Article Text

Download PDFPDF

Forty years later: inconsistencies in reporting of child abuse
  1. Robert D Sege1,
  2. Emalee G Flaherty2
  1. 1
    Boston Medical Center and Boston University School of Medicine, Boston, Massachusetts, USA
  2. 2
    Children’s Memorial Hospital and Feinberg School of Medicine, Northwestern University, Chicago, Illinois, USA
  1. Robert D Sege, Boston Medical Center and Boston University School of Medicine, Boston, Massachusetts, USA; robert.sege{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

In the late 1960s, clinical reports of battered children led to the enactment of state laws throughout the United States that required medical providers to report instances of suspected child maltreatment to state child protection agencies.1 In the ensuing 40 years, a great deal has been learned about child maltreatment. However, although medical professionals generally receive some training on the recognition and reporting of suspected child maltreatment, many cases of child abuse go unrecognised and unreported, occasionally with tragic consequences.2

Child abuse reporting by medical professionals requires both knowledge and experience. Specific knowledge helps practitioners recognise known patterns and mechanisms of injury associated with abuse, as well as those patterns and mechanisms of injury that commonly result from childhood unintentional injuries. Basic knowledge of normal child development helps flag those incidents in which the history provided does not match the child’s age or ability. Moving beyond direct pattern recognition, recent research has also demonstrated that even when abuse is recognised, professionals must make an active decision to report the child to state authorities.37 Their training, personal experiences and beliefs about the system in place for child protection all influence this decision.

This review will describe recent health services research that reveals the current barriers to reporting. Broadly classified, these can be divided into two major areas: failure to identify maltreatment, and deciding not to report suspected abuse to state authorities. From the point of view of the abused child, these barriers result in remarkably inconsistent medical evaluation and management.

Although child maltreatment consists of physical abuse, emotional abuse, sexual abuse and neglect, this review will focus on child physical abuse. Even though the elements of decision making vary somewhat depending on the type of maltreatment, the barriers we describe and the conclusions we present are likely to be broadly applicable to all types of maltreatment.


Continuing medical education concerning child maltreatment begins with the identification of possible child abuse.8 For example, the American Academy of Pediatrics has produced photographic sets that demonstrate the typical physical and radiographic findings associated with child battering.9 Clinical syndromes, including head injuries in young infants,10 11 long bone fractures12 and abdominal injury,1316 have been described. Key elements in the clinical history – including absent or changing histories – indicate a higher likelihood of child abuse.17 Nevertheless, even serious abuse-related injuries may be missed. Carole Jenny and colleagues have described a series of hospitalised children diagnosed with abusive head trauma.18 In 31% of these cases, a physician had evaluated the child on a previous occasion and failed to diagnose abuse-related injuries.

Physically abused children may be treated in a variety of settings, including the emergency department and paediatric or family practice primary care offices. Lack of knowledge about child abuse and common presenting complaints may be a substantial cause of under-reporting. While the efficacy of continuing medical education to improve decision making about possible child abuse remains controversial, it appears that residency training and professional subspecialisation play an important role in correctly identifying child abuse. In one study, primary care clinicians who had received some formal education about child abuse were much more likely to report suspected child abuse to child protective services (CPS).5 Reviewing data from the United States, Trokel reported that infants admitted to general hospitals for the treatment of highly suspicious injuries – long bone fractures or intracranial bleeding with fracture – were less than half as likely to be diagnosed as having been abused than similarly injured infants treated at children’s hospitals.19 This report suggests a substantial difference in the rate of recognition of severe child abuse between paediatricians (who generally work in children’s hospitals) and the emergency physicians who provide care for injured children treated at general hospitals.

Unfortunately, racial bias also continues to play a role in the identification of child abuse. In one recent study conducted in Pennsylvania, emergency department records were reviewed for all children with an injury type that may have resulted from child abuse.20 The investigators discovered a disturbing racial bias in completing a full evaluation of these children, including appropriate radiographic studies. African American children were more likely to have had a skeletal survey performed (OR 8.75) and were more likely to be reported to CPS (OR 4.3).


The clinical path towards appropriate management of child abuse progresses from recognition of common patterns of physical abuse to an assessment of the likelihood that the injuries sustained by the individual child were, in fact, the result of maltreatment. There are very few instances in which a childhood injury is pathognomic for maltreatment, but in many cases the physical findings, imaging studies and a careful clinical history raise the possibility of abuse. Clinicians evaluating these data then make a judgment about the probability that the injury resulted from abuse prior to reporting.

In theory, the clinician need not ascertain whether or not the injury did, in fact, result from abuse: statutory language typically sets a low bar for reporting abuse.1 In most states, healthcare providers are instructed to report whenever there is “reasonable” suspicion that a child has been abused. Levi et al21 reported that paediatric residents varied greatly in their interpretations of this phrase: nearly two thirds reported that the likelihood of abuse needed to be over 50% before this threshold was met. However, these same residents gave inconsistent responses to several other measures that attempted to ascertain the threshold likelihood of abuse that should trigger a report to the state CPS. In an accompanying commentary, Flaherty suggests that the resident’s responses reflect a general lack of consensus among physicians about the meaning of “reasonable cause” and the level of suspicion that should prompt reporting.22


The Child Abuse Recognition and Experience Study, a prospective assessment of 15 000 cases of children with physical injuries, demonstrated that few children were reported to state CPS until the clinician judged that abuse was at least possible (at the midpoint of a five-point scale). Only one in four children with injuries judged “possibly”, “likely” or “very likely” to have been caused by abuse were reported to CPS.23 This study demonstrated that even at the highest level of suspicion (“very likely”) many cases went unreported.

