< Back to Articles Index             

Pediatrics in Review, Vol. 2, No. 7, January 1981, 197-207.

Pediatric Understanding of Child Abuse and Neglect

Stephen Bittner, MD and Eli H. Newberger, MD

When C. Henry Kempe and his colleagues coined the term "battered child syndrome" in 1961, the attention of the American medical community was focused on one of the most dramatic manifestations of family violence. Since then family violence has been perceived as a major social problem, and the eyes of pediatricians have been opened to familial causes of morbidity and mortality.

In this paper the term "child abuse" is used to encompass all the symptom indicators of maltreatment of children, including physical injury, physical neglect, sexual abuse, and some ingestions of harmful substances. We address these problems not as discrete illness entities or syndromes, but as symptoms of different issues and risks for particular children in individual families.

Kempe noted that notwithstanding a long history of concern with child welfare, the pediatric community ignored the implications of injury and neglect of children because of a "process of denial that was unequal to anything...previously seen in pediatrics." This denial continues today in spite of an increasing and visible literature on child abuse.

The task of this paper is to summarize current knowledge about the causes, differential diagnoses, and management of child abuse in a fashion accessible to pediatricians and members of the colleague disciplines.

HISTORY, DEFINITION, AND PREVALENCE

Violence toward children has been condoned and endorsed throughout recorded history. In the view of many historians and social theorists, it is deeply embedded in the social institutions and legal structures of industrialized society.

Societies for the prevention of child abuse developed in the 19th century, and government and private agencies dealt with abused children throughout this century. Following Kempe's awakening of the medical community, there was an editorial outcry in professional and lay media. This led to a model Child Abuse Reporting Law promulgated by the US Children's Bureau. By the middle 1960s the law was adopted in some form by all states. Underlying these statutes, however, was a shaky knowledge base and a confusing set of criteria for reportable maltreatment.

Whether, indeed, child abuse can be defined with precision remains a question full of conflict even for experts. Gelles, a sociological scholar of family violence in the United States, identifies the term "child abuse" as "a political concept that is the single greatest obstacle which stands in the way of gaining an insight into the problem."

With so much confusion about its definition, the prevalence of child abuse has been difficult to determine. In 1969, Gil undertook a nationwide study of the cases that had been reported subsequent to the passage of the state laws; at that time only 6,000 cases of physical injury had been reported each year.

The concept of child abuse has been broadened in the last decade. Now each state's reporting criteria include neglect, sexual abuse, emotional abuse, and deprivation of necessary physical and moral supports for a child's development. Additionally, the list of professionals mandated to report has been increased to include virtually all who are responsible for the care of children. In 1978, there were more than 600,000 reports of the many problems that are now considered to be child abuse.

A nationally representative sample of American families was asked about their methods of child rearing in a recent survey by Straus, Gelles, and Steinmetz. The responses suggested a far greater prevalence of physical violence toward children than that suggested in the case report data. This survey, which asked respondents in intact families specifically whether children had been kicked, punched, bitten, beaten up, or threatened with a knife or gun, suggested that 3.6% of children between the ages of 3 and 17 are at risk for serious physical injury every year.

Official case reports are comprised almost entirely of poor families. But the finding of a high prevalence of violent acts toward children in middle and upper class homes and an absence of differences between ethnic groups in this survey suggests that the case report statistics are heavily biased toward reporting of poor and minority children; the children of the affluent may receive different diagnostic labels for their problems ("accidents" rather than "abuse"), and practitioners may feel an obligation to protect more affluent families from the stigma of reporting to public agencies.

MODEL FOR UNDERSTANDING CHILD ABUSE

The initial efforts to understand child abuse focused on the psychological problems of the parents of the victims. An influential study by Steele and Pollock pointed to abusing parents' distorted expectations of their children, frustrated dependency needs, personal isolation, and histories of having themselves been abused as children. Helfer has suggested that the propensity for abuse may have three elements: a child with qualities that are provocative; a parent with the psychological predisposition; and a stressful event that triggers a violent reaction. Social and cultural, in addition to psychological and familial, factors have been described in the recent literature, dispelling the widely accepted myth that child abuse results solely from individual deviant behavior.

It is useful to think of child abuse as culminating from a series of stresses that impinge on parents and children. Table 1 offers a framework for understanding the causes of child abuse in a social context. These "causes" have been identified as risk factors in populations of parents and children. The research has not uncovered causal connections for child abuse in the usual sense of illness pathogenesis. How risk may operate for any individual family must be assessed at the clinical level. Only by understanding the social, familial, psychological, and physiologic concomitants of child abuse can the pediatrician form a comprehensive management plan.

