a b s t r a c t

Background: Computed tomography (CT) is frequently used to identify intra-abdominal injuries in children with blunt abdominal trauma (BAT). The Pediatric Emergency Care Applied Research Network proposed a prediction rule to identify children with BAT who are at very low risk for clinically-important intra-abdominal injuries (CIIAI) in whom CT can be avoided.

Objective: To determine the sensitivity of this prediction rule in identifying patients at very low risk for CIIAI in our pediatric trauma registry.

Methods: Retrospective review of our institutional trauma registry to identify patients with CIIAI. CIIAI included cases resulting in death, therapeutic intervention at laparotomy, angiographic embolization of intra-abdominal arterial bleeding, blood transfusion for intra-Abdominal hemorrhage, and administration of intravenous fluids for two or more nights for pancreatic or gastrointestinal injuries. Patients were identified using ICD diagnosis and procedure codes. Kappa was calculated to evaluate inter-reviewer agreement.

Results: Of 5743 patients, 133 (2.3%) had CIIAI. 60% were male and the mean age was 8 (SD 4.4) years. One patient with CIIAI met the proposed very low risk criteria, resulting in a prediction rule sensitivity of 99%, 95% CI [96-100%]. This patient also had extra-abdominal arterial bleeding requiring revascularization, offering an alter- native reason for transfusion. Kappa was 0.85, 95% CI [0.82, 0.89], indicating strong inter-rater agreement.

Conclusions: One out of 133 patients with CIIAI met very low risk criteria based on the PECARN prediction rule. This study supports the PECARN clinical prediction rule in decreasing CT use in pediatric patients at very low risk for CIIAI.

(C) 2018

Introduction

unintentional injury is the leading cause of death in pediatric pa- tients, causing approximately 6000 deaths annually in children between 1 and 19 years old [1]. Injuries in children younger than 15 years of age account for over 7 million emergency department visits per year, with the majority of these visits for unintentional falls, bicycle, and motor ve- hicle collisions (MVCs), which are frequently associated with blunt ab- dominal trauma (BAT) [2,3]. Computed tomography (CT) is the criterion standard for identifying intra-Abdominal injury in patients with BAT, and is thus frequently used in the evaluation of both adult and Pediatric trauma patients [4,5].

? This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

* Corresponding author at: Division of Pediatric Emergency Medicine, Monroe Carell Jr. Children’s Hospital at Vanderbilt, 2200 Children’s Way, Suite 1025, Nashville, TN 37232, USA.

E-mail address: [email protected] (A.A. Vukovic).

¹ Pediatric Emergency Medicine Associates, Children’s Hospital at Erlanger, 910 Blackford Street, Chattanooga, TN, 37403 (present address).

Compared to adults, children have a higher predicted rate of devel- oping radiation-induced fatal cancer, with an estimated lifetime attrib- utable risk of death from cancer up to 0.14% per abdominal CT scan [6,7]. Although Pediatric institutions have lower rates of CT utilization in children, most pediatric patients are initially evaluated in general emergency departments [8-11]. Additionally, young children with mild to moderate injuries undergo more abdominal CTs than their severity-matched adult counterparts [12]. The benefit of identifying clinically important intra-abdominal injury (CIIAI) must be balanced with the risk of radiation-induced malignancy from abdominal CT. Sev- eral prediction rules have been proposed to risk-stratify these patients and to identify low-risk children in whom abdominal CT can be avoided [13-20].

In 2013, Holmes et al. published a prospective study through the Pe- diatric Emergency Care Applied Research Network (PECARN), which of- fered a clinical prediction rule to identify a cohort of children with BAT who are at very low-risk of CIIAI [21]. The rule uses seven readily avail- able clinical variables based on history and physical exam: evidence of abdominal wall trauma or seatbelt sign, Glasgow coma score, abdominal tenderness, thoracic wall trauma, abdominal pain, breath sounds, and

https://doi.org/10.1016/j.ajem.2018.11.031

0735-6757/(C) 2018

vomiting. 12,044 patients were enrolled from 20 participating pediatric emergency departments (PEDs). The prediction rule is 97% sensitive (95% CI [93.7, 98.9]) in identifying patients at very low-risk of having CIIAI [21]. In that study, only 0.1% of patients who fulfilled the very low-risk criteria were found to have an intra-abdominal injury requir- ing acute intervention, resulting in a less favorable risk-benefit ratio in relation to radiation-induced malignancy. Our objective was to deter- mine the sensitivity of the PECARN prediction rule by determining the proportion of patients with CIIAI who failed the proposed very low- risk criteria and to identify patients who would not have been identified using the prediction rule.

