CT with 3-D reconstruction is equivalent to skull radiographs in identifying skull fractures. When a head CT is indicated, skull radiographs add little diagnostic value.

Child physical abuse is a leading cause of traumatic injury and death in children. In 2017, child abuse was responsible for 1688 fatalities in the United States, of 3.5 million children referred to Child Protection Services and 674,000 substantiated victims. While large referral hospitals maintain teams trained in Child Abuse Pediatrics, smaller community hospitals often do not have such dedicated resources to evaluate patients for potential abuse. Moreover, identification of abuse has a low margin of error, as false positive identifications lead to unwarranted separations, while false negatives allow dangerous situations to continue. This context makes the consistent detection of and response to abuse difficult, particularly given subtle signs in young, non-verbal patients. Here, we describe the development of artificial intelligence algorithms that use unstructured free-text in the electronic medical record—including notes from physicians, nurses, and social workers—to identify children who are suspected victims of physical abuse. Importantly, only the notes from time of first encounter (e.g.: birth, routine visit, sickness) to the last record before child protection team involvement were used. This allowed us to develop an algorithm using only information available prior to referral to the specialized child protection team. The study was performed in a multi-center referral pediatric hospital on patients screened for abuse within five different locations between 2015 and 2019. Of 1123 patients, 867 records were available after data cleaning and processing, and 55% were abuse-positive as determined by a multi-disciplinary team of clinical professionals. These electronic medical records were encoded with three natural language processing (NLP) algorithms—Bag of Words (BOW), Word Embeddings (WE), and Rules-Based (RB)—and used to train multiple neural network architectures. The BOW and WE encodings utilize the full free-text, while RB selects crucial phrases as identified by physicians. The best architecture was selected by average classification accuracy for the best performing model from each train-test split of a cross-validation experiment. Natural language processing coupled with neural networks detected cases of likely child abuse using only information available to clinicians prior to child protection team referral with average accuracy of 0.90±0.02 and average area under the receiver operator characteristic curve (ROC-AUC) 0.93±0.02 for the best performing Bag of Words models. The best performing rules-based models achieved average accuracy of 0.77±0.04 and average ROC-AUC 0.81±0.05, while a Word Embeddings strategy was severely limited by lack of representative embeddings. Importantly, the best performing model had a false positive rate of 8%, as compared to rates of 20% or higher in previously reported studies. This artificial intelligence approach can help screen patients for whom an abuse concern exists and streamline the identification of patients who may benefit from referral to a child protection team. Furthermore, this approach could be applied to develop computer-aided-diagnosis platforms for the challenging and often intractable problem of reliably identifying pediatric patients suffering from physical abuse.

Rib fractures are considered highly suspicious for nonaccidental injury in the pediatric clinical literature; however, a rib fracture classification system has not been developed. As an aid and impetus for rib fracture research, we developed a concise schema for classifying rib fracture types and fracture location that is applicable to infants. The system defined four fracture types (sternal end, buckle, transverse, and oblique) and four regions of the rib (posterior, posterolateral, anterolateral, and anterior). It was applied to all rib fractures observed during 85 consecutive infant autopsies. Rib fractures were found in 24 (28%) of the cases. A total of 158 rib fractures were identified. The proposed schema was adequate to classify 153 (97%) of the observed fractures. The results indicate that the classification system is sufficiently robust to classify rib fractures typically observed in infants and should be used by researchers investigating infant rib fractures.

Pediatric patients ≤24 months of age presenting to the ER in delayed fashion with scalp swelling after minor head trauma-who were otherwise nonfocal on examination-did not require surgical intervention and did not experience any neurologic decline. Further radiographic investigation did not alter neurosurgical management in these patients; however, it should be noted that workup for child abuse and social care may have been influenced by CT findings, suggesting the need for the future development of a clinical decision-making tool to help safely avoid CT imaging in this setting.

Background In infants and children with fractures from an unclear cause, Osteogenesis Imperfecta (OI) is often included as a potential etiology. In infants and children with OI there exists a gap in the published literature regarding the fracture pattern seen at the time of diagnosis. As an additional aid to the diagnosis of OI, we sought to characterize the fracture patterns in infants and children at the time of their diagnosis. Methods We performed a retrospective chart review of a series of infants and children under 18 years of age who have the diagnosis of OI (any type) from a single institution. Results We identified 68 infants and children with OI: 23 (34%) type 1, 1(2%) type 2, 17(25%) type 3, 24(35%) type 4 and 3(4%) unknown type. A family history of OI is present in 46% of children. Forty-nine (72.0%) patients were diagnosed solely on clinical characteristics, without genetic or fibroblast confirmation. Rib fractures were noted in 21% of the subjects with none being identified during infancy. The number of fractures identified at diagnosis ranged from 1 to >37 with 7 (10%) having more than 2 fractures. All subjects with more than 2 fractures were diagnosed prenatally or in the immediate newborn period. Seventeen (25%) infants were diagnosed after 1 week of age but prior to 12 months of age. None of these infants had either rib fractures or more than 1 fracture at the time of diagnosis. Conclusion The majority of children diagnosed with OI are diagnosed by clinical features alone. The fracture pattern at the time of diagnosis in OI is variable with 10% having more than 2 fractures. The diagnosis of OI was made in utero or at delivery in 43% of children. Multiple rib fractures in an infant would be an unexpected finding in OI. Level Of Evidence Level III