SOCIOCULTURAL FACTORS

At the top of Table 1 are summarized the social and cultural factors that guide individual behavior. Gil notes that "the most fundamental causal level of child abuse consists of a cluster of interacting elements, to wit, a society's basic social philosophy, its dominant value premises, its concept of humans. "

Zigler (former chief of the US Children's Bureau) states that "undoubtedly the single most important determinant of child abuse is the willingness of adults to inflict corporal punishment upon children in the name of discipline."

Confusion remains on the legitimacy of violence toward children, and the support of corporal punishment in the schools by such institutions as the Supreme Court may sanction violent practices in American homes, some of which culminate in incidents of serious harm.

Other national institutions have been criticized for their role in impeding the development of children and fostering violent behavior in families. Keniston (Chairman of the Carnegie Council on Children) asserts that the low levels of support in public welfare programs also assure a "perpetuation of exclusion" of children from the mainstream of American life. Poverty, not parental failure, is cited by Gil as the principal "abuse" of children, and its continuation is an example of "socially structured and sanctioned child abuse." Many poor children, reported as victims of child abuse and neglect, are placed in foster home care because of serious economic and familial problems and a shortage of services in the home to enable parents to care more adequately for their offspring. Too often these foster homes and institutions are inadequate or even harmful.

Violent entertainment in the cinema and on television may also affect how adults and children approach issues of conflict. Whether media violence is associated with childhood aggressive behavior remains a subject for debate, but there is a developing consensus that a milieu of violence fosters actions of violence.

FAMILY STRESSES

Social and cultural factors affect individual behavior and the quality of relationships within the family (see Table 1). It is helpful to think of "stresses" within the family as a way of organizing the complex data that accompany a child with the symptom of child abuse. Many families have some of the stresses listed, but we do not yet know enough to be able to predict child abuse in a given family with certain levels and qualities of stress.

SOCIAL-SITUATIONAL STRESSES

Structural Factors

Situational realities such as unemployment, geographic mobility, low levels of parental education, and inadequate, crowded living quarters may importantly increase the risk of child abuse. In the survey of Straus, Gelles, and Steinmetz certain types of violence against children occurred twice as frequently in low income families as in high income families, but this survey also found that the poorest and the most well-to-do families appeared to abuse their children least often. Members of the middle class suffered stresses including conflict between adult members around roles, money, and sex, and these were associated with high rates of violence. No differences were found between white and nonwhite families.

Social isolation has been suggested as a principal concomitant of child abuse. If parents have no friends, have no telephone contact with the outside world, or have lost contact with other adults, they will have no access to support in times of trouble. Many families feel detached from their communities, and employment demands keep many families on the move.

Parental Relationships

Although most psychological studies of child abuse have focused on the adult adaptation of individual parents, recent work suggests that the relationship between a child's parents may have as much to do with the occurrence of child abuse as any psychological qualities of individual adults. In homes where husbands and wives had used violence on each other in the past year, Straus found a 129% greater incidence of severe violence toward children. The inter-generational nature of child abuse was supported in this study which found that' 'respondents who reported that they had observed their parents hit one another had a much higher rate of violence towards their own children than respondents who said they had never seen their parents hit each other."

A high level of "verbal violence" has been suggested as a risk factor for physical violence. How decisions are allocated in a family may also be associated with a risk of child abuse; a power domination by either parent, characterized as a dominant-submissive interpersonal pattern, has been identified as a family concomitant in cases of child abuse (see Table 1).

Parent-Child Relationship

Low birth weight infants are at increased risk for child abuse, and this may be due to some failure of development of mother-infant attachment. Nursery routines that oblige protracted separations of newborns from their parents may contribute to this perceived risk. Associated with prematurity are other early stresses in the parent-child relationship: difficulties with the pregnancy and delivery, separation of the child and parent during the first months of life, or illness in the mother or child during the first year. How these stresses increase the probability of physical injury is not known.

Corporal punishment is accepted in all social classes, but when violence is a major mode of communication between parent and child, the relationship may take on a quality that maintains the violence. Families in which abuse has occurred may be administering corporal punishment inconsistently—in response to the child's failure to accomplish tasks and behavior that may be impossible at the child's chronological age, or when the child is insufficiently responsive and nurturing to the parent. This so-called "role-reversal" phenomenon is noted often in clinical work with families.

The child may behave provocatively in order to attract attention, even violent attention, from a parent, who in turn may express remorse for "having" to beat the child. There may also be a passive acceptance or an active encouragement by the other parent, which may affect both the intensity and quality of the violence.

The number of children in a family may promote intrafamilial stress in ways that are not clearly understood. Straus, Gelles, and Steinmetz found that the highest frequency of potentially injurious parental behavior took place in families with five children; families with two or three children or more than six children had much lower overall levels of violence.

