generated by a surveillance system where every part of the system uses the same operational definitions of child abuse and neglect and the same data reporting methods. The WHO agrees that for good surveillance, operational case definitions should be clearly set out and agreed upon by the different sectors involved in the data collection (2006, p. 28). Finally, the ability to answer the why question is largely dependent upon the quality of what has gone before. Indeed, a high quality surveillance system that gathers a wealth of who and what data and uses common operational definitions over time (when) and place (where), can provide valuable information about the problem of child abuse and neglect that will assist governments and community organizations as they plan and prioritize prevention and response programs. Further benefits of a high quality data surveillance system Reliable who, what, when and where data might show an association between rising numbers of re-notifications and rising numbers of parental drug and alcohol use in a particular geographical area. Then it might be surmised that there should be an increased investment in drug and alcohol programs (that focus on parents but with a child- centred approach) in that same area. Alternatively the where, when and who data might show an association between rising measures of child abuse and neglect in a particular area, and a rising number of notifications made by a particular professional group in that area and over that timeframe many of which were later determined to be unsubstantiated (more so than for other notifier groups). An association such as this then might point to a need to educate the professional group about child abuse and neglect, mandatory reporting, and when to make a notification. In both of these examples it is of course important not to overstep the data (Grimes & Schulz, 2002, p. 148), and mistake what may be a temporal association for a causal one, or to even to attach our own meanings and ideas to the data which may result in a complete misunderstanding of reality. But these are mistakes that can be avoided by first undertaking some targeted case level research, before implementing strategies (Grimes & Schulz, 2002). Page 13

14 A high quality data surveillance system allows interventions to be evaluated The provision of early intervention and prevention is integral to public health practice. However, in order for a public health system to work effectively, the early intervention and prevention programs must also work effectively. In order to ensure that this is happening, programs must be evaluated. Within the four step model this is the essential third step, and this cannot be done without steps one (knowing the magnitude of the problem through systematic data collection) and two (establishing causes and correlations) being complete. Community - based programs that are run by community service organisations must be evaluated in order to ascertain whether they actually reduce the incidence of child abuse and neglect. New policies, laws and regulations that are introduced and affect whole populations should also be evaluated, to ensure that they are achieving their goals. WHO describes an effective prevention program as being one that reduces the incidence of child maltreatment in the intervention population, or at least lowers the rate at which incidence is increasing (2006, p. 32). The importance of evaluation cannot be overstated. If early intervention and prevention programs are not evaluated comprehensively, and are not effective, it is possible that the result will be an increased need for tertiary level services, without any understanding of the causes. For example, Fluke and colleagues have noted: The frequent association of services with rereporting in the literature raises the question of whether repeated reports of maltreatment result from an increase in the intrinsic risk among certain families that contribute to both service provision and rereports or whether families are subject to greater surveillance in the community when they receive services and are therefore more likely to be rereported (2008, p. 78). Whilst in their conclusion Fluke and colleagues do not resolve whether intrinsic risk or surveillance has primacy regarding the likelihood of subsequent report events (2008, p. 87), and recommend further research in this area, the point is clear. Indeed it is possible that increased primary and secondary support of vulnerable children and families may result in more cases being identified, more reports being made, and more rather than less tertiary intervention. We believe that this will most certainly be the case if the programs are not effective. The type of epidemiological study that can be used to evaluate the effectiveness of policies or laws in reducing the incidence of child abuse and neglect that requires high quality surveillance data is the ecological study. The ecological study is a type of observational study, where the investigator has no control over the intervention and where populations rather than individuals are the unit of analysis (Rivara, 2008; Webb & Bain, 2011). On the negative side, ecological studies can be difficult to interpret because the populations being compared may differ in a number of ways. However, on the positive side, they are inexpensive and easy to do, especially if the data is readily accessible. Then, if strong associations are observed, they can point the way toward some more targeted research (Grimes & Schulz, 2002; Webb & Bain, 2011). An example of an ecological study where populations were compared cross-sectionally (using the where data) is in relation to the impact of gun-control laws on the risk of homicide. Sloan and colleagues made comparisons between Vancouver BC and Seattle WA and found that from the homicide rate in Seattle was 63% higher than the homicide rate in Vancouver and this provided strong evidence for the relationships between handgun regulation and risk of homicide (Rivara, 2008). In relation to the prevention of child abuse and neglect, WHO agreed that ecological studies that compare populations cross-sectionally can be useful and that sometimes the comparison groups can be constructed at the beginning of the program, through some type of matching process (2006, p. 48) or after the fact (2006, p. 48), meaning after the intervention, policy or law has been introduced. Whatever the case, it is important that the two populations are as similar as possible. Then, if there is a reduction in the incidence of child abuse and neglect, it is more likely to be the result of the intervention, policy or law, rather than other factors (World Health Organization, 2006). Gates agrees that one of the benefits of measurement is: the ability it gives government leaders to make comparisons across countries, find who s doing well and then learn from the best (2013, p. 6). As previously stated, it is important not to overstep the data, (Grimes & Schulz, 2002, p. 148) but rather to view these studies as the first scientific toe in the water (Grimes & Schulz, 2002, p. 145), and one that can lead to more targeted research. Another type of ecological study that examines the effect of an intervention is one that makes the comparison over time (using the when data). In this Page 14

