Elsevier

The Lancet

Volume 379, Issue 9817, 25 February–2 March 2012, Pages 758-772
The Lancet

Review
Child maltreatment: variation in trends and policies in six developed countries

https://doi.org/10.1016/S0140-6736(11)61087-8Get rights and content

Summary

We explored trends in six developed countries in three types of indicators of child maltreatment for children younger than 11 years, since the inception of modern child protection systems in the 1970s. Despite several policy initiatives for child protection, we recorded no consistent evidence for a decrease in all types of indicators of child maltreatment. We noted falling rates of violent death in a few age and country groups, but these decreases coincided with reductions in admissions to hospital for maltreatment-related injury only in Sweden and Manitoba (Canada). One or more child protection agency indicators increased in five of six countries, particularly in infants, possibly as a result of early intervention policies. Comparisons of mean rates between countries showed five-fold to ten-fold differences in rates of agency indicators, but less than two-fold variation in violent deaths or maltreatment-related injury, apart from high rates of violent child death in the USA. These analyses draw attention to the need for robust research to establish whether the high and rising rates of agency contacts and out-of-home care in some settings are effectively reducing child maltreatment.

Introduction

2 years ago, The Lancet published a Series of four reports on child maltreatment.1, 2, 3, 4 The Series was intended to provide professionals with a rigorous and up-to-date overview of the scientific evidence. 1 year on, The Lancet asked leading professionals in child health and welfare what question they most needed answered by the scientific published work. Their response, “Are trends in child maltreatment decreasing?”, is addressed by this Review.

Whether trends in child maltreatment are changing is of great importance for children and their families, and for those whose job it is to reduce maltreatment and its consequences. Policy makers and professionals involved in child protection services will hope for a downward trend to vindicate the cost, effort, and painful media scrutiny that they have endured.5 Public health practitioners will draw attention to the contribution of economic and welfare changes, legislation against corporal punishment, and initiatives to improve child wellbeing and parent functioning.5, 6, 7 Over the past 30 years, developed countries' tolerance of child maltreatment has decreased sharply.8 Another major shift has been the broadening of responsibility to all professionals to be alert to the possibility of child maltreatment and to act when they have concerns.9

However, increased responsiveness to child maltreatment inflates the number of reported cases throughout the system.10 The result is more notifications to child protection agencies, more concerns recorded by professionals, and more interventions, including placement of children in care, than previously.10, 11 Expansion of definitions of maltreatment to include emotional abuse and witnessing of intimate partner violence, and changing thresholds for moving from recognition to recording and action, have further increased rates.10, 12, 13, 14 This report addresses these issues in three ways. First, we tried to minimise reporting biases by comparing trends with multiple measures of child maltreatment within a specific country. Second, we examined differences between six countries or states. Our underlying premise is that comparison of trends in countries with similar challenges but different policies provides a natural experiment that could provide insight into the effects of policies. Third, we used routinely recorded data sources and standardised definitions in each country and examined the extent to which variation in indices of child maltreatment is likely to be explained by true differences in occurrence—possibly linked to policy differences—or by random chance, data quality, or case mix.

Key messages

  • We recorded no consistent evidence for a decrease or increase in all types of indicators of child maltreatment across the six countries or states (Sweden, England, New Zealand, Western Australia, Manitoba [Canada], and the USA) despite several policy initiatives designed to achieve a reduction.

  • Large differences between countries in the rate of contacts with child protection agencies contrasted with little variation in rates of maltreatment-related injury or violent death. This discrepancy shows that governments' responses differ.

  • Overall, one or more child protection agency indicators (notification, investigation, officially recognised physical abuse or neglect, or out-of-home care) increased in five of six countries and states, particularly in infants, possibly as a result of early intervention policies.

  • Lower levels of maltreatment indices in Sweden than in the USA are consistent with lower rates of child poverty and parent risk factors and policies providing higher levels of universal support for parenting in Sweden.

  • High and rising rates of out-of-home care affect a substantial minority of children, especially those of non-white or Aboriginal origin, despite no policy advocating this option and little evidence for its effectiveness.

