History
As with other forms of child maltreatment, child sexual abuse (CSA) has likely occurred since the dawn of human history. But unlike physical abuse, neglect and psychological maltreatment, CSA has been shrouded by the cloak of social taboo surrounding sexual contact with children and human sexuality in general. This made determining the true number of CSA cases difficult, leading physicians and other scientists to believe it was an uncommon problem. In the 1970s in the United States, reports of CSA grew dramatically as the social changes associated with the women’s movement revealed the plight of sexually victimized children. Early counts of CSA rose dramatically from a few thousand, to 44,700 annually in 1979. CSA now consistently comprises 10% to 15% of child maltreatment (CM) reports in the United States and Canada. , Similar patterns have been noted in other countries, with initial reports of CSA being low or “nonexistent” in number, and more recently increasing case identification and reporting associated with social acceptance and improved professional response. Despite improved identification and reporting, a large proportion of CSA cases are thought to remain hidden from public view or investigation while real numbers appear to be declining in the United States.
Terminology
A variety of sources reports aspects of the incidence and prevalence of child sexual victimization. Unfortunately, varying definitions of the type of sexual contact (direct or indirect, penetrative or nonpenetrative, harm or endangerment) and what constitutes a “child” can make assessment problematic. Rape, which is often reported by law enforcement and criminal justice systems, has been generally defined as forceful, penetrative contact, and is further specified in state penal codes. Sexual assault refers to a broader collection of acts, including fondling and other nonpenetrating acts, and also is further refined in state penal codes. Other terms imply the relationship of the offender to the victim. Incest refers to sexual contact between family members, which is sometimes limited to immediate family but in other contexts can extend to fifth degree relationships (second cousin, once removed). Sexual exploitation generally refers to acts without sexual contact, such as having children pose for sexually explicit photographic or video images, having them witness sexual acts, or by adults exposing themselves to children inappropriately for the sexual gratification of the adult. Thus a broad definition of child sexual abuse has been taken as the “… involvement of dependent, developmentally immature children and adolescents in sexual activities that they do not fully comprehend, to which they are unable to give informed consent, or that violate the social taboos of family roles. ” This has been modified for practical application to “… an act of commission, including intrusion or penetration, molestation with genital contact, or other forms of sexual acts in which children are used to provide sexual gratification for the perpetrator. This type of abuse also includes acts such as sexual exploitation and child pornography. ”
Case Finding
David Finkelhor has noted that “because sexual abuse is usually a hidden offense, there are no statistics on how many cases actually occur each year. Official statistics include only the cases that are disclosed to child protection agencies or to law enforcement.” There are several ways, however, that CSA can be identified. Cases are most often reported by witnesses or disclosed by the child. These reports are transmitted to law enforcement and child welfare agencies (child protective services [CPS] in the United States) as “suspected cases” until an investigation identifies credible evidence to make a determination that the child is a victim and/or that a crime has occurred. To identify more cases, screening has been proposed to find victims in the general pediatric population. Screening procedures have been devised which use information from the parents, characteristics of the child, interview or physical examination findings, and other case factors. However, while some case characteristics have been found to be more predictive of CSA determination, there is no single “test” that identifies a child as a CSA victim. , That determination usually requires a finding by an investigatory agency, and the variability of these findings leads to variations in case findings in official statistics.
Incidence
Incidence refers to the number of CSA cases that occur each year, whereas prevalence is defined as the number of people who, at a given time, have been the victim of at least one act of CSA during their lifetime. These two approaches, measuring different aspects of the occurrence of CSA, come from different types of analyses and often appear to reach different conclusions about the extent of the problem. One can sometimes estimate the population prevalence of a condition from annual incidence statistics.
There are three principle sources of data on the incidence of CSA in the United States. Traditional criminal justice agencies collect information about a variety of crimes in the United States, including violent crimes such as homicide and rape, and property crimes. The U.S. Bureau of Justice reports that while violent crime decreased 26.3% from 1996 to 2005, the rate increased 1.3% from 2004 to 2005. Although the National Crime Victimization Survey estimated there were 197,000 incidents of forcible rape and 110,000 other incidents of sexual assault of victims ages 12 and older in the United States, only one third were estimated to have been reported to law enforcement agencies in 1996. In the Federal Bureau of Investigation’s Uniform Crime Reports in 12 U.S. states during 1991-1996, two thirds of the 60,991 sexual assault victims were less than 18 years of age. Juvenile victims accounted for 75% or more of incidents of fondling, sodomy, and forcible assault with an object, but only 46% of rapes. Most offenders were male (96%) and older than 18 years (76.8%), but only 34% were family members, suggesting that only a relatively small proportion of the cases in this dataset are true CSA cases as defined by child protective services agencies and collected in the National Child Abuse and Neglect Data System.