15 instance, a policy or law might be introduced in one or more jurisdictions and before and after data compared. This allows for jurisdictions to serve as their own control group (Rivara, 2008). WHO (2006) argued that programs that are developed to address entire populations, such as multimedia campaigns that seek to educate the public about an aspect of child abuse and neglect, can be evaluated by comparing the situation before and after the intervention. In order to compare populations cross-sectionally and over time, it is clearly important to have reliable when and where data. Public reporting of surveillance data It is essential that surveillance data be made available to the public so that it can be widely referred to and used by the public health community to address the problem of child abuse and neglect. Lee and colleagues have maintained that there should be: timely dissemination of these data to those who need to know [because the] final link in the surveillance chain is the application of these data to prevention and control (2010, p. 3). The World Health Organization (2006) agreed that surveillance data should be made widely available so that it can be used to design prevention and response interventions. In essence, it is important that surveillance data is publicly reported because this enables community engagement and collaboration. When surveillance data is available in the public domain, lobby groups are able to use the information to inform their interests and to advocate more effectively for their cause. Researchers and academics are able to use the information to inform their research, and in this way they can more effectively contribute to the child and family welfare knowledge base. Policy makers, those in the community sector, health services and schools, are also able to use this information, to inform, evaluate and improve their work with children and families, and ultimately the outcome will be more effective services for children and families. Page 15

16 Part 2: The Australian Context Unfortunately, in Australia, there is no high quality surveillance data system. It is almost impossible to know the magnitude of the child abuse and neglect problem, because there are no clear and uniform definitions of child abuse and neglect across jurisdictions. It is also almost impossible to know whether incidence of child abuse and neglect is increasing or decreasing because there is no consistent use of definitions over time. For almost twenty years Goddard and his colleagues stressed this point. For example, in 1993, Goddard and Carew stated that within Australia child abuse is defined differently and this meant that any figures must be treated with extreme caution (1993, p. 208). In 1995, Goddard argued that the failure of state governments to agree on almost every aspect of child abuse and child protection is an obstacle in the path of reducing the incidence of childhood physical abuse (1995, p. 13). In 2006 Goddard and Mudaly stated there are no national standards for child protection no national definitions. and no reliable national data (2006, p. 12). In 2006 Liddell, Donegan, Goddard and Tucci expressed their dismay at the continued absence of reliable child protection statistics in many areas of child welfare and protection activity (2006, p. 14). In 2008 Goddard and Tucci stated that in Australia there is a complete lack of reliable national data (2008, p. 9). In 2008 Goddard and Tucci again complained official figures are unreliable even basic measures are not available and to spend so much and to know so little defies belief (2008a, p. 12). In 2011 Goddard and Hunt stated the base data are clearly deficient (2011, p. 414). Cummins, Scott and Scales in the Report of the Protecting Victoria s Vulnerable Children Inquiry concur that: comprehensive and robust data over time to provide the basis for... overarching assessments for the statutory child protection system in reducing the incidence and impact of child abuse and neglect are not available for Victoria or indeed most other jurisdictions (2012, p. 77). Page 16

17 Firstly, there is a complete lack of reliable where data. For more than ten years the Australian Institute of Health and Welfare (AIHW) has released an annual Child Protection Australia report. This report is the main source of publicly available data relating to state and territory child protection systems. However for more than ten years the report has warned that the data from the different jurisdictions are not comparable. As the reports explain, this is because there are major differences between jurisdictions in policies and practices, these differences affect the data provided, and data from different jurisdictions should not be used to measure the performance of one jurisdiction relative to another (Australian Institute of Health and Welfare, 2000, p. 6; 2001, p. 6; 2002, p. 4; 2003, p. 4; 2004, p. 4; 2005, p. 4; 2006, p. 4; 2007, p. 5; 2008, p. 13; 2009, p. 14; 2010, p. 7; 2011, p. 10; 2012, p. viii; 2013a, p. 4). Some of the key differences include: the types of groups that are mandated to notify their concerns about the safety of children vary significantly across jurisdictions (Bromfield & Higgins, 2005). varying definitions of child. For example in New South Wales the mandatory reporting obligation does not include young people aged 16 and 17 years, whereas in many of the other jurisdictions it does (Bromfield & Higgins, 2005). in some jurisdictions mandated reporters are only obligated to report concerns about physical and sexual abuse whilst in other jurisdictions they are obligated to report every type of abuse (Bromfield & Higgins, 2005). in some jurisdictions a notification is caller defined whilst in others the child protection worker together with the Team Leader has the authority to decide whether the alleged events fit within the grounds for a notification (Bromfield & Higgins, 2005). some jurisdictions substantiate the harmful consequences to the child and other jurisdictions substantiate the actions or behaviours of the parent. This means that abuse suffered by a child at the hands of a stranger may result in a notification and substantiation in New South Wales but not in Victoria (Australian Institute of Health and Welfare, 2011, 2013a; Bromfield & Higgins, 2005). There is also a complete lack of reliable when data. For many years the AIHW Child Protection Australia report has warned that comparisons even within many of the states and territories should be treated with caution. This is because many of the states and territories have introduced new data recording systems over recent years and this means comparisons cannot be made with earlier years (Australian Institute of Health and Welfare, 2005, 2010, 2011, 2012). This makes it impossible to identify trends over time. For example, changes were made in the Victorian system in 2007, in South Australia in 2009, in Queensland in 2005 and 2007, in Western Australian in 2006 and 2010, in New South Wales in 2003 and 2010, and in Tasmania in 2003 (Australian Institute of Health and Welfare, 2005, 2010, 2012). There are no reliable data in relation to child protection notifications Queensland provides an example of how unreliable the when data are. In Queensland there were 19,057 notifications in , this peaked at 40,829 in (just prior to introducing new data reporting arrangements), then dropped to 21,655 notifications in (Australian Institute of Health and Welfare, 2009, 2012). The figures have since risen to 24,823 (Australian Institute of Health and Welfare, 2013a). The Child Protection Australia report explains that prior to March 2005: reports that could be responded to by way of protective advice (rather than investigation) were recorded as notifications. This practice ceased from March 2005, and reports dealt with by way of protective advice are now recorded as Child Concern Reports (Australian Institute of Health and Welfare, 2009, p. 23). In other words, there was a change in data recording methods with a significant impact on the data. The New South Wales data provide another example. The Child Protection Australia reports reveal that after 2003 (the time of the first system change) the New South Wales notification rates rose significantly. The Child Protection Australia report states: The number of child protection notifications, investigations and substantiations in New South Wales in differs significantly from the numbers in previous years. This difference is a direct result of changes to the Department of Community Services client information system which were implemented to reflect amendments to legislation and associated practice changes. For this reason, New South Wales Child Protection data for this year is not comparable with the data for previous years published in this report (Australian Institute of Health and Welfare, 2004, p. 12). Page 17

