A neurodevelopmentally impaired 6-year-old girl is referred for examination of asymptomatic skin lesions on the right forearm due to concern for possible child abuse. What's the diagnosis?
You are asked to provide an urgent evaluation of asymptomatic skin lesions on the right forearm of a neurodevelopmentally impaired 6-year-old girl (Figure 1). The skin findings were first noted by a radiology nurse while the child was undergoing an outpatient MRI; the patient was referred to a pediatrician because of concern of possible child abuse.
Diagnosis: Kitten bites misdiagnosed as child abuse
The lesions were apparently fresh and probably occurred that morning because the child’s mother first became aware of them when the nurse inquired about the wounds. The mother reported that the family had a rambunctious 8-month-old kitten in the home that loved to play and sleep with all the family members, particularly the patient. The mother had similar lesions from a week ago, but the puncture wounds had healed, leaving flat pinpoint and surrounding ring purpura (Figure 2). The mother reported that except for the time of the bite, the lesions were not painful and healed quickly.
One week prior to discovery of the child’s new lesions, the mother observed the kitten “attacking” the child’s leg and causing scratches in addition to claw and bite punctures. This behavior had not previously occurred. The child was not in distress, and the lesions required no additional care. The mother obtained and placed claw caps on the kitten.
Unfortunately, at initial evaluation in the pediatric emergency department (ED), the history of the kitten as the cause of the injury was rejected, and there was concern for the injury occurring in the setting of child abuse because of the wound pattern. Specifics about what made the wound pattern concerning for abuse were not given.
The child was transferred to a children’s hospital. The diagnosis of cat bite was not considered by the general pediatric staff. Two child abuse pediatricians (CAPs) rejected the mother’s history of the kitten being in close contact with the child and concluded that because the wounds could not have been self-inflicted, there was no plausible history and, therefore, the child’s injuries were consistent with child abuse.
Apparently, none of the physicians or nurses were familiar with wounds from cat and dog bites.
The literature suggests that approximately half of all people in the United States will be bitten by either an animal or another human being at some point.1,2 Most of the estimated 2 million annual mammalian bite wounds1 are minor, and the victims never seek medical attention. The lack of medical consultations in these cases could perhaps explain why many primary care physicians and some urgent care and ED physicians are not knowledgeable of the many presentations of animal bites.
Cat bites are the second most common form of mammalian bite. Ginsburg et al note that approximately 450,000 cat bites occur annually in the United States, nearly all of which are inflicted by household cats.3 Many of these bites are not reported.
The general literature on feline behavior notes that even domesticated cats can exhibit their instinctual predatory nature, which includes pursuing and possessing moving objects. This behavior is more unrestrained in kittens; they can be more boisterous and exhibit such predatory behaviors as part of playing.4
Animal bites can present with different lesions, such as punctures, avulsions, tears or lacerations, crush injuries, and/or abrasions.5 Also, puncture wounds can appear with irregular spacing like those seen on the patient’s mother (Figure 2). The literature regarding dog bites shows that similar irregular spacing can occur from their bites as well.6
A potential cause of confusion between animal bites and other causes of bruising is the peripuncture blue discoloration visible in the patient’s and mother’s wounds. The literature does not address the causes of these discolorations. Bilo et al attribute this to trauma from jaw pressure.6 Another consideration may be an inflammatory process from an animal salivary protein or its microbiota that can cause small-vessel inflammation and diapedesis of red blood cells.
Following this injury, the mother removed the kitten from their home. The child was asymptomatic, and no treatment was needed. The lesions cleared within a week.
It is important to analyze the cause of this misdiagnosis.
The CAPs who evaluated the skin findings rejected the mother’s history based on their erroneous conclusion that the pattern of skin lesions was not compatible with kitten bites. They did this without producing any objective supporting facts and without the support of medical literature. Consultants were not used. These physicians did not describe any abusive mechanisms that could have caused the child’s skin lesions. A differential diagnosis was not offered. The lack of putting forth a diagnosis based upon the medical facts of the case makes the diagnosis of abuse in this case a default diagnosis that lacked objective evidence and proper scientific reasoning.7-9
In a formal inquiry regarding unsubstantiated diagnoses of abuse, Justice Stephen T. Goudge, who served on the Court of Appeal for Ontario, Canada, for many years, stated that the use of a default diagnosis is unscientific and does not represent the evidence-based scientific approach. Default diagnoses infer rather than prove causality. They lack transparency, which prevents independent reviewability. Furthermore, a default diagnosis can be produced from confirmation bias.10
A similar term for a default diagnosis is diagnosis of exclusion (DOE). Without objective support for a given diagnosis, a DOE is an educated guess that could be unreliable and unsafe.11
Because no objective evidence of inflicted trauma was appreciable in this case and the specific abusive mechanisms that could have caused the child’s lesions were not given, “undetermined”—rather than abuse—would have been a proper default diagnosis by the CAPs.
The mother was accused of abusing the child. Custody was removed from the mother, and the child was placed in foster care. The following day, the foster parents brought the child back to the ED because of the medical complexities of the child’s care. The child was then released to the biological mother’s home; the kitten had already been removed from the home. However, the foster mother still retained custody, so the mother had to be supervised by the child’s step-grandmother.
A second child abuse consultation with a CAP who presented no object evidence rejected the mother’s history and experience with their kitten and affirmed, by default, the erroneous diagnosis of child abuse.
Three months later, the mother’s attorney presented to the family court officer and prosecutor a comprehensive medical report objectively confirming that the skin lesions were due to the kitten bite/paw punctures. They dismissed the case, and there was no adjudication of abuse on the mother’s record.
Bjӧrk et al states that a physician’s private values can modulate reasoning and may play a significant role in the diagnostic process.12 Physicians’ biases are paramount considerations in evaluating possible cases of child abuse because those cases can involve significant medical debate and societal pressures.
We believe that all physicians must insist upon ethical and medical precision in diagnoses of child abuse so that unequivocally abused children will be removed from abusive environments and innocent families will not be subjected to the psychological, financial, and societal stresses associated with being accused of abuse.
We agree with the statement by Findley et al—regarding a diagnosis of abuse, we should all strive to “get it right.”13