Methods

Study design

We performed a retrospective review of pediatric patients evaluated and treated in a pediatric emergency department in an academic, ter- tiary care children’s hospital and American College of Surgeons (ACS) level 1 trauma center from January 2011 to August 2016. The pediatric trauma service maintains a registry of all patients treated for traumatic injuries, excluding minor extremity trauma. Patients are identified by International Classification of Diseases (ICD) diagnosis and procedure codes, and trauma team members manually abstract data from the charts. This includes, but is not limited to, date of injury, trauma level, initial vital signs, transfusions required, length of hospital stay, and dis- position. We used this registry to identify patients with CIIAI requiring acute intervention (defined below) and reviewed these charts for the presence of very low-risk criteria. This study was approved by the insti- tutional review board (Protocol #161912) prior to commencement.

Study setting and population

All pediatric trauma patients b16 years of age are evaluated and treated in the children’s hospital, as are those from 16 to 18 years of age who do not meet adult level one or two trauma activation criteria. The institutional trauma registry includes pediatric trauma patients treated in and admitted to the pediatric hospital, as defined by ICD codes in Appendix A. During the study period, the institution transitioned from ICD-9 to ICD-10 coding.

Study protocol and measurements

We initially sought to identify any patient within the trauma registry who sustained any intra-abdominal, spinal or pelvic injury after sustain- ing blunt abdominal trauma. To accomplish this, we narrowed our list to all patients in the registry with ICD codes in Appendix B. We then sought to specifically identify those patients who sustained CIIAI as de- fined below. To achieve this, we further narrowed our patient list to only those with applicable ICD procedure codes, administration of blood products, hospitalization for two or more nights in the hospital, or death were included (Appendix C). Each of these remaining charts was reviewed and abstracted.

PECARN proposed a prediction rule to identify children with BAT who are at very low risk for CIIAI in whom CT can be avoided. In this study, intra-abdominal injury and acute intervention were defined using criteria from the original PECARN study [21]. Intra-abdominal in- jury included any radiographically- or surgically-apparent injury to the following structures: spleen, liver, urinary tract (from the kidney to the urinary bladder), gastrointestinal tract (including the bowel and associ- ated mesentery from the stomach to the sigmoid colon), pancreas, gall- bladder, adrenal gland, intra-abdominal vascular structure, or traumatic fascial defect (traumatic abdominal wall hernia). Acute intervention was defined by an intra-abdominal injury associated with any of the fol- lowing: death caused by the intra-abdominal injury, a therapeutic inter- vention at laparotomy, angiographic embolization to treat bleeding

from the intra-abdominal injury, blood transfusion for anemia as a re- sult of hemorrhage from the intra-abdominal injury, or administration of intravenous (IV) fluids for two or more nights in patients with pan- creatic or gastrointestinal injuries. Therapeutic laparotomy was defined as any surgical intervention to treat an intra-abdominal injury. Thus, for this research we defined CIIAI as (1) Intra-abdominal injury that re- quired (2) Acute intervention, both defined in accordance with the PECARN study.

Spinal and pelvic injuries were not included in the PECARN co- hort. Our study group elected to include both spinal and pelvic inju- ries in addition to intra-abdominal injuries, to determine if any patients with these injuries required transfusion and met very low- risk criteria. Specifically, the authors were interested in patients with pelvic fractures and associated shear injuries causing sufficient hemorrhage to require blood transfusion. For these patients, ortho- pedic surgical intervention for these fractures was not considered an acute intervention.

Patients with penetrating trauma, injuries occurring over 24 h prior to presentation, known pregnancy, and pre-existing neurologic disor- ders precluding reliable examination were excluded from the PECARN cohort. We had two notable differences in exclusion criteria from the original study. First, to increase our sensitivity in identifying patients with CIIAI, patients with injuries occurring N24 h from time of presenta- tion, or with an unknown injury time or mechanism, were included. Second, children transferred from a referring facility with a previous ab- dominal CT or Diagnostic peritoneal lavage were included. It is the stan- dard practice of our institution to have all referring facility trauma CTs reinterpreted by our institution’s radiologists. The CT reinterpretation was referenced for the presence and grading of all intra-abdominal inju- ries. If a reinterpretation report was not available, the referring facility Radiology report was used.

