The Australian Child Maltreatment Study: National prevalence and associated health outcomes of child abuse and neglect

In reply: We welcome the correspondence by Segal and Gnanamanickam1 regarding the Australian Child Maltreatment Study (ACMS),2 and their endorsement of the need to understand the extent of child maltreatment. The authors pose the question of where society should set the threshold for defining child maltreatment, particularly to identify those most in need of support, and highlight the need for prudent judgements about the circumstances justifying policy and service responses by child protection systems.

While the definition and extent of statutory child protection concerns are important questions, they did not underpin the design of the ACMS, or its main findings and recommendations. It is essential to distinguish between statutory child protection systems and technical legislative definitions of “child maltreatment” engaging their operation, and epidemiological studies designed to measure population‐wide prevalence and characteristics of child maltreatment, and associated health outcomes. Child protection systems are created by governments, regulated by complex legislative and policy frameworks, primarily being focused on tertiary responses — not to all child maltreatment but to designated significant levels of child maltreatment, generally both after the event and where there is no protective parent. These systems, and their definitions of child maltreatment warranting formal state intervention, are political artefacts. Between jurisdictions, and at different points in time, they bear variable connection to scientific evidence, bioethical principles, lived experience, and clinical need.

In contrast, being a comprehensive epidemiological study, the ACMS has a different purpose. Driven by public health framing, the ACMS measured the national prevalence of five types of child maltreatment, their associated mental disorders and health risk behaviours through life, and associated burden of disease. As we showed,3 such studies need to be driven by robust definitions of each child maltreatment type, derived not from child protection statutes, but from the best consensus of decades of theoretical, conceptual and empirical analyses.

The ACMS approach to defining and measuring child maltreatment was conservative. While soundly based in rigorous conceptual models,4 operational examples were narrower than many used internationally.3,4 We applied chronicity thresholds,5 and we elected not to measure some subdomains of maltreatment types at all, including several acknowledged in child protection statutes.

The ACMS identified the prevalence of each of the five maltreatment types, and differential prevalence by gender and age group.5 It identified associations between any maltreatment, and differential impacts of specific maltreatment types, and mental disorders and risk behaviours.6,7 Forthcoming work will isolate the contribution to these associated outcomes of subdomains of specific maltreatment types, further advancing understanding of which particular experiences present greater and lesser threats to health and development. These new understandings will bear relevance for tertiary and secondary prevention through child protection systems, and can inform legal, ethical and normative analysis of appropriate thresholds for any statutory response, and different levels of response. However, the fundamental motive for these scientific advances is to inform enhanced and targeted primary prevention of child maltreatment through health and social systems, especially of those experiences found to be most widespread and harmful, and to support appropriate social and clinical responses throughout childhood and adulthood for those with lived experience.8