18 Over the next seven years, the New South Wales notification rates soared. In fact, when compared to the notification rates in other jurisdictions, the New South Wales figures were disproportionally high. Whilst the population of New South Wales represents approximately 32% of the Australian population (Australian Bureau of Statistics, 2012), from until , notifications in New South Wales accounted for more than 50% of total national notifications (Australian Institute of Health and Welfare, 2005, 2011). In fact, in the and years the New South Wales figures accounted for more than 60% of the total notifications in Australia (Australian Institute of Health and Welfare, 2011). These figures, or course, influenced the overall picture, so much so that national notifications almost tripled between and (Australian Institute of Health and Welfare, 2005, 2011). However, in 2010, the New South Wales Government introduced another change to their system, and again this affected the data, but this time by reducing the numbers. The Child Protection Australia report states New South Wales figures are not comparable with those of other jurisdictions (Australian Institute of Health and Welfare, 2011, p. 14) and the data reported for part of reflect legislative changes data are not comparable to previous years (Australian Institute of Health and Welfare, 2011, p. 120). Consequently in and the number of notifications recorded for New South Wales dropped significantly, and national data were again dramatically affected (Australian Institute of Health and Welfare, 2012). For the first time in more than a decade Australia had a decrease in notifications: a decrease of 16% in (Australian Institute of Health and Welfare, 2011, p. 16) and of 17% in (Australian Institute of Health and Welfare, 2012, p. 17). In , however, notifications increased by 7% (Australian Institute of Health and Welfare, 2013a, p. 19). A report released by the New South Wales Department of Community Services has commented that: It is of interest to compare child protection data over time and across jurisdictions. However, issues of data comparability such as jurisdictional differences and changes in the data over time can result in inaccurate and potentially misleading conclusions (Zhou & Chilvers, 2007, p. 3). It is not just out of interest that we seek to compare data over time and place it is crucial to do so. It is essential to have the ability to compare the incidence of child abuse and neglect across jurisdictions (the where data), and it is essential to have the ability to monitor changes within jurisdictions (the when data). In relation to notifications about child abuse and neglect, it is crucial to know whether they are increasing or decreasing nationally, and although many researchers suggest that notifications have increased (Allen Consulting Group, 2008; O Donnell et al., 2008; Scott, 2006), without reliable data it is impossible to claim this. There are no reliable data in relation to child protection substantiations There are also no reliable figures in relation to the types of abuse or neglect that are substantiated by the different child protection authorities in Australia. For example, in substantiations in relation to physical abuse ranged from 14% in Tasmania to 31% in Victoria, substantiations in relation to emotional abuse ranged from 19% in Western Australia to 52% in Victoria, and substantiations in relation to neglect ranged from 8% in Victoria to 49% in the Northern Territory (Australian Institute of Health and Welfare, 2012, p. 8). Possibly the most startling variation in substantiations is in relation to sexual abuse where it ranged from 5% in the Northern Territory to 23% in Western Australia. Indeed it is very concerning that three years after the commencement of the Federal Government Northern Territory Intervention, which was targeted at the significant problem in Northern Territory Communities in relation to sexual abuse of children (Northern Territory Board of Inquiry, 2007, p. 6) there appear to be significantly fewer substantiated cases of child sexual abuse in the Northern Territory than other parts of the country. But of course the data is completely unreliable. Indeed the Child Protection Australia report concludes that the differences may be partially due to the varying mandatory reporting requirements across jurisdictions (Australian Institute of Health and Welfare, 2012, p. 8). However, the same report states that: Although there are differences between states and territories that affect the comparability of the data on children in out-of-home care and on care and protection orders, the differences between jurisdictions are greatest in relation to child protection notifications, Page 18

