Canadian Task Force on Preventive Health Care

Full Text Review

Please note: In 2003, the CTF updated its Grades of Recommendations to include an "I Recommendation" for situations where insufficient evidence exists to allow a recommendation to be made.  (Formerly, these situations were captured under a "C Recommendation".)  This change is not retroactive, and all "C Recommendations" made prior to 2003 have not been reevaluated in light of the new "I" recommendation grade.  For a discussion of these recommendation grades, please link to the 2003 article in the Canadian Medical Association Journal here.

Primary Prevention of Child Maltreatment

Prepared  by Harriet L. MacMillan, MD, FRCPC, Departments of Pediatrics and Psychiatry, McMaster University, James H. MacMillan, MSc, Glaxo Canada Inc, and David R. Offord, MD, FRCPC, Department of Psychiatry, McMaster University

These recommendations were finalized by the Task Force in January 1993

Up Contents

UpOverview

Child maltreatment includes the categories of physical abuse, neglect, sexual abuse and emotional abuse. In 1979, the Canadian Task Force on the Periodic Health Examination reported that there was fair justification for recommending that parenting problems, including child abuse and neglect, be included among those conditions considered in a periodic health examination. In evaluating the evidence since that time, this chapter considers the possibility of using screening to identify individuals at risk of maltreating children. It also examines programs for primary prevention of child maltreatment, such as perinatal and early childhood support programs (e.g. home visitation) and educational programs designed to teach children to recognize and respond to potentially abusive situations. The latter were included recognizing that parents may consult physicians regarding school-based programs. This chapter evaluates preventive programs based on the outcomes most closely related to maltreatment – for example, reports of verified or suspected abuse or neglect. Other more remote indices such as parenting attitudes were beyond the scope of the review.

Up Burden of Suffering

The term maltreatment includes all types of child victimization but is commonly divided into the subcategories of physical abuse, neglect, sexual abuse and emotional abuse. No national data are available in Canada regarding reports of child maltreatment. In the U.S., an estimated 45 reports of suspected child maltreatment per thousand children were received in 1992. However, many episodes of child maltreatment go unreported because of failure to detect, recognize or officially report the abuse. Deaths resulting from child maltreatment are also drastically underreported.

Estimates of both incidence and prevalence rates of child maltreatment have focused on two subcategories of child maltreatment: physical and sexual abuse. A U.S. national survey conducted in 1985 estimated the rate of severe physical violence against children by their caretakers as 11%. In the 1983 National Population Survey in Canada an estimated 15% of women and 8% of men reported that they had been victims of attempted or actual sexual intercourse before the age of 18 years. Other estimates of prevalence of childhood sexual abuse among girls 13 years or less vary from 10% to 12%.

Maltreated children are adversely affected in many ways. They may suffer from cognitive, emotional and social impairment in addition to physical disabilities. Many studies have pointed to an association between a history of childhood maltreatment and various psychiatric disorders including depression, personality disorders, anxiety, substance abuse, suicidal behaviour, conduct disorder and criminal behaviour. The human and fiscal costs associated with child maltreatment are clearly enormous.

Up Maneuver

Screening for Risk of Child Maltreatment

Most work in this area has focused on identifying people at increased risk for committing physical abuse or neglect. Methods of screening include three main approaches: a staff-administered checklist, a self-administered questionnaire and a standardized interview. The Family Stress Checklist<1> and the Dunedin Family Services Indicator<2,3> are examples of the first approach. Self-administered questionnaires include the Child Abuse Potential Inventory,<4> the Michigan Screening Profile of Parenting,<5> the Adult-Adolescent Parenting Inventory<6> and the Parent Opinion Questionnaire.<7> A standardized interview format has been used by Altemeier and collaborators.<8> Gray and colleagues<9> used a combination of approaches including interviews, questionnaire results and observations of parental behaviour. Leventhal, Garber and Brady<10> examined whether clinicians could correctly identify infants at high risk for abuse and neglect without the use of specific instruments.

Risk Indicators

A risk indicator is a factor associated with an increased likelihood of child maltreatment; it does not necessarily imply causation. Rates of physical abuse and neglect are similar among boys and girls. In contrast, girls are reported to be victims of sexual abuse 2.5 times more frequently than boys. Although children of all ages are at significant risk for physical abuse, those less than 5 years old and youths between 15 and 17 years old are at greater risk. Girls between 10 and 12 years are at increased risk of sexual abuse.

Risk indicators for committing physical abuse include low socioeconomic status, young maternal age, large family, single-parent family, parental childhood experience of physical maltreatment, spousal violence, lack of social support, and unplanned pregnancy or negative parental attitude toward pregnancy. The evidence regarding alcohol abuse and illicit drug use is unclear. Risk indicators of sexual abuse include: living in a family without a natural parent, growing up in a family with poor marital relations between the parents, presence of a stepfather, and poor child-parent relationships or unhappy family life.