CHILD-PRODUCED STRESSES

There has been a growing recognition that a child's own qualities may stimulate violence in his family. These qualities may include physical deformities, acute or chronic illness, slow intellectual development, psychiatric problems, or a temperament that is inadequately understood or tolerated by the parents. Historical circumstances may be associated with child-produced stresses, even in the absence of physical or behavioral idiosyncracies; a child may be born at a time of crisis and be identified as the bearer of trouble, or the child may have been unwanted. Stepchildren and foster children may be more susceptible to violence.

It has been suggested that vulnerable children may remind parents of their own traumatized past through behavior, physical characteristics, or developmental qualities. Self-destructive behavior has also been observed in victims of child abuse. This may be an expression of the child's acceptance of his family's rejection and may result from an inadequate sense of personal worth.

PARENT-PRODUCED STRESSES

Most adults who abuse children are not crazy. In comparison with other adults, they may be lonely, feel helpless or depressed, express unfulfilled dependency needs, and have little control over impulses. Other attributes that have been observed among abusive parents are inadequate understanding of child care and child development, misinterpretation of the meaning of such childhood behavior as crying or slow feeding or awakening at night, and a low tolerance for frustration. Alcoholism is often noted.

Because a parent may have psychological difficulties does not confirm the diagnosis of child abuse in a situation where there is doubt; and, equally importantly, cases should not be dismissed because the parents don't seem sick. Experienced clinicians in the field can all recount situations where victims of child abuse escaped detection for long periods because their parents appeared "so normal."

TRIGGERING SITUATION

There has been no systematic study of what happens to precipitate abusive events; some instances are acute and self-limited; other cases are of long duration. It is helpful nonetheless to consider the circumstances of the family's life in the period immediately prior to the injury which brings a child for care. Examples of triggering situations from clinical experience include a baby who, on a particular evening, would not stop crying; an alcoholic who was fired from his job; a mother who, after being beaten by her husband, could not make contact with her own mother; and the service of an eviction notice. Any of the stresses originating in a social situation, the child, or the parent, can become the triggering event for abuse. It is helpful to think of the situation of the family with an abused child being brought for care as a "crisis." A sensitive exploration of the origins of the problem and considerate attention to the recent family history by physician, nurse, and social worker can lead to clinical understanding of a given child's risk and can point the way to diminishing that risk.

DIAGNOSIS

History

The initial interview of a family in which child abuse or neglect is suspected has five objectives: (1) to understand the historical and especially the medical antecedents of the child's injury, and to assess the plausibility of the history; (2) to determine the dimensions of the ongoing risk to the child, so as to inform the choice of protective or family supportive interventions; (3) to gather the past medical history of the child and the family members; (4) to form a relationship with the family which will foster and support their participation in subsequent diagnostic and therapeutic work with other professionals; and (5) to explain the case report and other aspects of the protective service process: what the pediatrician and others will be doing to protect the child and to help the parents. This includes an honest reckoning with the parents about the professionals' concerns.

Physical Examination

Table 2 summarizes the differential diagnosis of symptoms of child abuse. A complete examination including developmental assessment should be performed on any child who may be a victim of child abuse. The child's affect and his verbal and behavioral interactions with his family and other adults should also be observed and carefully, noted. In assessing the origins of the symptoms, it is wise to keep an open mind. Do not rush to conclude that a given symptom or family problem is "diagnostic." Rather, the findings of the history and physical examination should guide the formation of hypotheses and a problem list. These can be more fully delineated by subsequent laboratory studies and social work and psychiatric consultations.