Three independent reviewers abstracted information from pro- vider documentation, radiology reports, trauma nursing flowsheets, lab results, scanned referring facility records, and discharge summa- ries. Prior to the start of data collection, two sets of 10 charts were co- reviewed to ensure agreement in the method of data abstraction and to modify the data collection instrument as needed. A data source hi- erarchy was created to improve inter-rater agreement when the chart contained conflicting information. Patient demographics and details of injury mechanism and mode of arrival were obtained from a trauma flowsheet, a standard form filled out for all patient en- counters with a trauma designation. Emergency physician documen- tation was used for subjective and physical examination findings, supplemented by the trauma surgeon’s documentation when data elements were not present. Laboratory data were obtained from the electronic medical record, and surgical interventions from the operative note. A random number generator was used to select 10% of charts for co-review in order to assess inter-rater reliability. Co- reviewers determined the presence or absence of all low-risk criteria and a qualifying acute intervention. While reviewers were not blinded to the objectives of the study, they were blinded to the data abstracted from co-reviewed charts.

Data analysis

Descriptive statistics were calculated to report patient demographics, injury patterns, and interventions required for intra-abdominal injuries. The PECARN blunt abdominal trauma prediction rule, as well as other validated and widely implemented clinical prediction rules, have high sensitivity to identify patients at very low risk of injuries requiring inter- vention [22]. We followed this rationale and calculated sensitivity as the proportion of patients with CIIAI who failed one or more very low risk criteria. Sensitivity of the prediction rule and Cohen’s kappa statistic to determine inter-rater agreement were generated using Stata Version

14.1 (StataCorp, College Station, TX).

Results

Of 5743 total patients in the registry, 671 (11%) had ICD codes indi- cating possible intra-abdominal injuries (Fig. 1). 148 (2.6%) of these pa- tients did not require acute intervention and were therefore excluded. The remaining 523 charts (9.1%) were reviewed and abstracted. 81 pa- tients (1.4%) had penetrating trauma and were not included in the final data set. The Cohen’s kappa values for inter-rater agreement comparing very-low risk criteria fulfillment and acute intervention requirement were 0.81, 95% CI [0.76, 0.86] and 0.92, 95% CI [0.86, 0.97], respectively. The composite kappa for both observations was 0.85, 95% CI [0.82, 0.89], indicating strong inter-rater agreement [23].

133 patients (2.3%) were identified with CIIAI requiring acute inter- vention (Table 1). 39 patients (29%) were transferred to our PED from referring hospitals. 80 patients (60%) were male, with an average age of 8 (SD 4.4) years. Children 3 years of age or less comprised 18.8% of the total study sample (n = 25). The most commonly injured organs were bowel or mesentery, liver (median grade 3 [IQR 3,4]), and spleen (median grade 3 [IQR 2,4]). Motor vehicle collisions were the most fre- quent mechanism of injury. For six patients (4.5%), non-accidental trauma was ultimately determined to be the mechanism of injury. The most common intervention was blood transfusion, followed by two or more nights of IV fluids, and therapeutic intervention at laparotomy. No patients required angiographic embolization, and one patient died. Four patients with CIIAI met low-risk criteria during initial chart ab- straction and triggered further review (Table 2). Three of these patients were excluded, as noted in Table 2, resulting in only one case (patient 1) in which a patient with CIIAI would have been identified as very- low risk by the PECARN prediction rule. This resulted in a clinical predic- tion rule sensitivity of 99%, 95% CI [95.9, 100]. A sensitivity analysis was performed in which patients transferred from referring hospitals were not included in our estimate, resulting in a sensitivity of 99%, 95% CI

[94.2, 100].

An additional three patients met very low risk criteria with pelvic or spinal injuries requiring acute intervention, excluding operative fixation of their fractures. These patients had no additional intra-abdominal inju- ries. A 13-year-old female in a rollover all-terrain vehicle (ATV) incident with neurologic deficits on exam was found to have an L2 fracture- dislocation that required operative fixation and intra-operative blood transfusion. However, her associated spinal cord injury prevented reli- able Abdominal examination, excluding her from the low-risk group. Following an ATV collision, a 4-year-old male required a blood transfu- sion after sustaining a fracture of the anterior superior iliac spine, in ad- dition to a large groin laceration which required operative laceration repair and repair of the femoral artery by vascular surgery. A 4-week- old female with somnolence and possible Seizure activity was found to have Skull fractures, subdural and Subarachnoid hemorrhages, and pa- renchymal contusions on Head CT. A skeletal survey additionally noted Pelvic fractures and Abdominal ultrasound showed no intra-abdominal injury or hemoperitoneum. She required blood transfusion for her inju- ries. These two latter patients required blood transfusion but had signif- icant alternate sources of blood loss, and neither had hemoperitoneum suggestive of pelvic shear injury.

