223 - Racial Disparities in Diagnostic Imaging and Diagnosis of Abuse among Children 0-2 with Extremity Fractures

Poster Number: 223
Publication Number: 223.401

Margaret Tashjian, University of Colorado School of Medicine, Aurora, CO, United States; Marion R. Sills, University of Colorado School of Medicine, Denver, CO, United States; Matthew Hall, Children's Hospital Association, Lenexa, Kansas, KS, United States; Daniel Lindberg, University of Colorado School of Medicine, Denver, CO, United States; Henry Puls, Children's Mercy Hospitals and Clinics, Kansas City, MO, United States; James Anderst, Children's Mercy Hospital, KANSAS CITY, MO, United States; Kavita Parikh, Children's National Health System, Washington DC, DC, United States; Joanne Wood, The Children's Hospital of Philadelphia (CHOP) - - Philadelphia, PA, Philadelphia, PA, United States; Megan E. Collins, Children's Mercy Hospitals and Clinics, Mission, KS, United States; Denise C. Abdoo, Children's Hospital Colorado, Aurora, CO, United States

Background: Previous studies have demonstrated racial disparities in clinicians’ evaluation of child abuse, with some studies showing both increased use and decreased yield of occult injury tests like skeletal surveys. Most of these were published before recent national guidelines (2009, 2015) suggested a standardized approach to occult injury testing.

Objective: To measure racial disparities in skeletal surveys (SS) performed and diagnoses assigned during hospital-based acute care of children less than 24 months old with long-bone fractures.

Design/Methods: This retrospective study used the Pediatric Health Information System (PHIS), which incorporates billing data from 49 tertiary children’s hospitals, to identify children with an emergency or inpatient visit that included an ICD-10-CM code for extremity fracture between 2017-2019. Rates of ordering SS and rib fractures identified were compared among three race and ethnicity groups: Hispanic (H), Non-Hispanic Black (NHB), and Non-Hispanic White (NHW). Rib fractures were measured as a proxy for abuse because they are the most common occult injury identified on SS, rarely a presenting complaint, and highly specific for non-accidental injury. Generalized estimating equations with p < 0.05 for significance were used during multivariable analysis to calculate adjusted odds ratios (AOR) for age, sex, payor, and hospital.

Results: Among 20,842 children diagnosed with an extremity fracture, the median age was 16 months, 48% were female, and 55% had government insurance. Racial distribution included 23% H, 18% NHB, and 46% NHW (Table 1). Overall, 21% received a SS, by subgroup: 31.3% in NHB, 20.9% in NHW, and 15.2% in H patients. A rib fracture was detected in 17% (Table 2). Pairwise AOR for ordering a SS differed by racial groups: 1.97 for NHB versus H (95% confidence interval (CI) 1.74, 2.23), 1.17 for NHB versus NHW (CI 1.05, 1.31), and 0.59 for H versus NHW (CI 0.53, 0.67). Among children who received a SS, we identified no difference in AOR of rib fractures by racial group: 0.86 for NHB versus H (CI 0.57, 1.3), 0.83 for NHB versus NHW (CI 0.6, 1.14), and 0.96 for H versus NHW (CI 0.65, 1.41) (Table 3).Conclusion(s): In children with extremity fractures, we identified race and ethnicity disparities in AORs of ordering skeletal surveys but found no difference in the frequency of rib fractures identified. These findings suggest that disparities in SS use may be guided by unmeasured, confounding factors that are associated with both abuse risk and race.
Table 1: Study Sample Demographics
Table 2: Diagnostic Imaging and Diagnoses Assigned by Race and Ethnicity: Number (Percentage) of Study Patients (n = 20842)