As shown previously in the Child Abuse Recognition Experience Study (CARES), primary care providers reported that they decided not to report a substantial proportion of injuries that they suspected might have resulted from abuse. The most serious cases result in death. This article provides detailed reports of 2 illustrative cases from the author's experience as a member of a multidisciplinary child fatality review team and discusses several alternative legal mechanisms for improving mandated reporting. Regional centers of excellence in child protection may be an effective way of improving child protection; current proposals to establish such centers are reviewed.
The finding of the Child Abuse Recognition Experience Study (CARES) that many cases of suspected child abuse go unreported was disconcerting but not wholly surprising.1,2 This article provides descriptions of 2 cases reviewed by our child death review team that illustrate some of the key considerations revealed by the CARES and discussed at the Child Abuse Recognition, Research, and Education Translation (CARRET) conference. Finally, the proposal to establish local centers of excellence in child protection is briefly reviewed.
Child death review teams often review cases in which failure to recognize and report abuse are contributing factors to a child's death. It is through a review of the autopsy findings and a reconstruction of the events surrounding a child's death that the abusive nature of the injuries is recognized.3
A 34-month-old girl was seen in the emergency department of a community hospital with the complaint that she was not using her arm. The history offered was that she had fallen off her bed. The nursing note documented the absence of pain, and a radiograph showed a supracondylar fracture. The medical notes showed that the child was examined. When questioned, the physician explained that the examination was performed without removing the child's clothing. An orthopedist was not available. There were 3 unsuccessful attempts to draw blood from the child, and a computed tomography scan of the head was requested. It was not obtained before the family left at 1:00 am, saying that they would return in the morning. The next morning the child was found apneic and without a heartbeat. She could not be resuscitated. At autopsy she was found to have a complete transection of the duodenum at the ligament of Treitz, a tear of the falciform ligament, and a hematoma of the liver. There was evidence of peritonitis. She had visible bruises on her body, including a knuckle mark on her abdomen and a hematoma of the right pinna.
Is this a case of failure to recognize abuse, or a failure to report it? Did the abdominal injury occur after the child left the emergency department? Should the staff at the community hospital have been suspicious? Is it ever acceptable for a child with a traumatic injury to be examined without being completely unclothed? The nurses were worried that something was wrong, but they deferred to the physician. The staff was reassured by the family's promise to return in the morning, but by morning the child was dead. Would intervention at 1:00 am have saved her life?
It is apparent that some cases represent failure to recognize that abuse occurred. Failure to recognize can be related to 3 different situations:
Lack of adequate knowledge about the medical findings related to child maltreatment, including failure to recognize even classic cases of inflicted trauma:Occasionally, there is an overestimation of one's knowledge and skills, and the marginally informed think they are more expert than they are.4 Insufficient knowledge stems from both inadequate didactic education and insufficient clinical experience, and although education is essential for the recognition of abuse, education in and of itself is not sufficient to ensure that abuse will be either recognized or reported. At community hospitals in which physicians are less likely to have pediatric experience, children are less likely to have child abuse recognized, compared with those seen at pediatric facilities.5
Unusual presenting complaint (the possibility of a traumatic injury is not considered):Related to this scenario is a presenting complaint that is so nonspecific that the traumatic nature of the finding is unrecognized (eg, vomiting misdiagnosed as acute gastroenteritis rather than traumatic head injury).6
A traumatic injury that is not consistent with the child's developmental level or with the proposed biomechanics, but the physician chooses to believe the story offered by the parents:Although pediatricians may sometimes accept a developmentally incredible story, this scenario more often relates to nonpediatricians such as emergency medicine physicians or orthopedists.7
STRENGTHEN MANDATE TO REPORT
Aside from mandating child abuse education for licensure, another legal approach is to strengthen the mandate that suspicion of child abuse be reported by threatening prosecution for those who fail to report their suspicions to the state child protective service agency. In our first case, some members of the child death review team questioned the failure to report a supracondylar fracture in a 2-year-old, not recognizing that supracondylar fractures are common accidental injuries. A lack of familiarity with medical conditions seen in nonabused children is the focus of the second case presented below.
