Retrospective Review of Hematologic Evaluation in Children with Suspected Non-Accidental Injury, A First Step towards Evidence Based Guidelines.

Child maltreatment is a frequent cause of injury in the United States, occurring in approximately 695,000 children per year. Bleeding disorders can exacerbate and be confused with non-accidental injury (NAI); both of these diagnostic errors are life altering for the child and family. Despite the high incidence of child abuse and the independent relative high incidence of bleeding disorders, the optimal hemostatic evaluation is unclear for children who may be victims of NAI. Expert hematologic opinion recommends a multi-tiered approach, first investigating for common bleeding disorders, and subsequently investigating for rare defects in the coagulation and fibrinolytic pathways, if necessary. Further research is needed to develop evidence-based guidelines for the evaluation of bleeding disorders in children who may be victims of NAI.

To review and analyze a five-year history of the hematologic investigation of children who presented with bleeding and/or bruising that was suspicious for NAI at Vanderbilt Children's Hospital (VCH). Our hypothesis is that there is a lack of a systematic approach for the hemostatic evaluation of children who present with bleeding symptoms and concern for NAI.

A retrospective cohort study design was employed. ICD-9 codes for NAI (995.5, 995.50, 995.54, 995.55, 995.59) were used. 354 medical records from 2007 – 2011 were reviewed and screened for inclusion and exclusion criteria, resulting in 198 fully evaluable patients. Medical records were then queried for details of clinical and laboratory evaluation that occurred at the initial presentation concerning for NAI. We defined a basic hematologic evaluation as a CBC, PT and PTT; and a comprehensive hematologic evaluation as a CBC, PT, PTT, factor VIII, IX and XI activity and von Willebrand evaluation. Data was analyzed using SPSS; statistical analysis was performed using frequencies and Chi-Square analysis.

The mean age for the studied population was 445 days (max 4687 days, minimum 6 days); 37% were male. Bleeding symptoms included intracranial hemorrhage (ICH) (40%) and bruising (58% without associated ICH, 73% with and without associated ICH), with approximately 60% demonstrating additional non-hematologic symptoms (i.e. fractures, burns). Hematologic laboratory tests performed included CBC in 66%, PT in 58%, and PTT in 55%; factor activity levels in 12% (primarily consisted of factors VIII and IX); and von Willebrand disease evaluation in 12% of subjects. Table 1 shows the percentage of laboratory tests obtained in the patients based on symptoms at presentation.

Abnormal coagulation labs were seen in 33% of performed PT and 55% of PTT tests; 55% of abnormal PTs and 44% of abnormal PTTs repeated. Our defined basic evaluation was completed in 79% of patients with ICH and 36% of patients without ICH (p <0.00). Our defined comprehensive evaluation was completed in 9% of patients with ICH and 2% of patients without ICH (p 0.32). Only 1% (2 of 198) children in our cohort were diagnosed with hematologic disorders that explain their increased bleeding symptoms.

Complete hematologic evaluation of children who present with bleeding symptoms and concern for NAI is inconsistent. While some children with other findings diagnostic for NAI may not require a hematologic work-up from a medico-legal perspective, it is still prudent to consider common bleeding disorders as a potential contributing factor in the severity of symptoms. At VCH, laboratory evaluation was obtained with greater frequency in patients with hematologic symptoms only. Given the variability of tests obtained and the disparity between expert opinion and our historical evaluation, further investigation into the optimal evaluation of these patients is warranted at this time. A prospective cohort study would allow comprehensive evaluation of all children suspected of NAI resulting in a clear understanding of laboratory abnormalities and incidence of bleeding disorders.

No relevant conflicts of interest to declare.