  • To improve the evidence base for child protection policies, governments should facilitate use of anonymised, linked, population-based data from health-care and child protection services to establish the effect of policy on trends in child maltreatment. Rising placements of children in out-of-home care demand urgent assessment with randomised controlled trials.

Section snippets

What is known?

Existing research of how child maltreatment is changing in developed countries is conflicting. Studies that rely on officially recorded or substantiated maltreatment measure only a small part of the bigger picture—eg, in some settings as few as one in 30 of the children who experience physical abuse every year have their abuse officially recognised.2, 15, 16 One reason is that most child maltreatment is hidden and not recognised by professionals dealing with children. Another reason is that

Case definitions

Child maltreatment refers to any act of commission or omission by a parent or other caregiver that results in harm, potential for harm, or threat of harm to a child.2 Because this definition is broad and difficult to measure, we restricted our case definition to indicators of physical abuse or neglect that were likely to have been measured in a similar way over decades and that indicate sufficient professional certainty that child maltreatment has occurred to warrant recording.43 Cases of

Variation in child maltreatment indicators within country

Figure 3 shows trends over time in maltreatment indicators. Agency indicators are confined to placement in out-of-home care in Sweden and Manitoba because no other agency data were available. Figure 4 shows rates of placement in out-of-home care for all six countries. Table 2 shows rate ratios resulting from the time-trend regression analyses, and webappendix p 12 shows actual rates and figures in 2005–06.

Variation in trends between countries

After the mid-1990s, all countries showed stable or falling mortality rates, but these changes were significant in only five of 12 age and country groups (children 1 year and older in Sweden and the USA, both age groups in Canada, and infants in England; table 2). These downward trends are unlikely to be explained by violent deaths being labelled as accidents in recent years (diagnostic transfer), because the rate of non-motor vehicle accidents either paralleled trends in violent deaths or

Children recorded in multiple data sources

We measured the proportion of children identified by more than one data source in Western Australia and Manitoba, where contacts with child protection agencies, hospital admission records, and mortality data were linked. In both countries, a high proportion of children admitted to hospital with a maltreatment-related injury were in contact with the child protection agency at some point before the age of 11 years or the end of the data collection period. For example, in Western Australia between

Variation in maltreatment indicators between countries

Figure 5 shows rate ratios for each maltreatment indicator compared with the mean yearly rate in Western Australia in 2004–06, where data were available for all indicators (webappendix p 12 shows actual rates). All countries apart from England exceeded a two-fold higher rate than Western Australia for at least some agency indicators (lower 95% CI excluded 2·0). In New Zealand and the USA, rate ratios for most child protection agency indicators exceeded a two-fold increase in most age groups (

Summary of main findings

We examined variation in indicators of child maltreatment across six developed countries, with two approaches. First, we sought evidence for significant trends over time. We noted that after the mid-1990s, the rates of violent deaths and maltreatment-related injury admissions remained stable in most settings. Only in Sweden and Manitoba did decreases in violent deaths coincide with decreases in admission related to maltreatment injury. Officially recognised physical abuse or neglect mostly

What do the results mean?

Our analyses of routinely recorded indicators of child maltreatment represent an advance over previous patchy evidence because we used population-based, individual child-level datasets from several sources over many years. Policy makers need to know what the results mean, but inferences need to take into account the limitations of these observational data. Can we conclude that recorded variation is indicative of real variation (or absence of it) in maltreatment indicators? And can we infer that

Effect of policy on child maltreatment

Our results indicate relative stability or increases in child maltreatment indicators over time with little evidence of a decline, apart from in Sweden and Canada where rates of violent deaths fell. Although insufficient power to detect change is one explanation for the apparent stability, another explanation is countervailing policy influences.