Up Effectiveness of Screening

The main problem with the available approaches is the high false positive rate. For example, assuming a high prevalence rate for child maltreatment of 20%, screening 1,000 children with an instrument whose sensitivity is 80% and specificity 90% would result in 33% of the positive test results being false positive. With a lower prevalence rate of abuse, the number of false-positive results would be even higher. A sizeable number of individuals identified by such techniques as being "at risk for child maltreatment" would never go on to commit abuse. Such labelling may put people under increased stress and interfere with their ability to function as parents. Further, the validity of many of the screening approaches has not been adequately evaluated.

Overall, screening may do more harm than good. Nevertheless, knowledge of risk indicators for child maltreatment can assist clinicians in making decisions regarding the provision of preventive interventions to individuals and families in high-risk populations. Although screening of individuals is not recommended, interventions can be targeted at all members of high-risk communities.

Up Maneuver

Preventive Interventions

For the purpose of this review, primary prevention is defined as any intervention provided to stop maltreatment before it occurs. This includes interventions aimed at high-risk populations, sometimes referred to by others as secondary prevention. These interventions can be considered in two categories: 1) perinatal and early childhood hospital support, home visitation and parent training programs and; 2) education programs for children, parents and teachers. The former have generally focused on the prevention of physical abuse and neglect, whereas the education programs have primarily centred on the prevention of sexual abuse or abduction. Programs in both categories frequently used a spectrum of measures to evaluate effectiveness. This review includes the perinatal and early childhood programs that used official reports of verified or suspected abuse and neglect, in addition to the following proxy measures of maltreatment: rates of hospital admission, emergency visits and injuries. Effectiveness of education programs is frequently evaluated using measures of knowledge or behavioral responses under simulated conditions. Two systematic reviews provide a detailed description about the evaluation of these interventions.<11,12>

Up Effectiveness of Prevention

Perinatal and Early Childhood Programs

A randomized controlled trial (RCT) evaluated intensive pediatric contact plus home visitation by public health nurses (PHNs) and lay health visitors for 100 mothers identified as being at risk for abnormal parenting practices.<9> The number of verified reports of child maltreatment and accidents did not differ significantly between the intervention and control groups, however, the children of women in the control group were significantly more likely than those of women in the intervention group to require inpatient treatment for serious injuries. The number of central registry reports was greater in the intervention group, although the difference was not significant; one possible explanation for this result was increased surveillance among families in the intervention group.<9> This study suffered from inadequate follow-up; outcome was evaluated in only 50% of the families. Since this study examined a combined intervention of intensive pediatric contact plus home visitation, the lower number of seriously injured children in the intervention group cannot be attributed to intensive pediatric contact alone.

Several RCTs have evaluated home visitation as the primary preventive intervention. The two most rigorous studies demonstrated a reduction in the incidence of child maltreatment and outcomes related to physical abuse and neglect in the intervention groups. Olds and associates<13> evaluated home visits by nurses made to white primiparous women who were primarily young, single or of low socioeconomic status (85% of the 400 women met at least one of these criteria). Women in the control group received no services during pregnancy or free transportation for prenatal and well-child care; their infants underwent developmental screening. Of the two treatment groups, one was visited by a nurse during pregnancy (pregnancy-visited group) and the second during pregnancy and after birth until the child’s second birthday (infancy-visited group). The babies in the latter group were taken to the emergency department significantly less often in the first (p=0.04) and second (p=0.01) years of life and were seen less frequently for accidents and poisonings in the second year (p=0.03) than babies in the comparison group. In a subgroup of mothers at highest risk for maltreatment (poor, unmarried teens), 19% of those in the comparison group and 4% of those in the infancy-visited group had instances of verified abuse and neglect (p=0.07). The incidence of outcomes in the pregnancy-visited group generally fell between the rates in the infancy-visited group and the comparison group.

In a RCT, 290 black mothers of low socioeconomic status were assigned to receive either home visits beginning in the newborn period until the infant was 24 months of age or no such intervention.<14> Seventy-eight percent of the women were single and 23% primiparous. The home visitor was a woman from the community, with support provided through a health care program for children and youths. Children in the intervention group had significantly fewer admissions to hospital (p<0.01) and fewer episodes of suspected abuse or neglect (p<0.01) than those in the control group. They also had fewer episodes of definite physical abuse and neglect (p<0.01). (We conducted additional analyses using Fisher's exact test.)