19 investigations and substantiations (Australian Institute of Health and Welfare, 2012, p. 96). The first comment implies that children in some parts of Australia are being abused, but because of differences in mandatory reporting requirements these children are not coming to the attention of child protection, their abuse is not being substantiated, and therefore they are not being protected. However the latter comment reminds us that the data is not comparable, and therefore we cannot know the reasons for these differences. As Liddell, Donegan, Goddard and Tucci have noted it is not obvious that such differences can be explained simply on the basis of procedural differences between the jurisdictions (Liddell et al., 2006, p. 15). This is a very serious problem indeed. In relation to substantiations there is also no reliable when data in other words, the new data recording systems that have been introduced by many of the states and territories have affected the reliability of the data. Again these impacts have been most noticeable in New South Wales and Queensland. For example, in New South Wales the trend in relation to substantiations has been very similar to the trend in relation to notifications. There were 16,765 substantiations in , this peaked at 37, 094 in , then dropped to 18,596 in (Australian Institute of Health and Welfare, 2008, 2012), and increased to 23,175 in (Australian Institute of Health and Welfare, 2013a). Again these New South Wales figures have influenced the overall picture. In Australia substantiations of child abuse and neglect increased in for the first time since (when it reached a peak of 60,230) (Australian Institute of Health and Welfare, 2011). Even though the media statement related to the most recent report announced that: substantiated child abuse and neglect has risen in Australia.. over the past 12 months, the number of children who were the subject of substantiated abuse increased from 31,500 to 37,800 (Australian Institute of Health and Welfare, 2013b). There has been very little concern expressed by the media, politicians or academics and why would there be when the data are completely unreliable. There are no reliable data in relation to the ethnicity of children, or the characteristics of the parents coming to the attention of child protection authorities In Australia there is also a lack of detailed who and what data. Again we refer to notifications to illustrate our point. In Australia, there is no information about the number of children and young people from culturally and linguistically diverse families notified to child protection (Cummins et al., 2012; Kaur, 2009). Even the data in relation to indigenous children is unreliable. This is confirmed by the Child Protection Australia report which states that: The practices used to identify and record the indigenous status of children vary across states and territories, with some jurisdictions recording large numbers of unknowns. No state or territory can validate the data on Aboriginal or Torres Strait Islander children by other means and the quality of the data is therefore unknown (Australian Institute of Health and Welfare, 2012, p. 82). The AIHW report states that: Over the last few years, several jurisdictions have introduced measures to improve the identification of Indigenous clients. However, in some jurisdictions, the high proportion of children whose Indigenous status is unknown still affects the quality of data on Aboriginal and Torres Strait Islander children. Improvements to the quality of Indigenous identification in the national notifications, investigations and substantiations data set have affected the comparability of these data prior to (Australian Institute of Health and Welfare, 2013a, p. 7). Whilst it is positive that some attempts are being made to improve data collection and reporting in relation to indigenous clients, it is crucial to note that if improvements are not made by all jurisdictions then comparisons will still not be possible over time for many years to come, and comparisons may never be possible across place. Page 19

20 There is no information about the percentage of notifications where there is parental drug and alcohol use, mental illness or family violence, or the number of notifications that are re-notifications, (for the first time, or how many times children have been re-notified) (Cummins et al., 2012). The Child Protection Australia report acknowledges that there has been no way of: determining how many children appear within the system on multiple occasions. To help overcome these and other limitations generated by reporting from aggregate data from national child protection data will be collected and analysed at the unit record (child) level (Australian Institute of Health and Welfare, 2013a, p. 139). Only time will reveal the value of these changes. There are no useful data in relation to the source of the notification Whilst there is information about the type of notifier (for example police, school employee, or parent/guardian) for investigated cases, the same information is not provided for total notifications, or for substantiated cases. This means that whilst there is information about the type of notifier (or source of the notification) for the 127,759 cases that were investigated in the year, it is not possible to know the type of notifier (or source of the notification) for the 237,273 notifications, or for the 40,466 substantiated cases in that same year (Australian Institute of Health and Welfare, 2012, pp 6, 7 & 48). This information might be very useful if, for example, it were shown that particular reporter groups were making comparatively high numbers of notifications that resulted in comparatively low numbers of investigations and substantiations. Anecdotal evidence, for example, suggests that high numbers of reports are being made by non-custodial parents, and that this might currently be as high as 30% of the total number of reports made in Victoria each year. In some instances, the non-custodial parent makes a false notification in the hope that child protection will remove the children from the other parent (the custodial parent) and grant custody to him or herself. The child protection system is viewed by the notifier as being a cost free alternative to Family Law proceedings. Some non-custodial parents also make false notifications to exact revenge upon their former partners. This claim is somewhat supported by ten year old research material on the Melbourne Children s Court website where it is reported that 17% of notifications made to child protection in were made by a parent/guardian (Children s Court of Victoria, 2012). However the Child Protection Australia report of the same year stated that only 9% of finalized investigations had parent/guardian listed as the source of the notification (Australian Institute of Health and Welfare, 2001, p. 48). In essence, this means that in , in Victoria, 17% of reports were made by parent/guardian, with a large number of them being closed at the intake phase and not investigated. The only other notifier groups showing a similar pattern for this same year were friends/ neighbours (they made 8% of the notifications and 6% that resulted in a finalized investigation), anonymous (they made 5% of the notifications but none that resulted in a finalized investigation), and other (they made 6% of the notifications and 5% that resulted in a finalized investigation). Unfortunately these data are the only available data in the public domain, so it is not possible to ascertain whether this phenomenon (high numbers of notifications made by parent/ guardians resulting in a low proportion of finalized investigation) has increased over the past ten years. If this is so, it may indicate that more and more non custodial parents wanting residency of their children are willing to make false notifications, because they view child protection intervention as a better alternative to costly family court action. If this is the case, it is essential to know, because on the one hand malicious and vexatious reports will further overload the system. On the other hand, it is possible that some notifications that are actually true may not be taken seriously. Serious questions must also be asked about whether some children in these families are suffering emotional abuse, parents in some instances using their children to exact revenge upon former partners. This is a problem worthy of closer examination. The only way to ascertain this is through high quality, reliable surveillance data. The lack of reliable surveillance data limits the ability of the Australian Government and public health community to effectively respond to the problem of child abuse and neglect In Australia, the Council of Australian Governments have developed the Protecting Australia s Children is Everyone s Business: National Framework for Protecting Australia s Children (here on referred to as the National Framework) which aims to achieve the high level outcome Australia s children and young Page 20