Discussion

In this study, we sought to validate the sensitivity of the PECARN blunt abdominal trauma clinical prediction rule in a cohort of patients with CIIAI by identifying those who screened as very low risk for the outcome. Our results indicate that the PECARN blunt abdominal trauma clinical prediction rule has high sensitivity to identify patients at very low risk of CIIAI who are thus unlikely to benefit from CT imaging and should not be exposed to the radiation dose of this procedure.

Several other sets of low-risk criteria have been proposed to de- crease unnecessary CT use in pediatric BAT. Hynick, et al. retrospectively reviewed data from a single trauma center and found that hematuria,

Fig. 1. Patient inclusion flow chart.

Table 1

Demographics, injury type and mechanism, and interventions for 133 patients with CIIAI evaluated for proposed low-risk criteria.

Age (years, SD) 8.0, 4.4

Male n (%) 80 (60)

Total number of injuries 314

Injuries per patient (Median [IQR]) 2 [1,3] Injury types (n [%])

Bowel or mesentery 74 (24%)

Liver 39 (12%)

Spleen 37 (12%)

spinal fracture 32 (10%)

Pelvis fracture 30 (10%)

Pancreas 20 (6%)

Abdominal hematoma 19 (6%)

Kidney 16 (5%)

Adrenal gland 15 (5%)

Other 11 (4%)

Vascular 10 (3%)

Bladder or other urinary tract 6 (2%)

Traumatic fascial defect 5 (2%)

Gallbladder 0 (0%)

trauma mechanism (n [%])

Motor vehicle collision 69 (52%)

ATV or other non-street motorized vehicle 20 (15%)

Pedestrian vs. automobile 14 (11%)

Other 9 (7%)

Fall 7 (5%)

Bicycle, skateboard, or rollerblade 6 (5%)

Crush 5 (4%)

Horse-related 2 (2%)

Assault 1 (1%)

Total number of interventions 211

Interventions per patient (Median [IQR]) 2 [1,3] Interventions (n [%])

Blood transfusion 83 (39%)

>=2 nights of IV fluids 65 (31%)

Therapeutic intervention at laparotomy 62 (29%)

Death 1 (0.5%)

Angiographic embolization 0 (0%)

elevated serum alanine aminotransferase (ALT), and clinically concerning abdominal exam findings were significant predictors of a clinically important intra-abdominal injury (IAI) [16]. Acker, et al. also proposed low-risk criteria comprised of GCS 15, normal pediatric age- adjusted Shock Index (heart rate/systolic blood pressure), and injury at- tributable to a single, non-motorized, blunt force to the abdomen [13]. The criteria included readily available information from history and vital signs, and thus did not depend on rapid laboratory availability or ultrasonography expertise. The authors reported that of 101 low-risk children, only one required blood transfusion (1%) and no child died or required laparotomy or discharge to a rehabilitation facility. However, the study included a relatively small sample size and the

experience of only two trauma centers, limiting its external validity. Ad- ditionally, the study excluded injuries from motorized vehicles, which comprise the majority of pediatric BAT injuries.

Holmes, et al. prospectively identified 6 findings associated with IAI: low systolic blood pressure, abdominal tenderness, femur fracture, as- partate aminotransferase (AST) N 200 U/L or ALT N 125 U/L, urinalysis with N5 red blood cells per high-powered field, and initial hematocrit b30% [17]. The criteria are 98% sensitive and 49% specific. However, pa- tients were enrolled at a single center, limiting generalizability. Karam et al. proposed the BATiC score, which is calculated using abnormal ab- dominal Doppler ultrasound, abdominal pain, peritoneal irritation, he- modynamic instability, AST N 60 IU/L, ALT N 25 IU/L, white blood cell count N 9.5 g/L, LDH N 330, lipase N 30 IU/L, and creatinine N 50 ug/L [20]. A BATiC score <= 7 is 91% sensitive and 84% specific to identify chil- dren with IAI. However, the study was small, with only 31 patients with IAI identified during the study period.

Streck et al. prospectively derived a prediction rule including readily available history and Physical exam findings, in addition to lab and chest x-ray data. Patients were enrolled at fourteen trauma centers, improv- ing generalizability from single center studies. The resulting prediction rule is 98.4% and 100% sensitive for identifying IAI and IAI requiring in- tervention, respectively [14]. However, the rule is less specific (38% and 34.7%) than the PECARN study, and depends on the availability of rapid laboratory results, which may not be available at all facilities.