TABLE 2. Differential Diagnosis of Child Abuse
Clinical Findings	Differential Diagnosis	Differentiating Test
Cutaneous lesions
1. Bruising	Trauma
	Hemophilia	PT, PTT
	Von Willebrand's	Bleeding time
	Anaphylactoid purpura	R/O sepsis
	Purpura fulminans	R/O sepsis
	Ehlers-Danlos	Hyperextensibility
2. Local erythema or bullae	Burn
	Staphylococcus impetigo	Culture, Gram stain
	Bacterial cellullitis	Culture, Gram stain
	Pyoderma gangrenosum	Culture, Gram stain
	Photosensitivity and phototoxicity reactions	History of sensitizing agent, orally or topically
	Frostbite	Clinical history and characteristics
	Herpes zoster/herpes simplex	Scraping
	Epiderolysis bullosa	Skin biopsy
	Contact dermatitis allergic or irritant	Clinical characteristics
Occular findings
1. Retinal hemorrhage	Shaking or other trauma
	Bleeding disorder	Coagulation studies
	Neoplasm
	Resuscitation	History
2. Conjunctival hemorrhage	Trauma
	Bacterial or viral conjunctivitis	Culture, gram stain
	Severe coughing	History
3. Orbital swelling	Trauma
	Orbital or periorbital cellulitis	Complete blood cell count (CBC), culture, sinus x-rays
	Metastatic disease	X-ray, CT scan; CNS examination
	Epidural hematoma	X-ray, CT scan; CNS examination
Hematuria	Trauma	R/O other disease
	Urinary tract infection	Culture
	Acute or chronic forms of glomelular injury (eg, glomerulonephritis)	Renal function tests; biopsy
	Hereditary or familial renal disorders (eg, familial benign recurrent hematuria)	History
	Other (vasculitis, thrombosis, neoplasm, anomalies, stones, bacteremia, exercise, etc)	History, cultures, IVP
Acute abdomen	Trauma	R/O other disease
	Intrinsic gastrointestinal disease (eg, peritonitis, obstruction, inflammatory bowel disease, Meckel's)	X-ray studies, stool tests, etc
	Intrinsic urinary tract disease (infection, stone)	Culture, IVP
	Genital problem (torsion or spermatic cord, ovarian cyst, etc)	History, physical examination x-ray, laparoscopy (?)
	Vascular accident as in sickle cell crisis	Angiography, sickle prep
	Other (mesenteric adenitis, strangulated hernia, anaphylactoid purpura, pulmonary disease, pancreatitis, lead poisoning, DKA, etc)	As appropriate
Osseous lesions
1. Fractures (multiple or in various stages of healing)	Trauma
	Osteogenesis imperfecta	X-ray and blue sclerae
	Rickets	Nutrition history
	Birth trauma	Birth history
	Hypophosphatasia	Decreased alkaline phosphatase
	Leukemia	CBC, bone marrow
	Neuroblastoma	Bone marrow, biopsy
	S/P osteomyelitis or septic arthritis	History
	Neurogenic sensory deficit	Physical examination
2. Metaphysical and/or epiphyseal lesions	Trauma	X-ray and nutrition
	Scurvy	History
	Menkes syndrome	Decreased copper, decreased ceruloplasmin
	Syphilis	Serology
	Little league elbow	History
	Birth trauma	History
3. Subperiosteal ossification	Trauma
	Osteogenic malignancy	X-ray and biopsy
	Syphilis	Serology
	Infantile cortical hyperostosis	No metaphyseal irregularity
	Osteoid osteoma	Response to aspirin
Sudden infant death	Unexplained	Autopsy
	Trauma	Autopsy
	Asphyxia (aspiration, nasal obstruction, laryngospasm, sleep apnea)	"Near-miss" history
	Infection-botulism?	Cultures, bacterial and viral
	Immunodeficiency?	Immunoglobulins
	Cardiac arrhythmia?	Autopsy
	Hypoadrenalism?	Electrolytes
	Metabolic abnormality--calcium?--magnesium?	CA2+, MG2+
	Hypersensitivity to cow's milk protein?

The cutaneous manifestations of child abuse may be ambiguous (see Fig 1). In the first 24 hours, a bruise may be reddish. blue, or purple; from the first to the third day. the color becomes blue or blue-brown; with further metabolism of heme, the bruise acquires a greenish cast in the fifth to seventh day; and by the tenth day the bruise may appear to be yellow. Before disappearing completely in two to four days the bruise may take on a brownish hue. The child's skin color may make some of these transitional colors variable and difficult to interpret.

Other cutaneous manifestations that have been identified include bruises in the shape of a handprint, linear bruises, abrasions from whipping with a cord or rope, loop-shaped marks from cords that have been folded over, crescentic bite marks, alopecia or subgaleal hematoma from pulling of the hair, and areas of abraded skin which may be caused by being bound or restrained. Impetigo in its various forms may be confused with burns or inflicted injuries.

Wilson notes that in addition to recording the distribution, shape, color, location, and approximate measurement of bruises, it may be helpful to have photographs for purposes of documentation.

The significance of obtaining informed consent for such procedures cannot be overemphasized. Parents will quickly sense any inquisitorial intentions on the part of the pediatrician, and the utility of the photographs needs to be explained openly and honestly; often, it may be advantageous to omit the pictures in the interest of building and sustaining a helping relationship.

Four distinct patterns of inflicted burns have been described by Lenoski and Hunter. Forced immersion yields a doughnut-shaped distribution of the burn with a spared area, frequently the buttocks or back, where the body may have been in contact with a container and, thus, may have been protected from the heat of the water. A splash burn may produce non-uniform multiple noncontiguous burn areas, sometimes with "arrowhead" patterns where the water has spread laterally as it rolls off the skin. If the child's body is immersed in a maximally flexed position, the skin folds of the thorax will be spared. This will give a striped pattern. When a burn is caused by contact with a hot object, such as an iron, the object may leave a distinctive mark on the skin.