Siegel and colleagues<15> evaluated the effects of three types of intervention: 1) early and extended hospital contact after delivery between women and their newborns; 2) home visits by paraprofessionals during the first three months after birth; and 3) both. At one year follow-up, the three intervention groups did not differ from the control group in the number of reports of abuse and neglect, hospital admissions or visits to the emergency department. However, the visits continued only during the first three months of the infant’s life.

O’Connor and associates<16> compared the effects of extended postpartum hospital contact (rooming-in) with routine care. Although the experimental group showed a reduction in parenting inadequacy, no significant differences were found in the number of hospital admissions, accidents, emergency department visits or reports of maltreatment to protective services. The outcome of parenting inadequacy was too broad to draw conclusions about prevention of child maltreatment. Evidence from the study was weakened because the study was not truly randomized.

A controlled trial without randomization and involving high-risk families evaluated contact by the project social worker after the mother’s discharge from hospital and access to a drop-in-centre.<17> The proportion of children on the child abuse register, the rate of hospital admissions and rate of admissions because of trauma did not differ significantly between the intervention and control groups. Our statistical analysis of descriptive data provided by the authors revealed that the number of children seen in the emergency department was lower in the intervention group than in the control group (p<0.001). However, no statement was made about randomization or baseline comparison of groups. Given these problems, no conclusions can be drawn from the trial.

One RCT<18> and a non-randomized prospective controlled study<19> evaluated the effectiveness of parent training programs for mothers at risk of committing child abuse. Neither study evaluated reports of abuse or events (e.g. hospital admissions) related to child maltreatment, so once again no conclusions can be drawn about the prevention of child abuse and neglect.

Overall, the evidence regarding intensive pediatric contact,<9> home visitation over the short term (three months or less),<15> early and/or extended postpartum hospital contact,<15,16> use of a drop-in centre<17> and parent training programs<18,19> remains inconclusive. Several of the studies lacked sufficient statistical power to detect a difference between the groups in the outcomes evaluated in this review.<15-17> Two trials<13,14> provide evidence that home visitation can prevent child physical abuse and neglect or associated outcomes among disadvantaged families.

Specific recommendations about the intervention cannot be made based on the available evidence. Authors of the most rigorous trials of home visitation have emphasized three essential aspects: 1) the importance of building a supportive relationship between the home visitor and mother over time; 2) flexibility on the part of the home visitor; and 3) adequate backup support for the home visitor. Both programs extended until the newborn child reached two years of age. In the trial by Olds and colleagues<13> visits occurred every two weeks initially, gradually tapering to every six weeks by the time the infant was two years of age. In the second trial,<14> visits were scheduled initially every two months tapering to every three months. The duration of visits in the two trials ranged from 40 to 75 minutes. The visitor was free to tailor the curriculum to the specific needs of the parents.

Education Programs

The second group of interventions comprises primarily school-based programs aimed at identifying potentially abusive situations and teaching strategies to prevent sexual abuse or abduction.<20-28> The target group for most education programs has been children from 3 to 12 years of age. Other trials evaluating the effectiveness of preventive education for teachers and parents are beyond the scope of this review.

Identifying inappropriate touching or advances by an adult and saying "No" were common elements of the educational curricula.<20> Some programs also taught children to report advances of an adult or that a victim must report the abusive episode. Interventions used various modes of presentation including instruction (verbal and/or written material), film or videotape plus instruction, skits plus instruction, film plus instruction and printed material (e.g. a colouring book), behavioral rehearsal plus instruction, and a combination of instruction, film or videotape and behavioral rehearsal. The frequency and duration of the training sessions varied. The interveners also had a range of qualifications. Outcomes reported across trials fell into four main categories: 1) knowledge of prevention concepts; 2) assessment of behavioral skills using responses to hypothetical vignettes; 3) behavioral responses under simulated conditions; and 4) disclosures of sexual abuse.

Numerous RCTs have demonstrated that education programs significantly increase knowledge about sexual abuse, enhance awareness of safety skills and modify children’s behaviour in response to hypothetical vignettes.<20-25> In two studies<26,27> education programs were effective in modifying children’s behaviour in response to a simulated abduction by a stranger. Disclosure of sexual abuse by children has been evaluated in a few trials. Due to methodologic problems (e.g. small control group), however, no conclusions could be drawn from these studies.<25,28> Participation in the trials has not been shown to be associated with negative effects but further assessment of both positive and negative long-term outcomes is necessary.<24>

The interpretation of outcome assessments remains a major dilemma. Researchers have assessed predominantly changes in knowledge assuming that increased knowledge leads to changes in behaviour. However, the appropriate response of a child in a research situation does not guarantee that the child will avoid abduction in real life. Although prevention of abduction and sexual abuse by strangers is a high priority, most sexual offenses are committed by people known to the child.

