Date of Award

January 2015

Document Type


Degree Name

Medical Doctor (MD)



First Advisor

Melinda S. Sharkey

Subject Area(s)




Victoria V. Tate1, Kelsey B. Loeliger2, Julie R. Gaither2, Daniel R. Cooperman1, Lauren J. Ehrlich3, Cicero Silva3, Chang-Yeon Kim1, Rebecca Moles4, John M. Leventhal4, and Melinda S. Sharkey1

1Department of Orthopaedics and Rehabilitation, Yale University School of Medicine, New Haven , CT; 2Department of Epidemiology and Public Health, Yale University School of Medicine, New Haven, CT; 3Department of Diagnostic Radiology, Yale University School of Medicine, New Haven, CT; 4Department of Pediatrics, Yale University School of Medicine, New Haven, CT.

The purpose of this study was to determine the patterns, characteristics, and incidence of abusive fractures in children less than three years of age that presented to the Yale New Haven Hospital pediatric emergency department from 2007 to 2010. Our second objective was to investigate changes and trends in abusive fractures over four time periods (1979-1983, 1991-1994, 1999-2002, and 2007-2010) at a level 1 pediatric trauma center.

We conducted a retrospective chart review for all children <36 months who were diagnosed with a fracture after presenting to the emergency department of a single level 1 trauma center between 2007 and 2010. For each patient the following data were collected: social demographics, mechanism of injury, fracture(s) characteristics, treatment, involvement of social services, and outcome. Each case was rated and scored independently for likelihood of the fracture(s) being caused by abuse on an established 7-point scale by two pediatric orthopaedic surgeons, two child abuse experts, two pediatric radiologists, and one senior medical student. When there were disagreements in a rating, the raters discussed the case and came to a consensus. A best-fit multivariable logistic regression model was created to determine predictors associated with abusive ratings in the 2007-2010 data sample. These data were compared to previously published data from the same level 1 pediatric trauma center using the same methodology. Chi-square was used to compare proportions; Cochran-Armitage trend test was used to evaluate temporal trends.

From 2007 to 2010 there were 593 children with 690 fractures, including 33 children with 58 abusive fractures. Age <12 months old, a delay in presentation, presence of bruising, previous injuries or concerns of abuse, and history of DCF involvement were the highest predictors of a fracture from abuse. The incidence of children <36 months of age presenting with an abusive fracture in this region was 2.8 per 10,000. The incidence of children presenting with an abusive fracture per ED visit was 3.2 per 10,000.

When compared to the previous three time periods the proportion of children with abusive fractures decreased from 22.5% in 1979-83 to 5.6% in 2007-2010 (p<.001). However, the incidence of children with abusive fractures both regionally (P=0.837 for trend) and per ED visit (P=0.189 for trend) remained stable over time. Abusive fractures are more likely to occur in children <12months in all time periods (p=.004) and rib fractures were the only fracture type that had >90% association with abuse in all time periods.

In this single institution review of fractures in children less than 3 years old, younger age is consistently one of the most important determinants in diagnosing abusive fractures, and while they are not pathognomonic, rib fractures in young children should instigate a high suspicion of abuse. The regional incidence of young children presenting with abusive fractures has not changed significantly over the past 30 years.


This thesis is restricted to Yale network users only. This thesis is permanently embargoed from public release.