Less obvious physical examination signs of child abuse may reflect internal injuries. A ruptured tympanic membrane may result from a blow to the side of the head or from a basilar skull fracture. Multiple class I fractures of the teeth may result from repeated blows to the chin. A lacerated frenulum or other intra-oral trauma may result from forced feeding with a spoon.

Ocular injuries may include hyphema, corneal abrasion, subconjunctival hemorrhage, dislocation of the lens, detached retina, or retinal hemorrhages (see Figs 2 and 3). Caffey has described the syndrome of the severely shaken infant; this syndrome may include osseous lesions (Fig 4) such as metaphyseal avulsions or subperiosteal hematomas, intracranial hemorrhage, or retinal hemorrhage. The shaken infant may be difficult to identify as there may be no obvious cutaneous signs. Sudden thoracic compression and consequent increased intravascular pressure may also lead to retinal hemorrhages, the so-called Purtscher retinopathy.

Occult internal injuries include rupture of the pancreas and pseudocyst formation, lacerated liver or spleen, intramural hematoma of the bowel, retroperitoneal hemorrhage, renal laceration or contusion, intestinal perforation, rupture of the ureter or bladder, and chylous ascites.

The concomitant presentation of unusual chronic illness and child abuse has been noted in several recent reports. The term "Munchausen Syndrome by Proxy" has been applied to clinical situations in which parents make their children sick in order to attract attention to their own problems. A child with mysterious relapsing coma was found to have been given sublethal doses of chloral hydrate by his mother. Other reports include fever, bacteremia, recurrent idiopathic lesions of the skin, and sclerosed lesions of the cornea.

Laboratory studies are helpful in delineating the nature and extent of current trauma and in defining the presence of previous trauma. The radiologic findings of child abuse include multiple long bone fractures in various stages of healing, spiral fractures, epiphyseal fractures, and exaggerated periosteal reaction.

A coagulation profile (prothrombin time, partial thromboplastin, platelet count, and bleeding time) will exclude endogenous disorders of bleeding. Hemoglobinuria and hematuria are known to occur with major trauma.

OTHER SYMPTOM ENTITIES AND CHILD ABUSE

Family stresses may culminate in and be discovered in hospitals as symptoms other than the physical injury. The goals of the initial assessment are nevertheless the same: to assess the stresses, to determine the ongoing risks, and to begin the helping process.

Childhood ingestions are seldom random, isolated events. Although their prevalence and morbidity have been reduced by the use of secure containers, recent studies suggest that family crisis is frequently present when a child is poisoned.

In many states, the physician is obliged to report failure to thrive as a symptom of child abuse. The research on these children has been characterized by one recent reviewer as "primitive." There is sufficient doubt that failure to thrive is a variant of child maltreatment to lead us to recommend that care be taken not to assume that parents are at fault.

CHILD ABUSE AND SUDDEN INFANT DEATH SYNDROME (SIDS)

Each year approximately 10,000 infants die in the United States without apparent cause. Sudden infant death syndrome (SIDS) is a common cause of death for infants between the ages of 1 week and 1 2 months, and it is the second most common cause of death (after accidents) for all children between 1 week and 1 5 years of age. The relationship between SIDS and infanticide has been explored epidemiologically, although detailed comparative family studies are absent in the literature. SIDS is reported 100 times more frequently than infanticide, and there has been no change in the relative frequency since 1963, notwithstanding the increased awareness of both causes of death.

Despite the importance of autopsy for determining the origin of infant death and enlarging our knowledge base of SIDS, only a fraction of SIDS cases are examined pathologically.

In SIDS the autopsy may not delineate the precise cause of death, yet the information revealed from the pathologic study is important in allaying the guilt and anguish of the family. Dilemmas of interpretation may follow the autopsy. It is difficult to determine whether an autopsy finding, such as blood in the abdomen, occurred before or after the time of death. Resuscitation attempts may result in injuries (such as fractures and hemorrhages) that are difficult to distinguish from inflicted trauma.

A full history and review of any written records that may be available will help the pediatrician in the perplexing differential diagnosis of SIDS (see Table 2). If there is a suspicion that the victim may have been abused, it is wise to obtain legal counsel immediately. In all states law requires that homicides be reported promptly, but when the possibility of SIDS is considered, an autopsy should be performed to guide the choice of action. The pediatrician can protect the parents and siblings of the victim from certain hurtful legal processes. As SIDS occurs more frequently in families of lower socioeconomic status, who may be more likely to attract the label of "child abuse" to their children's problems, it is best to avoid premature judgments and to provide support for the parents in their distress.