No study has produced evidence that the education of children about abduction and sexual abuse actually reduces the occurrence of such offenses. Without actual measures of abuse as outcome indicators, one cannot make firm recommendations about educational interventions for the prevention of sexual abuse and abduction.

Up Recommendations of Others

The second report of the U.S. Advisory Board on Child Abuse and Neglect released in September 1991, called upon the U.S. Federal Government to implement a universal voluntary neonatal home visitation system. The evidence reviewed in this chapter indicates that the effectiveness of home visitation programs in preventing child maltreatment has been demonstrated only in high-risk populations.

Up Conclusions and Recommendations

There is fair evidence not to recommend use of screening devices for identifying parents or families at risk for child maltreatment (D Recommendation) because of the high false positive rates and the harm associated with labelling parents as potential child abusers. However, physicians should know the risk indicators that characterize populations at increased risk for child maltreatment so that effective interventions for high-risk groups can be recommended (communities with high rates of poverty, and single and adolescent parenthood).*Although the recommendation has not changed, the evidence review has been updated for this recommendation.
Link to recommendation table of 2000 Update: Primary prevention of child maltreatment

Two RCTs<13,14> provide evidence that home visitation can prevent child physical abuse and neglect or outcomes associated with maltreatment among disadvantaged families characterized by one or more of single parenthood, teenaged parenthood, and poverty. Thus, there is good evidence to recommend referral for home visitation during the perinatal period and through infancy to prevent child physical abuse and neglect for families of low socioeconomic status, single parenthood or teenaged parenthood (A Recommendation).*Although the recommendation has not changed, the evidence review has been updated for this recommendation.
Link to recommendation table of 2000 Update: Primary prevention of child maltreatment

There is inconclusive evidence regarding including or excluding a referral for intensive contact with a pediatrician, early and extended postpartum hospital contact or both, use of a drop-in centre, or parent training programs in the prevention of child maltreatment (C Recommendation). These interventions may be beneficial for other reasons and should be assessed on an individual basis; whether they reduce the incidence of abuse and neglect remains to be established.

Whether education programs for children reduce the incidence of sexual abuse and abduction remains to be established. Physicians making recommendations regarding such programs must do so on other grounds (C Recommendation).

*Click here for additional prevention programs that were reviewed in the 2000 update.

Up Unanswered Questions (Research Agenda)

The following have been identified as research priorities:
  1. Measuring the prevalence rates and distribution of child maltreatment and its subcategories in the general population including identifying populations at high-risk.
  2. Investigating second-stage screening that would do more good than harm within high-risk populations.
  3. Determining for different populations the optimal content, duration and frequency of visits, and qualifications of providers for home visitation program.
  4. Further evaluation of interventions aimed at prevention of sexual abuse (effectiveness in reducing incidence of such episodes, in identifying children who have been sexually abused and any negative effects associated with the education programs).

Up Acknowledgements

The authors thank Drs. William R. Beardslee, Associate Professor of Psychiatry, Judge Baker Children’s Center, Boston, Massachusetts, Leon Eisenberg, Professor Emeritus, Department of Social Medicine, Harvard Medical School, Boston, Massachusetts, USA, Kenneth M. McConnochie, Associate Professor of Pediatrics, University of Rochester, School of Medicine and Dentistry, Rochester, New York, and David L. Olds, Professor of Pediatrics, University of Colorado Health Sciences Center, Denver, Colorado, for their helpful criticism of the background report.

Harriet MacMillan was supported by a New Faculty Research Fellowship from the Ontario Mental Health Foundation and the Chedoke-McMaster Hospitals Foundation. David Offord was supported by a National Health Scientist Award from Health Canada. Funding for this report was also provided by Health Canada under the Government of Canada’s Family Violence Initiative.

 Up Evidence

The medical literature was identified using the following search strategy for the years January 1979 to April 1993: 1) MEDLINE using the MESH headings child abuse, battered child syndrome, incest, and the textword prevent; 2) ERIC, PSYCINFO, CRIMINAL JUSTICE PERIODICAL INDEX and CHILD ABUSE AND NEGLECT using the descriptors child abuse, child neglect, incest, and prevention.

This review was initiated in 1992 and recommendations were finalized by the Task Force in January 1993. A report with a full reference list was published in January 1993.<29>

 Full Citation

Link to Structured Abstract of this review

Link to Summary Table of this review

Link to Selected References list of this review

Link to 2000 Update: Primary prevention of child maltreatment

Literature Search Update: Aug 31, 1998

Top of Page

Home PageCTFPHC Home Page

Reprinted in modified format by the Canadian Task Force on Preventive Health Care
with permission.
For any technical issues please contact: webmaster@ctfphc.org
Original Copyright © 1994 Minister of Supply and Services Canada.
Last modified April 1, 1998.