21 people are safe and well (Council of Australian Governments, 2009, p. 11). To measure this outcome governments and the non - government sector have set the following target a substantial and sustained reduction in child abuse and neglect in Australia over time (Council of Australian Governments, 2009, p. 11). The National Framework is based on a public health model of child protection. It states: A public health model offers a different approach with a greater emphasis on assisting families early enough to prevent abuse and neglect occurring. It seeks to involve other professionals, families and the wider community enhancing the variety of systems that can be used to protect children and recognizing that protecting children is everyone s responsibility (Council of Australian Governments, 2009, p. 8). This is a worthy goal but the lack of reliable data limits the ability of the public health community to effectively work toward achieving it (Leeb et al., 2008). Goddard and Hunt put it this way: the various governments have endorsed the National Framework for Protecting Australia s Children (Council of Australian Governments, 2009) which aims to achieve a substantial and sustained reduction in child abuse and neglect (AIHW, 2011, p.6), even though the base data are clearly deficient. The National Framework also emphasizes a public health approach thus producing a further contradiction. Such an approach requires transparent data, for if you don t know, you can t act (Krieger, 1992, p 412). This failure to collect compatible data is but one example of the history of silencing children and denial of abuse. (2011, p. 414) Firstly, the lack of reliable national where data makes it impossible to know which children, in which areas, are suffering more abuse and neglect. If the public health community knew this information then they could target their services toward these more needy areas. To tell the public health community that it is their responsibility to respond to the problem of child abuse and neglect, but not equip them with the information to carry out this mission is at best a contradiction, at worst sabotage. Not only has the Council of Australian Governments in their National Framework given the Australian public health community this impossible task, but so have many of the individual states and territories. For example: Page 21

22 The Australian Capital Territory s Sharing Responsibility: A Framework for Service Collaboration for the Care, Protection and Wellbeing of Children and Young People in the ACT states that: o The care and protection of children and young people is a shared responsibility. No one agency has the knowledge, skills or opportunities to respond entirely to the needs of children and young people who suffer abuse or neglect. An effective response requires the cooperation and coordination across agencies (Vardon and Murray Reports Steering Committee, 2005, p. 12). The Northern Territory s Safe Children Bright Futures Strategic Framework names child protection, other government agencies, the non-government sector and the community as being partners in change (Department of Children and Families, 2011, p. 5). Victoria s Vulnerable Children Our Shared Responsibility states: o The Victorian Government has a key role to play in protecting vulnerable children. However, protecting children is a community wide responsibility (and) in line with this shared responsibility, organisations should strive to provide connected services and get those services to those who most need them. Tailored, flexible approaches must be available for groups or communities with specific needs or preferences, including Aboriginal children and their families (Victorian Government, 2012, p. 3). And, finally, the Western Australia s Supporting Individuals and Families at Risk or in Crisis Strategic Framework and State Plan states that responding effectively to issues requires a shared responsibility between government agencies (p 6). All agencies have a responsibility to be aware of the evidence that exists in their areas of specialisation, and to ensure that this informs the development of their services and interventions (Department for Child Protection, 2009, p. 19). The lack of reliable surveillance data makes it impossible to know whether early intervention and prevention efforts are effective The lack of reliable national data makes it impossible to monitor the effectiveness of any early intervention and prevention efforts. As Goddard and Tucci (2012) have argued, such data provide essential measures of effectiveness. Indeed one of the most basic questions to be asked about the National Framework, and other state and territory frameworks and plans, is have they improved the safety of children? In order to find answers it is necessary to have reliable when data. This is the only way that the goal of a substantial and sustained reduction in child abuse and neglect can be measured. The first Annual Report to the Council of Australian Governments (2010) on implementing the National Framework (here on referred to as the Annual Report) acknowledges that this is an important question. However it also acknowledges that: An ongoing and significant challenge for the Performance and Data Working Group (PDWG) is to specify relevant and feasible indicators of change. This is an ongoing task and an important area to establish in the early phase of the implementation (Council of Australian Governments, 2010, p. 2). The Second Annual Report (here on the Second Annual Report) does not make this same acknowledgement. Ironically it points to the Child Protection Australia report and the deficient data as being one of the main measures to track progress (Council of Australian Governments, 2012, p. 6). Of the 28 indicators of change that are being used to assess progress in achieving the goal, 11 of them are listed as requiring development not reported in this document (Council of Australian Governments, 2010). Whilst the other 17 indicators are reported there are deficiencies within most of them (refer to Appendix One for list of the 28 indicators). The main deficiency in relation to these 17 indicators is that many of them rely on data that are unreliable, as we have noted in detail above. Of the 17 indicators, those that rely on unreliable data are: indicator 7, rate of child protection notifications, indicator 24, rate per 1000 indigenous children with substantiated cases compared to other children, indicator 25, rate per 1000 indigenous children in out of home care compared with other children, indicator 26, proportion of indigenous children placed in accordance with the indigenous child placement principles, and indicator 28, number and rate of children in substantiations by abuse type, Page 22