SEXUAL ABUSE

Kempe has recer.cly pointed out that sexual abuse ren,ains a "hidden pediatric problem ar, j a neglected area. " Cases of sexual 1buse present enormous difficulties for pediatricians as well as other professionals.

As with inflicted injuries, there is conflict and ambiguity with regard to the definition of sexual abuse, the familial and parental concomitants, and the role of the victim. Brant and Tisza suggest that the concept" sexual misuse" might be more appropriate (than sexual abuse) to describe those situations in which there is "exposure of a child to sexual stimulation inappropriate for the child's role in the family," as distinct from sexual abuse in which there is forced sexual contact between a "victim" and a "perpetrator." Yet, the line between loving caresses and sexual misuse may often be unclear.

Incest is defined as sexual abuse by a parent, sibling, or other close relative, that includes intercourse. Molestation is defined as sexual abuse by a stranger with or without penetration. Modes of sexual contact include rape, defined as forced genital contact-usually introduction of the penis into the vagina of the female victim, but occasionally in children including the forced penetration of an adult female by a boy's penis; sexual assault, defined as violent or nonviolent manual, oral, or genital contact with the genitalia of the victim or the perpetrator; "immature gratification," defined as fondling, looking, or kissing the genitalia of the victim, and including confrontation by an exhibitionist.

An adult who seeks sexual contact with children, in whatever mode, is pedophilic. The sexual encounter may be coerced by the adult, or the child may be an active participant, seeking out or encouraging the adult because of needs for nurturance, mastery, or acting out aggressive feelings.

Our understanding of sexually abused children and their families is rudimentary. Finkelhor's recent survey of 795 undergraduate women revealed eight' 'vulnerability factors" in the women who identified themselves as having been sexually victimized as children. (A fifth of the women (and a tenth of the men) admitted to one or another sexually victimizing experience in childhood.) The factors were divorce and remarriage to a stepfather; an aloof, unavailable mother; the experience of having lived away from the mother as a child; a quality of "sexual punitiveness" by the mother toward her spouse; low educational status of the mother; few friends in childhood; and a family income under $10,000.

Finkelhor suggests that the current approach to sexual abuse is complicated by the historical context of sexual behavior between children and adults, and he notes that the process of discovery of sexual abuse has political dimensions. The women's movement has stimulated a consideration of the problem of sexual abuse as a variant of rape, and the proponents of protective services for child abuse have focused on the relationship with other forms of maltreatment. Yet, sexual abuse is unique in many ways. Unlike rape, sexual abuse is usually perpetrated by a friend or an adult known to the child. It is more frequently chronic, does not necessarily involve physical force, and seldom involves sexual intercourse. Unlike physical child abuse, the "injury" is usually psychological rather than physical, and the motivation is usually for sexual gratification rather than as an expression of anger. These differences create confusion when trying to comprehend the nature of the risk to a child, the needs of the family, and the therapeutic resources that might be employed.

The very suspicion of sexual abuse can be profoundly troubling to children and adults, and the force of the intrusions of the criminal justice and welfare systems when case reports are made can sometimes be more harmful than helpful to the psychological status of the victim and his or her family.

Sexual abuse may present with "nonspecific" symptoms such as enuresis and encopresis, hyperactivity, fears and phobias, sleep disorders, learning problems, compulsive masturbation, sexualized play, perineal irritation, other genital injury, and distorted, pseudomature personality development (Table 3). Incest frequently is discovered when the child or mother reports the problem to someone outside the family. By far the most frequent relationship is father-daughter, and often the child has been an active participant in the relationship over a long period of time. The incest is usually a symptom of a family system that includes a lonely, dependent father; a depressed, withdrawn, but subconsciously encouraging mother; and a needy daughter who may believe that her secret relationship with her father is the only thing that holds her family together.

When sexual misuse or abuse is suspected, the clinical evaluation should include a calm, careful, sensitive interview of the child alone, allowing the child to communicate with pictures, toys, and play. The parents, other close relatives, and other caretakers can be interviewed to assess risk factors in the home and to establish relationships that will carry beyond the crisis to support the family, even in the event of ambiguous medical findings and uncertain diagnostic conclusions.

The physical examination of suspected victims of sexual abuse should first be carefully explained (Table 3). Foreign bodies should be sought, and clothing examined for signs of semen or blood. The throat, rectum, and vagina should be culturedfor gonococcus, and a serologic examination for syphilis and serum assay for the beta-subunit of HCG (human chorionic gonadotrophin) should be performed. If vaginal discharge is present, microscopic examination for Trichomonas and Monilia as well as culture and Gram stain will be helpful in enabling prompt treatment of infection. Vaginal contents may be gently aspirated with an eye dropper. Venereal warts may also be found on careful examination. It is well to keep in mind that the data that are gathered may subsequently be reviewed in the criminal or civil court. For girls who have passed menarche, the pregnancy test is given as a matter of routine in cases where penetration might have taken place. Medical management at presentation may include prophylactic antibiotic therapy and diethylstilbestrol to prevent pregnancy.