A 2-month-old infant was found apneic in bed cosleeping with her mother. The infant was 1 of a pair of twins, and the mother and 3 of her other children (including the other twin) were sharing the same adult sleeping surface. The mother had delivered the twins while she was incarcerated for assault. The maternal aunt had cared for the twins until the mother's release from jail 3 weeks before the infant's death. The autopsy showed some red blood cells in the alveolar spaces (focal pulmonary hemorrhage) and some thymic petechiae. The cause of death and the mode of death were undetermined; this is the usual designation by the coroner in Los Angeles for deaths in which an infant is cosleeping.
The city attorney had asked for the case to be reviewed, because she wanted to charge the medical personnel involved with failure to report abuse. She believed that all cases of sudden unexpected death syndrome (SUDS)/sudden infant death syndrome (SIDS) required reporting to the child abuse hotline. The Health and Safety Code of the State of California requires that SIDS/SUDS be reported to the coroner's office, which then, on the basis of its investigation and findings, might report the case and thereby initiate an abuse investigation. The city attorney and law enforcement suggested that all pediatric deaths, even those felt to be natural, should be reported, because even children with chronic diseases may be murdered.
This suggestion seemed impractical and unnecessary to the medical personnel on the review board. If child abuse is suspected because of any physical findings or suspicious circumstances, then the physician should report the case. If there are no findings suggestive of abuse in an infant who has died unexpectedly, the case needs to be reported to the coroner. The coroner acts as an investigative officer to determine if, on the basis of his or her assessment, there is evidence of neglect or inflicted trauma. The hotline is then notified.
From an investigative perspective, this process may be slow, the ability of law enforcement to investigate the crime scene would be delayed, and compromised children may be left in an unsafe home with the perpetrator. However, it is unlikely that investigative agencies could investigate all pediatric deaths in a meaningful manner.
This particular case clearly illustrates the difference in perspective among the various agencies responsible for ensuring child safety. The medical providers want an orderly step-by-step procedure; law enforcement counters that delay may result in the loss of evidence for future prosecution, noting that, in fact, child abuse is as much a legal conclusion as it is a medical diagnosis. All are concerned about the potential safety of any other children who may be in the home.
A CONSULTATIVE APPROACH
One issue that has been noted as a barrier to reporting is lack of certainty. The physician asks, “What if I am wrong, and these findings are not the result of abuse but of some medical condition?” The CARES and other studies have shown that physicians may not report suspected abuse because they are uncertain if the child has been maltreated.1,8,9 The issue of uncertainty and fear of error is not unique to decisions related to child maltreatment. It is a factor that is intrinsic to medicine. Perhaps the notion of “reporting” serves as an impediment for some physicians. “Consulting” is a medical model with which physicians are familiar and comfortable. Some physicians may be more comfortable consulting about suspected abuse rather than reporting it. Consultation is significantly different from a legal model of reporting.
The current system of reporting is also framed as an all-or-none event; a consultative model might be more acceptable to health care providers who could then help families access services for children. The difficulty the primary care physician faces in reporting a family with whom he has a relationship needs to be understood and appreciated. The addition of medical consultation before reporting may help to remove the existing barriers and allow all children to have their pediatrician as their advocate. Perhaps now that our medical knowledge has expanded it is time to modify the all-or-none reporting system and permit greater medical assessment up front before the legal system is completely pulled into the investigation.
Some centers use telemedicine to obtain more expert advice. This modality was particularly popular in the 1990s for assessment of anogenital findings in cases of suspected child sexual abuse. Images were transferred over the Internet to allow physicians at 1 institution to assess the physical findings and render an opinion on cases being seen elsewhere. Telemedicine is one method that could potentially facilitate consultation.
