343 - Identifying Prior Concerning Injuries in Children with Abusive Fractures

Publication Number: 343.5263

Noelle Tran, Children's Mercy Hospital, Kansas City, MO, United States; Maria C. Antonucci, UPMC Childrens Hospital of Pittsburgh, Pittsburgh, PA, United States; Angela Bachim, Baylor College of Medicine, Houston, TX, United States; Colleen J. Bressler, Medical University of South Carolina College of Medicine, Charleston, SC, United States; Kristine A. Campbell, University of Utah School of Medicine, Salt Lake City, UT, United States; Nara Cho, Nationwide Children's Hospital, Columbus, OH, United States; Kristin G. Crichton, Nationwide Children's Hospital, Columbus, OH, United States; Lori Frasier, Penn state hershey, Hershey, PA, United States; Nancy S. Harper, University of Minnesota Masonic Children's Hospital, Minneapolis, MN, United States; Sara Kilbride, Children's Mercy Hospitals and Clinics, kansas city, MO, United States; Natalie Laub, University of California, San Diego School of Medicine, San Diego, CA, United States; Jan Leonard, Children's Hospital of Philadelphia, Philadelphia, PA, United States; Daniel Lindberg, University of Colorado Anschutz Medical Campus, Denver, CO, United States; Joanne N. Wood, Children's Hospital of Philadelphia, Philadelphia, PA, United States; Jim Anderst, Children's Mercy Hospitals and Clinics, Kansas City, MO, United States

Background: Existing data characterizes the relationship between prior concerning injuries (PCIs) in infants preceding abusive head trauma. Little data exists on the relationship between PCIs and abusive fractures.

Objective: We aim to compare the frequency and number of PCIs known to a Child Abuse Pediatrician (CAP) at the time of evaluation between children with abusive and non-abusive fractures.

Design/Methods: This was a retrospective, case-control study using data from CAPNET, a research network of 10 Child Protection Teams, over a 27-month period. Study subjects were less than 4 years old, had fractures, in-person CAP consultation, and did not have intracranial hemorrhage. The prevalence and number of PCIs was compared between abuse cases and non-abuse controls. Abuse was defined in two different ways to help mitigate circular reasoning. Analysis 1 compared the prevalence and number of PCIs among subjects with a score of 5-7 on a 7-point scale to identify likelihood of abuse cases to subjects with a score of 1-2. For Analysis 2, a combination of mobility, history, presence of concomitant suspicious injuries (CSIs), and fracture specificity for abuse were used a priori identify abused cases and controls. CSIs were defined as bruising and/or petechiae specific for abuse, subconjunctival hemorrhage outside of birth, burns, multiple fractures, fractures of different ages, and abdominal injury. Unadjusted odds ratios and 95% confidence intervals were calculated.

Results: There were 1339 study subjects with a median age of 8 months (interquartile range 3.8 to 13.9 months). The prevalence of PCIs in the abused groups in Analyses 1 and 2, respectively, were 30% and 22%. In Analysis 1, abused subjects had higher odds of any PCIs than the non-abused group (OR 4.5, 95% CI 3.2-6.1). This association remained in subjects with one (OR 3.4, 95% CI 2.4-4.9) and multiple PCIs (OR 11.7, 95% CI 5.8-23.8). In Analysis 2, any PCIs were also more common in immobile (OR 5.9, 95% CI 3.1-11.3) and mobile abuse group (OR 3.0, 95% CI 1.6 to 5.6) when compared to same level of mobility non-abuse groups. This association remained in immobile subjects with one (OR 5.1, 95% CI 2.5-10.4) and multiple PCIs (OR 9.5, 95% CI 2.2-40.0). Similarly, the association remained in mobile subjects with one (OR 2.1, 95% CI 1.1-4.0) and multiple PCIs (OR 33.3, 95% CI 1.9-572.6).

Conclusion(s): Children who suffered abusive fractures were more likely to have PCIs. This data supports the idea that abuse is commonly a chronic, escalating disease, and that improved early recognition has the potential to prevent significant injuries.

Table 1. Prevalence/type of prior concerning injuries by analysis
This table shows the prevalence and types prior concerning injuries found in each analysis

Figure (1): Patient population separated into Analysis 1 and Analysis 2
This table shows how the same patient population was separated into Analysis 1 and Analysis 2.