INTERDISCIPLINARY MANAGEMENT OF CHILD ABUSE

Table 4 summarizes the management of child abuse and divides the pediatric role into two phases: diagnosis and treatment. Principal questions are outlined, and interventions to protect the child and to help the family are summarized.

Since current understanding regards child abuse as a symptom of family distress and a problem with complex, multivariate origins, it should be managed by a diagnostic interdisciplinary team that includes a social worker, a pediatrician, a nurse, a psychiatrist, and an attorney. When such a diagnostic unit is not available, it may be necessary for the physician to help organize and to work with other professionals in the hospital or in the community. Management guidelines can be developed that utilize each community's resources and personnel. The protective service to which mandated reports are sent may not by itself be able to offer an adequate program of services. A social worker should be called promptly at the time of the family's presentation, both to facilitate the social assessment and also to form a helping relationship.

In the initial interviews and in subsequent contacts, no direct or indirect attempt is made to draw out a confession from the parent. Denial is a prominent ego defense in virtually all abusing parents, and their bizarre stories about how their children got their injuries ought not to be taken as intentional falsifications, but rather as repression of profoundly distressing realities.

Other defenses such as angry outbursts against the interviewer or refusal to talk limit both the process of information gathering and the prospects for continuing a helpful professional relationship. Breakdown of the assessment process may possibly endanger the child. It is appropriate to emphasize to the parent the child's need for care—which may include admitting the child to a hospital—and the need to ensure that the child is protected from harm. In explaining the legal obligation to report the case, the physician's compassion and honesty will go far to allay the family's anxiety. Other professionals may provide crucial aid in the evaluation process. The opportunity to observe parent-child interaction and the child's physical and psychological milestones (which might yield insight into the familial causes of a child's injury) may not be available to a physician in his office or in the emergency room. Nurses in clinical and public health settings can and do, however, make such observations, which are fundamental in case-finding and evaluation. The input of these nurses contributes uniquely to diagnosis, and their perceptions should be shared appropriately with the physician and social worker seeing the family.

A home visit by a public health nurse or social worker may be an important part of all initial assessments and is made to gather data on the child's home environment that may aid in making the disposition.

A psychiatric consultation is frequently obtained in cases of child abuse and neglect. Often this consultant's perceptions lead to understanding of what interventions can be most effective. However, a psychiatrist can rarely work magic; this consultation should not substitute for careful diagnostic work and energetic advocacy by the social worker, pediatrician, and nurse.

Several ethical dilemmas confront the pediatrician and his or her colleagues in the diagnosis and management of child abuse. The diagnosis itself is often impossible to make with certainty, and the physician, concerned with giving the parents the benefit of any doubt, may feel that the easiest, fairest, and most ethical approach is to send the child home without reporting. These clinical problems, once reported, may also consume substantial amounts of unremunerated time.

The reporting laws also oblige communication of confidential information when a child is suspected of being at risk, and this may place the pediatrician in conflict. The Hippocratic precept, "primum non nocere," is challenged when the reporting carries with it the risk of an incompetent intrusion into the life of the family by a poorly trained, inadequately supervised social worker from an overburdened and underfunded public child protection agency. The child may be separated from home, or help may not materialize. It is often necessary for the interdisciplinary team to choose "the least detrimental alternative," a concept suggested by Goldstein, Freud, and Solnit in Beyond the Best Interests of the Child, to guide a choice of management options when both may clearly carry the possibility of harm.

A consensus on seven axioms of child abuse management appears in the literature on child abuse.

1. Once diagnosed, abused children, especially infants less than 1 year of age, are at great risk for reinjury or continued neglect.

2. In the event the child is reinjured it is likely that the parents will seek care at a different medical facility.

3. There is rarely any need to establish precisely who it was who injured the child and if the injury was "intentional." The symptom itself should open the door to helping alliance and comprehensive service plans for the child and the family.

4. If there is evidence that the child is at major risk, hospitalization is appropriate to allow time for interdisciplinary assessment. The complex origins of the child's injury are seldom revealed in the crisis atmosphere at the time of presentation.

5. Protection of the child must be the principal goal of intervention, but protection must go hand-in-hand with the development of a family-oriented service plan.

6. Traditional social casework alone may not adequately protect an abused child in the environment in which he received his injuries. Multidisciplinary follow-up is also necessary, and frequent contact by all those involved in the service plan may be needed to encourage the child's healthy development.