23 These are all based on data from the Child Protection Australia report that are unreliable. Of the 17 indicators, still others are deficient. For example: Indicator 13, proportion of children 4 14 years with mental health problems. The Annual Report itself recommends that this be interpreted with caution (due to the fact that the data were collected in 1998). The second Annual Report states no current data source (Council of Australian Governments, 2012, p. 124). Indicator 21, proportion of investigations finalised by time taken to complete investigation. Again the Annual Report recommends these data be treated with caution, due to the fact that there are a number of factors that affect the time taken to complete an investigation and many of which are outside of the control of the department responsible (Council of Australian Governments, 2010, p. 62). We also suggest that whilst timeliness is undoubtedly an important performance measure, it may not directly be related to a reduction in the incidence of child abuse and neglect. Indeed, child protection workers could be very timely in their responses to reports of child abuse and neglect, but this in itself does not tell us anything about the quality of work that is done, or the decisions that are made. Indicator 20, retention rate of foster carers and child protection workers. This is related to the number of households entering and exiting foster care (Council of Australian Governments, 2010, p. 59). This indicator is deficient for a number of reasons. Firstly, because it is not related to retention rates of child protection workers. Secondly the New South Wales and Queensland data are not included. Thirdly, the Child Protection Australia report itself acknowledges that state and territory differences in policies and practices in relation to foster care and relative/ kinship care should be taken into account when interpreting the data (Australian Institute of Health and Welfare, 2012, p. 40). Finally, and most importantly, there is a weak relationship between the indicator and the ultimate goal of a substantial and sustained reduction of child abuse and neglect. Of the 17 indicators, still others are a measure of childhood wellbeing rather than abuse and neglect. For example: Indicator 6, rate per babies born with low birth weight. Whilst it is clear that this indicator is related to wellness it certainly does not follow that a decrease in the rate of babies born with low birth weight means that there has been a reduction in the incidence of child abuse and neglect. Some healthy babies are abused and neglected, and many underweight babies experience high levels of nurture and care. Indicator 10, proportion of communities with improved measures against the Australian Early Development index (AEDI). The report explains that the AEDI is a population measure of young children s development, measuring five domains including physical health and wellbeing, social competence, emotional maturity, language and cognitive skills, and communication skills and general knowledge (Council of Australian Governments, 2010). Again, whilst this indicator is related to child wellbeing, it is not related to a reduction in the incidence of child abuse and neglect. Similarly indicator 18, number of children living in jobless families. Whilst children living in jobless families clearly suffer a range of disadvantages, and whilst joblessness is undoubtedly a risk factor for child abuse and neglect, a reduction in the number of unemployed parents is not directly related to reducing the incidence of child abuse and neglect in Australia. Indeed we know that many jobless parents do not abuse their children, while some working parents do. Finally, in other indicators, it is not clear whether an increasing or decreasing trend will lead toward a reduction in the incidence of child abuse and neglect. For example: Indicator 8, number of at risk children and families accessing support services. This is reported as being a proxy indicator of the extent to which children and families are receiving the support they need to enable children to live safely in their home environment, with their family, without the need for out of home care (Council of Australian Governments, 2010, p. 45). The Annual Report states that: o In ,544 children aged 0-17 years started intensive family support services. Of these children, almost half (46 per cent) were aged 0-4 years. Across Australia the number of children commencing intensive family support services has increased from 2651 in (Council of Australian Governments, 2010, p. 45). We question whether an increasing or decreasing number of at risk children and families accessing support services is desirable. Whilst the report indicates that an increase is desirable (because it will lead to children and families receiving the support they need to enable children to live safely at home, without the need for out of home care), there is also the possibility that this may actually result in more children remaining Page 23

24 in abusive and neglectful families (especially if the support services are not effective). In this case, this would not result in a substantial and sustained reduction in child abuse and neglect (Council of Australian Governments, 2009, p. 11), but the opposite. Indicator 14, rate per 1000 children accessing assistance through homelessness services Again, we question whether an increase or decrease in numbers is desirable. Whilst the Annual Report suggests that a rise in these numbers is positive and indicates that more children and families are receiving the support they need, a rise in numbers could also be viewed as a concern. On the one hand, if rising numbers of children accessing homelessness services reflect that there are fewer numbers of children living on the streets, this is better for children. However, if rising numbers indicate that the homelessness problem is worsening overall, and there is the same proportion of children and young people living on the streets, this is clearly bad for children. The problem with such crude data is that it tells us very little. The states and territories As noted above, many of the Australian states and territories also have their own frameworks for change. The Australian Capital Territory s Sharing Responsibility: A framework for service collaboration, Western Australia s Supporting Individuals and Families at Risk or in Crisis Strategic Framework and State Plan , and Victoria s Vulnerable Children Our Shared Responsibility. However, there is not any mention within any of these documents about how, or even if, progress will be evaluated. The Northern Territory s Safe Children, Bright Futures Strategic Framework , lists seven areas of reform, one of which is keeping kids safe (Department of Children and Families, 2011, p. 8). The authors of the report ask the question How will we measure our success? (Department of Children and Families, 2011, p. 7). Their answer (in part) is that the Child Protection External Monitoring and Reporting Committee will work with the Department of Children and Families to develop a performance measurement and evaluation framework to measure the success of reforms to the child protection and child and family support systems, and to evaluate specific initiatives undertaken as part of the reforms (Department of Children and Families, 2011, p. 7). The authors of the report also state that Government will work with indigenous organizations and communities to develop new measures that are meaningful for indigenous children and families, and will involve non-government organizations and communities in the design and process of evaluating programs and services (Department of Children and Families, 2011, p. 7). We suggest that rather than developing new measures it would be preferable to have reliable basic measures. It would be better for example, to have uniform definitions of child abuse and neglect across place and across time, so that the effectiveness of interventions can be evaluated. Then there would be no debate about whether the extreme measures undertaken by the Federal Government in implementing the Northern Territory Intervention, to address the significant problem of child abuse and neglect that is facing so many indigenous communities, have been effective or not. The answers would be in the data. Good data provides protection. Good data needs competent systems of surveillance. Page 24