A key issue is whether we would expect to see an effect of the interventions promoted by these policies, on the basis of the available research evidence

Next steps

Our results show no clear evidence for an overall decrease in child maltreatment despite decades of policies designed to achieve such reductions. We urge caution in interpretation of these findings. Stable rates of officially recognised maltreatment could suggest that interventions are failing to achieve the reductions that were hoped for, or they could indicate improved recognition of maltreatment coinciding with decreases in the overall prevalence of maltreatment. Similarly, policies might be

References (86)

  • JV Roberts

    Changing public attitudes towards corporal punishment: the effects of statutory reform in Sweden

    Child Abuse Negl

    (2000)
  • JM Leventhal et al.

    Are abusive fractures in young children becoming less common? Changes over 24 years

    Child Abuse Negl

    (2007)
  • P Sidebotham et al.

    Child maltreatment in the children of the nineties: a cohort study of risk factors

    Child Abuse Neglect

    (2006)
  • J Waldfogel

    Welfare reform and the child welfare system

    Child Youth Serv Rev

    (2004)
  • E Munro

    Improving practice: child protection as a systems approach

    Child Youth Serv Rev

    (2005)
  • NM Trocme et al.

    Major findings from the Canadian incidence study of reported child abuse and neglect

    Child Abuse Negl

    (2003)
  • RA Minns et al.

    Incidence and demography of non-accidental head injury in southeast Scotland from a national database

    Am J Prev Med

    (2008)
  • GB Melton

    Mandated reporting: a policy without reason

    Child Abuse Neglect

    (2005)
  • LM Jones et al.

    Child maltreatment trends in the 1990s: why does neglect differ from sexual and physical abuse?

    Child Maltreat

    (2006)
  • S Janson et al.

    Sweden. A 30 year ban on the physical punishment of children

  • J Mansell

    The underlying instability in statutory child protection: understanding the system dynamics driving risk assurance levels

    Soc Pol J N Z

    (2007)
  • E Munro

    The Munro review of child protection

    (2010)
  • Comparability of Child Protection Data

    (1999)
  • AD Theodore et al.

    Epidemiologic features of the physical and sexual maltreatment of children in the Carolinas

    Pediatrics

    (2005)
  • AJ Sedlak et al.

    Fourth national incidence study of child abuse NIS-4: report to Congress

    (2010)
  • DK Runyan et al.

    International variations in harsh child discipline

    Pediatrics

    (2010)
  • A Zolotor et al.

    Child Abuse Rev

    (2011)
  • D Finkelhor et al.

    Trends in childhood violence and abuse exposure: evidence from 2 national surveys

    Arch Pediatr Adolesc Med

    (2010)
  • S Janson et al.

    Violence against children in Sweden. A national survey 2006-2007 (in Swedish)

    (2007)
  • Child maltreatment in the United Kingdom: a study of the prevalence of child abuse and neglect (executive summary). BOOK (NSPCC, London). NSPCC Research Findings

    (2000)
  • L Radford et al.

    Maltreatment and victimisation of children and young people in the UK

    (2011)
  • DP Jutte et al.

    Administrative record linkage as a tool for public health

    Annu Rev Public Health

    (2010)
  • CD Holman et al.

    A decade of data linkage in Western Australia: strategic design, applications and benefits of the WA data linkage system

    Aust Health Rev

    (2008)
  • C Paxson et al.

    Welfare reforms, family resources, and child maltreatment

    J Policy Anal Manage

    (2003)
  • JA Martin et al.

    Births: final data for 2005

    Natl Vital Stat Rep

    (2007)
  • M Keller et al.

    Policy benchmarking report on neonatal health and social policies in 13 European countries

    Acta Paediatr

    (2010)
  • Report on maternity: maternal and newborn information 2004

    (2007)
  • P Laws et al.

    Australia's mothers and babies 2005. Perinatal statistics series no. 20

    (2007)
  • Report Card 7, child poverty in perspective: an overview of child well-being in rich countries

    (2007)
  • C Waldegrave

    Contrasting national jurisdictional and welfare responses to violence to children

    Soc Pol J N Z

    (2006)
  • DK Runyan et al.
  • S Waldegrave et al.

    A differential response model for child protection in New Zealand: supporting more timely and effective responses to notifications

    Soc Pol J N Z

    (2005)
  • Cited by (250)

    View all citing articles on Scopus
    View full text