The presence of the Internet and the rapidity of consultation could facilitate both the investigation and the reporting process. Currently, challenging cases related to child abuse are often shared through expert Listservs. However, the validity of opinions reached on the basis of preliminary and incomplete information raises legal, medical, and ethical concerns. How can we be sure that non–face-to-face consultation is sufficiently complete, accurate, and fair? It would be important for there to be national standards created and a determination made regarding the issue of whether such opinions can even be entered into evidence in court or represent hearsay. The issue of liability would also have to be clarified. Current statutes protect mandated reporters from civil suit, but the statutes do not cover experts (not the mandated reporter) who render an opinion and do not protect against malpractice suits. It would be important for consulting programs to recognize the limitations of non–face-to-face interactions.
The challenge of changing from a reporting to a consultative model will be whether physicians will access consultative services even if they are available. This model would need to be tested for utility before it could be recommended.
CREATION OF CENTERS OF EXCELLENCE: A PUBLIC HEALTH MODEL
The influx of new knowledge about child maltreatment has made the diagnosis of child maltreatment more complex. This complexity may contribute to physician uncertainty. With the development of child abuse pediatrics as a board-certified subspecialty of pediatrics, perhaps physicians will be more comfortable referring children with suspected abuse to other physicians with the expertise to evaluate and make determinations about the causality of the injury or other problem. This insight has led many to advocate for changes in the model of interdisciplinary work, including the adoption of better consultation and the establishment of regional multidisciplinary centers of excellence.
The proposed Health Child Abuse Research, Education, and Services (CARES) Network would establish 25 regional child abuse health care consortia, modeled after the University Centers for Excellence in Developmental Disabilities. Although there are many goals of the proposal, including research, education of professionals, and data collection, a goal related to the current discussion on reporting would be to provide readily accessible consultative services to community physicians and child protective services. One could speculate that crafting the center as part of a health surveillance network, by using a public health rather than a prosecutorial model, might be more palatable to some physicians. There is a stigma associated with the allegation of child abuse, but there is also a stigma associated with other conditions such as tuberculosis or sexually transmitted infections. In cases of tuberculosis or sexually transmitted infections, a report is filed that allows for additional evaluation. The report is not viewed as punitive but, rather, in the best interest of the individual's health and that of the community. There have been proposals to treat child abuse prevention and domestic violence from a public health perspective.
Some regions currently have organized child abuse efforts around academic health centers (eg, Florida, Washington, Missouri, New Jersey, and Utah). It would be important to determine the impact that such centers have had on reporting in their areas and whether there are lessons to be learned from their experience.
There are numerous modalities that can address the issue of physician underreporting. Education remains high on the list, starting with medical school and continuing into the arena of continuing medical education. Improving communication between child protective services and physicians is also needed, as is the need to address confidentiality and sharing information. Moving away from an all-or-none mandated-reporting model to a consultative approach using centers of excellence and telemedicine would also create an environment that is more closely aligned with the medical approach to health care.
- Accepted May 28, 2008.
- Address correspondence to Carol D. Berkowitz, MD, Harbor/UCLA Medical Center, 1000 W. Carson St, Box 437, Torrance, CA 90502-2004. E-mail:
The author has indicated she has no financial relationships relevant to this article to disclose.
- ↵Flaherty EG, Sege RD, Griffith JL, et al. From suspicion to report: primary care clinician decision-making. The Child Abuse Recognition Experience Study Research Group. Pediatrics.2008;122 (3):611– 619
- ↵Jones R, Flaherty EG, Binns HJ, et al. Clinicians' description of factors influencing their reporting of suspected child abuse: report of the Child Abuse Reporting Experience Study Research Group. Pediatrics.2008;122 (2):259– 266
- ↵Markenson D, Tunik M, Cooper A, et al. A national assessment of knowledge, attitudes, and confidence of prehospital providers in the assessment and management of child maltreatment. Pediatrics.2007;119 (1). Available at: www.pediatrics.org/cgi/content/full/119/1/e103
- ↵Trokel M, Wadimmba A, Griffith J, Sege R. Variation in the diagnosis of child abuse in severely injured infants [published correction appears in Pediatrics. 2006;118(3):1324]. Pediatrics.2006;117 (3):722– 728
- Copyright © 2008 by the American Academy of Pediatrics