A number of studies have demonstrated that medical providers occasionally decide not to report suspected child abuse, even after identifying a particular case as suspicious for abuse.6 7 2430 The two most common reasons the medical providers gave for not reporting were lack of certainty that the injury was caused by child abuse and a belief that they could intervene with the family more effectively than CPS. They also expressed a concern that they would lose the family as patients and the fear that they would suffer negative personal consequences as a result of filing a report of suspected abuse or neglect. Clinicians appear to weigh their assessment of the CPS response and may decide not to report suspected abuse if they have had previous negative experiences with CPS around similar reports.5 31 Clinicians may be influenced by negative media reports and by published studies that suggest that children who have been reported may continue to experience abuse.32 33 Thus, the natural scepticism of clinicians may be exacerbated by concerns about the utility of filing reports. This effect may be magnified when, as often occurs, the clinician receives very little feedback from state CPS agencies after making a report.5 In the CARES study cited above, clinicians frequently cited their poor opinions of the likely CPS response in deciding not to report cases that they themselves suspect were probably caused by child physical abuse.34


There has been a great deal of discussion in the popular press about several high profile cases. In the United States, Louise Woodward’s defence team challenged the scientific literature describing abusive head trauma,35 even suggesting that child abuse paediatricians were members of a cult. In the UK, a firestorm of publicity surrounded the testimony of medical experts in cases of infant death.36 37 In both countries, defence attorneys attempted to cast doubt on the existence or frequency of abusive head trauma in children. While the scientific research has not changed, some reporters – including physicians – may have been unduly influenced by these press accounts. In the UK, two thirds of paediatric trainees have indicated they do not want to work with child abuse and about one third of the positions for child abuse physicians are unfilled.

A recent survey has shown that a large majority of child abuse paediatricians have experienced adverse personal consequences after reporting suspected child abuse. As malpractice claims have dramatically increased in the past decade,38 39 physicians have been sued because they reported suspected maltreatment to CPS. These suits have been brought despite laws protecting physicians who report suspected abuse in good faith. Fortunately, none of these lawsuits have been successful, but the physicians were still subjected to the stress and expense of defending themselves.7

Evaluating children for possible maltreatment is stressful for physicians. When physicians consider the possibility that a child may have been maltreated, they must consider the consequences of violence and of inappropriate treatment on the young and vulnerable victim. Although physicians generally work collaboratively with families for the best outcome for the child, the physician’s evaluation of the child as possibly having been abused may disrupt the usual collegial working relationship with the family. The evaluation of possible child maltreatment can be very time consuming because the physician needs to collect details of the history, collaborate with investigators and sometimes testify in court. Because of the adversarial nature of testifying, the experience is stressful for most physicians and often poorly compensated.40


The complete child protective program recommends that discerning clinicians reliably detect possible abuse and report their suspicions to competent state authorities for further investigation. If the child has indeed been maltreated, the child protective service should prevent further maltreatment by providing services or counselling for the family, or, occasionally, by finding a new loving home in which the child can thrive. Maltreated children should receive counselling and support, allowing them to heal physically and emotionally.

In the United States, this child protection system has been reinforced by laws requiring health care providers to report suspected child maltreatment. In point of fact, very few cases of failure to report have been prosecuted or litigated; however, this legal mandate provides an important departure point for training. Incidentally, the fact of illegality tends to make health care providers reticent to discuss difficult cases with colleagues unless they do plan to report, and reticent to cite their own decisions not to report all cases. Thus, the literature referred to in this review may underestimate the incidence of non-reporting of child maltreatment.

Forty years after implementing a mandatory child-reporting system in the United States, we continue to fall far short of protecting many vulnerable children. The diagnosis may be missed, suspicions of abuse are intentionally not reported, no intervention takes place, or the intervention is inappropriate or inadequate.2 32 33 39 41 42 Nevertheless, mandatory reporting sets a clear expectation for healthcare providers and offers strong support for protecting abused children. The consistent observation that reporting falls short even in the presence of a legal mandate should not be interpreted as undermining the importance of these laws. We believe, on the contrary, that the experience in the USA supports the importance of a reporting mandate in the UK as well, at the very least as a counterweight to the considerable systematic bias against proper child protection. Mandated reporting may be necessary but does not provide a sufficient basis for systematic improvements in child protection; it must be coupled with effective quality improvement43 and educational and policy efforts.44

Clinicians routinely fail to identify possible abuse, and, equally routinely, determine that reporting their suspicions to state authorities would not lead to benefit for the child or family. The next generation of child protection can begin with these data, and move forward to better accomplish the goal of child protection, with clinicians thoroughly ensconced in the web of services available to children.45 Our current understanding suggests that four steps would lead to an improved outcome for children:

  1. Broad education of clinicians likely to encounter children – including paediatricians, family practitioners, emergency medicine physicians, general, neurological and orthopaedic surgeons, and allied medical professionals – of the common patterns of child physical abuse.

  2. Support for front-line clinicians by providing access to child abuse physicians who could help guide the children’s work-ups and assess the likelihood of abuse.

  3. Process-oriented education so that clinicians understand the investigative process undertaken by child protection agencies and the services they provide.

  4. Finally, improved CPS assessments and interventions leading to better outcomes for children and families would go a long way towards improving the motivation of providers to report injured children to state agencies.

Rethinking the types of training we provide, the relationship between medical professionals and state agencies, and re-training medical and child protection professionals may be required to further improve the recognition and care of abused children.



  • None.

  • The authors would like to acknowledge the support of grant number 1R13HS01639-01 from the United States Agency for Healthcare Research and Quality.