25 Conclusion In Australia, many researchers and policy makers have called for a public health model of child protection and this is now shaping the policy framework nationally. Integral to a public health model, however, are four vital steps. The first is surveillance, to understand the size of the child abuse and neglect problem, the second is to identify risk and protective factors, and risk populations, the third is to develop and evaluate prevention and intervention strategies, and the fourth is to disseminate information about the effectiveness of intervention activities to the public health community, to enable widespread adoption of programs and policies that work. These steps rely on data that is comparable over place and time, and data that provides important information about the characteristics of children and families who come to the attention of statutory child protection services. In Australia, high quality data are not available, and no attempts have been made to collect this data. Yet a public health approach is still repeatedly espoused, and the public health community is still given the shared responsibility for protecting children. Astoundingly, a National Framework for Protecting Australia s Children has been implemented, with annual reporting of progress, but with no real ability to measure effectiveness. Experts continue to claim that child protection systems in Australia are overburdened with notifications that should not have been made in the first place. For example as recently as April 2013, Alan Hayes, the Director of the Australian Institute of Family Studies accused Australians of over-reporting, caving to moral panic in their attitudes to child protection (ABC Radio National, 2013), and used the completely unreliable data contained in Child Protection Australia reports as evidence. On 12 November 2012, Prime Minister Julia Gillard announced a Royal Commission into child sexual abuse within religious groups, sporting organizations, schools, welfare agencies and the like. The Prime Minister is reported as saying there has been a systemic failure in the response to child abuse that required a national solution (Australian Politics, 2012). We agree that when problems are serious, widespread, systemic, and seemingly intractable, a national solution requires a national approach. Page 25

26 For the same reasons we argue for a national approach to the problem of child abuse and neglect. This will require a new structure, a new system that involves all the states and territories employing the same child protection legislation and practices, the same operational definitions of child abuse and neglect over time, and the same data recording methods. There must be a shared agreement about the types of information that is needed (for example information about parental substance misuse, mental illness, disability, family violence, family ethnicity) and common definitions of each type. For the sake of children s safety, it is essential that the work is urgently undertaken, and essential to invest to ensure it is done accurately (Gates, 2013). Australian children need a response from Commonwealth and State Governments that moves beyond a tokenistic and partial (Goddard, Mudaly, & Frederick, 2011) approach to child protection. This can only be based on high quality, comparable, and reliable data that will inform future policy and practice. This is the first step toward a truly childcentred, national framework for protecting children. Child protection data must count for something. It needs to provide the knowledge base that will inform the development of effective policies and programs that stop child abuse. Page 26

27 Recommendations Recommendation 1. We recommend that a unified child protection system be created across Australia. States and territories should have the same child protection legislation and practices, the same operational definitions of child abuse and neglect, and the same data reporting methods. Recommendation 2. We recommend that the type of information collected be expanded so that the true extent and scale of the problem of child abuse and neglect in Australia is understood. It should include the current available information as well as the following data: Information about the children, including sex, age and ethnicity. Information about the families, including composition of household, presence of parental alcohol or drug use, mental health problems, criminal history, and previous child protection involvement. Information about the occurrence of family violence, including the number of notifications that are made by the police to child protection. Information about the source of the notification, the number of notifications that led to an investigation by type of notifier, the number of notifications that were substantiated by type of notifier, and the number of notifications that ledto a children s court intervention by type of notifier. Information that tracks the type of abuse notified and substantiated, into the child protection system, health, and criminal justice systems. Information about notifications that were not substantiated but would fit a new category of possible or probable abuse and neglect (for example, suspected child sexual abuse). These data are critical in public health surveillance where prevention is the goal, and the legal process of finding probable cause is much less important (Schnitzer et al, 2004, p. 383). Information about the number of times children are re-notified to child protection and the time frame in which second and subsequent reports were made. Recommendation 3. We recommend that the sources of information be extended beyond data provided by state-based child protection departments to include the following: Data relating to the number of notifications made by the Family Court of Australia to child protection authorities involving allegations of sexual or physical abuse, and the outcome of these notifications. Data relating to physical assault and sexual assault crimes against children (number of criminal charges and convictions) and injuries (number of hospital presentations and admissions). The number of child deaths that are a result of child abuse and neglect, and full de-identified details of the victims. Recommendation 4. We recommend that the Commonwealth Government show leadership on behalf of Australia s children and establish and adequately resource an independent National Child Protection Inspectorate: To collate all relevant child protection data to ensure that the true extent and nature of the problem of child abuse in Australia is understood: and, Regularly inspect, audit and review the effectiveness of all children s services, child protection systems and out-of-home-care. Page 27

28 Appendix One: Child Protection Australia reports stating that much of the data they report on is worthless For more than ten years the Australian Institute of Health and Welfare (AIHW) has released an annual report containing information relating to state and territory child protection systems. Every year the report acknowledges that there are major differences between jurisdictions in legislation, policy and practice. Every year the report acknowledges that much of the data is not comparable, and that comparisons between states and territories should be treated with caution. Child Protection Australia is the sixteenth annual comprehensive report on child protection. Although the processes used by each jurisdiction to protect children are broadly similar (AIFS; Bromfield & Higgins, 2005), there are some important differences between jurisdictions child protection policies and practices that should be taken into account when making comparisons across jurisdictions (p 4) Child Protection Australia is the fifteenth annual comprehensive report on child protection. The report provides detailed statistical information on state and territory child protection and support services, and some of the characteristics of the children receiving these services. In Australia, child protection is a state and territory government responsibility, and there are significant differences in how each deals with and reports child protection issues. These differences should be taken into account when making comparisons (AIHW, 2012, p viii) Child Protection Australia is the fourteenth annual comprehensive report on child protection. The report provides detailed statistical information on state and territory child protection and support services, and some of the characteristics of the children receiving these services. In Australia, child protection is a state and territory government responsibility, and there are significant differences in how each deals with and reports child protection issues. Statistical comparisons between states and territories should therefore be treated with caution. (AIHW 2011 p vii) Although the processes used by each jurisdiction to protect children are broadly similar protection policies and practices that affect the data presented in this report. The data from jurisdictions are therefore not strictly comparable and should not generally be used to measure the performance of one jurisdiction relative to another. (AIHW 2011 p 10) Child Protection Australia is the thirteenth annual comprehensive child protection report. The report provides detailed statistical information on state and territory child protection and support services, and some of the characteristics of the children within these systems. In Australia, child protection is a state and territory government responsibility, and there are significant differences in how each deals with and reports child protection issues. Statistical comparisons between states and territories should therefore be treated with caution. (AIHW 2010 p vii) Although the processes used by each jurisdiction to protect children are broadly similar (Bromfield & Higgins 2005), there are some important differences between jurisdictions in policies and practices in relation to child protection, and these differences affect the data presented in this report. The data from jurisdictions are therefore not strictly comparable and should not generally be used to measure the performance of one jurisdiction relative to another (AIHW 2010 p 7), Child Protection Australia is the twelfth annual comprehensive child protection report. With the many differences in the way each state or territory handles and reports child protection issues, one must interpret relevant statistical information with caution. (AIHW 2009 p viii) Page 28