The National Child Abuse and Neglect Data System (NCANDS) contains aggregate and case-level data on child abuse reports received by state agencies in the United States. Data were first collected in the late 1980s from a small number of states, but there are now more than 45 states and territories providing information annually about the outcomes of child abuse reports, types of maltreatment, child and family factors, and services being provided. National estimates of the overall numbers of CM victims (substantiated or indicated reports) and victims identified with the major types of CM (physical abuse, sexual abuse, neglect, medical neglect, and psychological maltreatment) are provided in Figure 3-1 . In NCANDS, the number of CM victims rose, fell, and then stabilized at approximately 900,000 annually since the year 2000, with rises in neglect and declines in physical abuse. The number of CSA victims, while rising during the late 1980s, actually declined during much of the 1990s and early into the twenty-first century. Cases declined from a peak of 144,760 cases in 1991 to 79,640 in 2006. CSA incidence rates also declined from 2.2 per 1000 children in 1990 to 1.1 per 1000 in 2006 ( Figure 3-2 ).
National incidence surveys are an additional source of information. The Canadian Incidence Study (CIS) reported that 11% of confirmed CM reports were for sexual abuse, affecting 0.93 children per 1000 in 1998. In the United States, the National Incidence Studies of child abuse and neglect (NIS) have provided separate, periodic estimates of a growing number of sentinel professionals in a representative group of U.S. counties to determine the actual number of CM victims. In 1993, NIS-3 sampled more than 5600 professionals in 842 agencies serving 42 counties to identify children in any or all of the agencies under two standards: The harm standard (relatively stringent in that it generally requires that an act or omission result in demonstrable harm to be classified as abuse or neglect) and the endangerment standard (which allows children who were not yet harmed by maltreatment to be counted if the CM was confirmed by CPS or identified as endangerment by professionals outside CPS, either by their parents or other adults). It was found that there was a two thirds increase in the overall number of CM victims since the previous study (NIS-2) in 1986. Sexual abuse nearly doubled during this time period, rising to an estimated 217,700 cases under the “harm standard” and 338,900 cases under the “endangerment standard” in 1993. Differences in these estimates from those reported by NCANDS are thought to be explained by: (1) The fact that NCANDS reports victims that have been investigated and determined to include CSA and do not include unsubstantiated or unfounded cases; (2) NIS includes cases identified by community professionals at schools and hospitals, but which have not been reported to CPS; (3) NIS includes cases under the “endangerment standard,” which do not meet CPS criteria for CSA case finding: and, (4) some cases are never revealed during the child’s lifetime. In one analysis, the true number of CSA cases was thought to be closer to NIS estimates. NIS-4 was conducted in 2006 and results are expected in 2009.
Prevalence
There are many studies which report the prevalence of CSA. Prospective designs may be more accurate than official CPS reports, but many prevalence studies are retrospective surveys in special populations at increased risk for CSA, suggesting potential biases might overestimate the true prevalence. , Early small studies reported prevalence rates as low as 3% for males and 12% for females, but with increasing social recognition and acceptance and improved survey techniques, rates of 25% or higher have been consistently identified. Prevalence studies have historically varied greatly in their definition of CSA and in their methods, but they also likely include cases that have not been reported in prospective incidence studies, creating an apparent disparity in the numbers of cases. It is estimated, for example, that less than one third of all CSA cases are reflected in current incidence figures, mostly because cases are not disclosed to authorities. Thus prevalence studies can offer an opportunity to “capture” more cases than are officially reported.
In the selected sample of studies presented ( Table 3-1 ), rates range from 1% in a population-based study in North and South Carolina to over 66% among pregnant adolescents in Washington. These studies have been completed over a wide span of years (1988 through 2002) and have wide variations in the self-reported rates of CSA based on locality, sampling technique (convenience vs. population), victim gender, age, type of sexual contact (CSA vs. rape vs. unwanted sexual contact), condition of interest (medical vs. psychological), or criminal justice status (incarceration). Women with pregnancy and men with sexually transmitted infections (consequences of sexual activity) had higher lifetime prevalence of CSA. University students, incarcerated men, and those with injection drug use also had greater rates. This does not mean that these populations are more likely to be abused; rather, it implies that a history of CSA, when obtained by retrospective self-report, is more likely to be found in groups with certain medical, psychological, and social problems.
In contrast, meta-analyses and studies with national samples offer potentially more accurate CSA estimates for the general population ( Table 3-2 ). For example, the National Family Violence Survey in 1985 reported that 27% of adult women and 16% of adult men reported sexual contact or sexual abuse during childhood, but their relationship to the offender (a key element of CSA) was not specified. Others later reported rates from as low as 4.5% to as high as 37%, varying by location and methodology. A meta-analysis of 59 studies from 1974-1995 noted that there were wide variations in definitions but that, in aggregate, college students reported rates of 16% for CSA with “close” family members and 35% for total CSA with “close” and “wider” family. These rates were 33% higher than the national studies used for comparison, but wide ranges of results were obtained depending on the sexual acts included in their definition.
