br Methods br Study design br This
This observational study was designed to evaluate the association between CT and treatment at different trauma center types among children hospitalized for injury in the U.S. We evaluated differences in overall and body-region specific CT scans across different trauma center types. Stratified analyses in children with different injury severities, mechanisms, and age were performed. This work was approved by the Research Ethics Board at Sunnybrook Health Sciences Center, Toronto, Ontario.
Data sources and participating centers
Data were derived from 130 U.S. trauma centers participating in the American College of Surgeons (ACS) Trauma Quality Improve-ment Program (TQIP) [17,18]. TQIP is a voluntary performance improvement program that was created to provide level I and II trauma centers with feedback on risk-adjusted outcomes for quality-improvement purposes . Data pertaining to TQIP centers is obtained from the National Trauma Databank, which utilizes the National Trauma Data Standard (NTDS)  to standardize how abstractors capture data. Inclusion criteria for entry into TQIP require an International Statistical Classification of Diseases, Ninth Revision (ICD-9CM) diagnosis of 800–959 exclud-ing late effects of trauma (905–909), superficial injuries (910–924), or foreign bodies (930–939) . TQIP data quality is evaluated by data checks, external validation, and specialized training of data collection personnel .
We identified patients aged 18 or younger who were hospital-ized at a TQIP center from January 2010 to December 2013. We included children with blunt trauma who had a minimum Abbreviated Injury Score  (AIS) 2 in any body region. We only included children transported directly from the scene, as we could not ascertain whether or not children transferred from another center had received a CT scan prior to arrival. We excluded patients with superficial injuries, those who were transferred to another hospital or home directly from the emergency depart-ment, and those who were dead on arrival. At the center-level, hospitals that had no recorded CT scans during four years were excluded (five centers), suggesting that they 2-Guanidinoethylmercaptosuccinic Acid were not routinely capturing imaging data as per the NTDS . The NTDS requires that abstractors capture all procedures, including CT, performed in the emergency department, ICU, ward, or radiology department that were essential to the diagnosis, stabilization, or treatment of the patient’s specific injuries .
Trauma center and outcome classification
Trauma centers were classified as ATC, MTC, or PTC. We defined an ATC as any center having adult ACS verification or adult state designation and no pediatric qualifications . PTC was defined as any center exclusively having either pediatric ACS verification or pediatric state designation. MTC was defined as any center having both adult and pediatric ACS verifications or state designations .
Children with CT scans were identified using ICD-9  procedure codes (Appendix 1). CT scans zone of physiological stress occurred within 24 h of arrival were included in our analyses. Outcomes were categorized into overall CT (head OR torso OR other), head CT, chest CT, or abdomen/pelvis CT to allow for the evaluation of each type of scan.
Baseline characteristics were summarized as counts and compared between those with and without CT scans. Overall and body-region specific CT rates were compared across different trauma center types using Chi-Square analyses. Data were plotted using bar graphs to visually depict differences in CT scan rates across centers types (exact numbers associated with percentages on bar graphs are in Appendix 2). Stratified analyses in children with high injury severity (Injury Severity Score  (ISS) 16) and low injury severity (ISS < 9) were undertaken to determine how differences in CT usage across centers varied in these subsets of injured children. To limit the potential effects of confounding, we performed additional stratified analyses in more homogenous cohorts of injured children - those injured as a result of motor vehicle collision (MVC) or a fall. Additional stratified analyses in different pediatric age groups  (young children (0–5 years), children (6–12 years), adolescents (13–18 years)) were performed
to evaluate which group may be particularly vulnerable to differential CT rates across trauma centers. All statistical analyses were performed using SAS, version 9.3, Cary, NC. All tests were two-sided with p-values <0.05 considered statistically significant. Lastly, in order to assess the variation of CT rates across individual centers within each trauma center type category we produced a box plot including the mean and median CT rate across each trauma center type, as well as interquartile range, maximum, and minimum.