29 Although the processes used by each jurisdiction to protect children are broadly similar (Bromfield & Higgins 2005), there are some important differences between jurisdictions in policies and practices in relation to child protection, and these differences affect the data presented in this report. The data from jurisdictions are therefore not strictly comparable and should not be used to measure the performance of one jurisdiction relative to another (AIHW 2009 p 14). Child Protection Australia is the eleventh annual comprehensive child protection report. With the many differences in the way each state or territory handles and reports child protection issues, one must interpret relevant statistical information with caution. (AIHW 2008 p x) Although the processes used by each jurisdiction to protect children are broadly similar (Bromfield & Higgins 2005), there are some important differences between jurisdictions in policies and practices in relation to child protection, and these differences affect the data presented in this report. The data from jurisdictions are therefore not strictly comparable and should not be used to measure the performance of one jurisdiction relative to another (AIHW 2008 p 13). Child Protection Australia is the tenth annual comprehensive child protection report. Each state and territory has its own legislation, policies and practices in relation to child protection, which accounts for some of the differences between jurisdictions in the data provided. Australian totals have not been provided for those data that are not comparable across the states and territories (AIHW 2007 p x). There are some important differences between jurisdictions in policies and practices in relation to child protection, and these differences affect the data provided. The data from different jurisdictions are therefore not strictly comparable and should not be used to measure the performance of one jurisdiction relative to another (AIHW 2007 p 5). Child Protection Australia is the ninth annual comprehensive child protection report. Each state and territory has its own legislation, policies and practices in relation to child protection, which accounts for some of the differences between jurisdictions in the data provided. Australian totals have not been provided for those data that are not comparable across the states and territories (AIHW 2006 p xii). There are some important differences between jurisdictions in policies and practices in relation to child protection, and these differences affect the data provided. The data from different jurisdictions are therefore not strictly comparable and should not be used to measure the performance of one jurisdiction relative to another (AIHW 2006 p 4). Child Protection Australia is the eighth annual comprehensive child protection report. Each state and territory has its own legislation, policies and practices in relation to child protection, which accounts for some of the differences between jurisdictions in the data provided. Australian totals have not been provided for those data that are not comparable across the states and territories (AIHW 2005 p xiii). There are some important differences between jurisdictions in policies and practices in relation to child protection, and these differences affect the data provided. The data from different jurisdictions are therefore not strictly comparable and should not be used to measure the performance of one jurisdiction relative to another (AIHW 2005 p 4). Child Protection Australia is the seventh annual comprehensive child protection report. Each state and territory has its own legislation, policies and practices in relation to child protection, which accounts for some of the differences between jurisdictions in the data provided. Australian totals have not been provided for those data that are not comparable across the states and territories (AIHW 2004 p xiii). There are some major differences between jurisdictions in policies and practices in relation to child protection, and these differences affect the data provided. The data from different jurisdictions are therefore not strictly comparable and should not be used to measure the performance of one jurisdiction relative to another (AIHW 2004 p 4). Child Protection Australia is the sixth annual comprehensive child protection report. There are some major differences between jurisdictions in policies and practices in relation to child protection, and these differences affect the data that are provided. The data from different jurisdictions are therefore not strictly comparable and should not be used to measure the performance of one jurisdiction relative to another (AIHW 2003 p 4). Child Protection Australia is the fifth annual comprehensive child protection Report. There are some major differences among jurisdictions in policies and practices in relation to child protection, and these differences affect the data that are provided. The data from different jurisdictions are therefore not strictly comparable and should not be used to measure the performance of one jurisdiction relative to another (AIHW 2002 p 4). Page 29

30 Child Protection Australia is the fourth annual comprehensive child protection report. These data are collected each year by the Australian Institute of Health and Welfare (AIHW) from the community service departments in each State and Territory. The data in this report cover the financial year. Each State and Territory has its own legislation, policies and practices in relation to child protection, so there are differences between jurisdictions in the data provided. Australian totals have not been provided for those data that are not consistent across the states and territories (AIHW 2001 p 1). There are some major differences among jurisdictions in policies and practices in relation to child protection, and these differences affect the data that are provided. The data from different jurisdictions are therefore not strictly comparable and should not be used to measure the performance of one jurisdiction relative to another (AIHW 2001 p 6). Child Protection Australia is the third annual comprehensive child protection report. As noted, each jurisdiction has its own legislation, policies and practices in relation to child protection. There are differences between the jurisdictions in these areas and these differences affect the data that are provided. The data from different jurisdictions are therefore not strictly comparable and should not be used to measure the performance of one jurisdiction relative to another (AIHW 2000 p 6). Page 30