7. Problems of public social service agencies in both urban and rural areas-specifically in numbers of adequately trained personnel and in quality of administrative and supervisory functions-militate against their effective operation in isolation from other care-providing agencies. Simply reporting a case to the public agency mandated to receive child abuse case reports may not be sufficient to protect an abused child or to help the family.

The development of programs that attend to these principles will require careful thought and planning. In the last analysis, the professionals' ability to convince patients or clients that they intend to help them depends on their ability to mobilize effective services. When case management programs and interdisciplinary cooperation improve, pediatricians and other professionals who work with children will find it easier and more rewarding to participate in comprehensive service plans.

ACKNOWLEDGMENTS

Partial support for this work came from grants from the Administration for Children, Youth, and Families, Department of Health, Education, and Welfare (Project OCD-CB141), and from the National Institute of Mental Health (grant 1 T01 MH15517 01A2 CD).

The authors wish to thank Dr Robert Peterson for the ophthalmologic illustrations, Dr Arthur Rhodes for critical review of the dermatologic material, Dr Robert Wilkinson for the radiograph illustration, Richard Gelles, PhD, for critical review of the manuscript, and Ms Debbie Sosin for assistance in preparing the manuscript.

FURTHER READING

Berwick D: Nonorganic failure to thrive. Pediatr Rev 1 :265, 1980

Bourne R, Newberger EH: Critical Perspectives on Child Abuse. Lexington, MA, Lexington Books, 1979

Brant R, Tisza V: The sexually misused child. Am J Orthopsychiatry 47:80, 1977

Caffey J: On the theory and practice of shaking infants: Its potential residual effects of permanent brain damage and mental retardation. Am J Dis Child 124:161, 1972

Finkelhor D: Sexually Victimized Children. New York, Free Press, 1979

Friedrich WN, Boriskin JA: The role of the child in child abuse. Am J Orthopsychiatry 46:580, 1976

Gil D: Unraveling child abuse. Am J Orthopsychiatry 45:346, 1975

Goldstein J, Freud A, Solnit AJ: Beyond the Best Interests of the Child. New York, Free Press, 1973

Halpern M: Fatalities from child abuse and neglect: Responsibility of the medical examiner and coroner. Pediatr Ann 5:42, 1976

Helfer RE, Kempe CH (eds): Child Abuse and Neglect: The Family and the Community. Cambridge, Ballinger, 1976

Kempe CH: Sexual abuse: The 1977 C. Anderson Aldrich Lecture. Pediatrics 62:382, 1978

Knitzer J, Allen ML, McGowan B: Children Without Homes. Washington, DC, Children's Defense Fund, 1979

Lenoski EF, Hunter KA: Patterns of inflicted burns. J Trauma 17:842, 1977

Mcintire MS, Angle CR: Psychological "biopsy" in self-poisoning of children and adolescents. Am J Dis Child 126:42, 1973

Meadow R: Munchausen syndrome by proxy: The hinterland of child abuse. Lancet 2: 343, 1977

Newberger EH, Hyde IN Jr, Holter JC, et al: Child abuse and child neglect, in Hoekelman RA, Blatman S, Brunell PA, et al (eds): Principles of Pediatrics. New York, McGraw Hill Book Co, 1978

Sobel R: Psychological implications of accidental poisonings in childhood. Ped Clin North Am 17:653, 1970

Steele Bf, Pollock CB: A psychiatric study of parents who abuse infants and small children, in Helter RE, Kempe CH (eds): The Battered Child, ed 2. Chicago, University of Chicago Press, 1974, pp 89-133

Stein AH, Friedrich L: Impact of television on children and youth, in Hetherington X (ed) Review of Child Development Research. Chicago, University of Chicago Press, 1975, vol 5, pp 183-250

Straus M, Gelles RJ, Steinmetz SK: Behind Closed Doors: Violence in the American Family. New York, Anchor Books/Doubleday, 1980

Taylor L, Newberger EH: Child abuse in the International Year of the Child. N Engl J Med 301:1205, 1979

Wilson EF: Estimation of the age of cutaneous contusions in child abuse. Pediatrics 60: 750, 1977

Zigler E: Controlling child abuse in America: An effort doomed to failure, in Bourne R, Newberger EH (eds): Critical Perspectives on Child Abuse. Lexington, MA, Lexington Books, 1979, pp 171-213

Stephen Bittner, MD: Assistant in Pediatrics, Department of Medicine, Children's Hospital Medical Center, Boston; Instructor in Pediatrics, Harvard Medical School, Boston.

Eli H. Newberger, MD: Director, Family Development Study, Children's Hospital Medical Center, Boston; Assistant Professor of Pediatrics, Harvard Medical School, Boston.