There are some notable differences between this study and the initial PECARN study. Patients transferred from referring institutions with prior abdominal CT or diagnostic peritoneal lavage were ex- cluded from the initial prospective study. To increase the sample size, the authors elected to include these patients in this study, de- spite the introduction of provider bias. In the PECARN study, an emergency department faculty or fellow physician performed an as- sessment of clinical variables, including physical exam findings. Given the retrospective study design, the authors chose to include both medical and surgical physician documentation to increase the likelihood of documentation of low-risk variables.

While four patients meeting very low-risk criteria with possible CIIAI were initially identified, on further chart review three patients met one or more exclusion criteria from the original PECARN study. This resulted in a prediction rule sensitivity of 99%.

The remaining patient had significant poly-trauma, including femo- ral arterial bleeding requiring emergent operative repair, in addition to liver and adrenal lacerations. As this patient had a clear source of extra- abdominal large-volume blood loss, it is difficult to definitively attribute his anemia and transfusion requirement to his intra-abdominal injuries. Exclusion of this patient would have resulted in 100% sensitivity; how- ever, this patient was included to provide a most-conservative estimate of sensitivity.

Table 2

Patients with CIIAI who met very low-risk criteria on initial chart abstraction.

Patient Age Intra-abdominal

injury (IAI)

Mechanism Qualifying intervention for IAI

Additional injuries Criteria for exclusion from low-risk cohort

Notes

1 15

years

Adrenal laceration, grade 3 Liver laceration

Rollover ATV incident

Blood transfusion

Femur fracture, superficial femoral artery damage

None arterial injury required

immediate operative revascularization

2 26

months

Grade 5 liver laceration

Presumed

non-accidental trauma

Blood transfusion

Unknown injury time, possible abdominal tenderness, faint abdominal bruising noted by consulting child abuse physician

No specific traumatic injury ever reported by caregivers

3 13

months

4 13

years

Rectal laceration Fall onto

bathtub stopper

Rollover ATV incident

Operative laceration repair, IV fluids Blood transfusion

Spinal and scapular fractures, thoracic spine epidural hematoma, large-volume hemopneumothorax (HPTX)

Penetrating trauma, injury to rectum not classified as intra-abdominal by PECARN study

Patient insensate below umbilicus in trauma bay secondary to thoracic spine hematoma, precluding reliable exam

HPTX required chest tube placement and blood transfusion in trauma bay

ICD-10 codes used “>While the prediction rule proposed by the PECARN study was devel- oped to identify patients with solely intra-abdominal injuries, we elected to also include pelvic and Spinal fractures to determine if any pa- tients with associated shear injuries required transfusion and met very low-risk criteria. No such patient was identified. Two patients with pel- vic fractures without hemoperitoneum suggestive of pelvic shear injury required blood transfusion, but both had significant alternate sources of blood loss. Thus, providers should be cautious in applying these criteria to rule out pelvic and/or spinal injuries, as the prediction rule was not derived to identify these injuries and they may be present despite fulfill- ment of the proposed very low-risk criteria.

Limitations

This study is a retrospective chart review at a single institution and thus has several inherent limitations. Blunt abdominal injuries in pedi- atric patients infrequently require procedural intervention, making ad- equate study samples difficult to achieve in single trauma centers [19,24]. In this study, while similar to the findings of the PECARN study, only 19% of patients with ICD codes suggestive of intra- abdominal injury were found to require acute intervention, producing a sample of just 130 patients in over 5 years.

Ascertainment of low risk criteria was based on initial treating phy- sician documentation, which occasionally was incomplete. Physical exam findings often conflicted both between medical and surgical teams and between trainee and attending physicians. Additionally, pa- tients are identified for inclusion in the trauma registry by ICD codes in- dicating known injuries. Because patients evaluated for BAT but without identified injuries would not be included in the registry, the authors were unable to calculate specificity and further test characteristics. Fur- thermore, the registry is generated by manual data entry and thus data entry errors may have occurred.

There were a number of differences in patient inclusion from the ini- tial PECARN study. Patients transferred with previous abdominal CT were included in this study, which introduces bias toward documenta- tion of such elements as abdominal pain and tenderness on examina- tion. However, time of documentation was not compared to time of availability of Imaging results for non-Transferred patients, and thus providers who complete documentation following the patient encoun- ter are theoretically subject to the same bias. Finally, poly-trauma re- quiring blood transfusion accounted for a large number of the total patients, and the etiology of blood loss is difficult to definitively attri- bute in these patients.

Lastly, we used a very specific definition of CIIAI. This effort at consis- tency with the original PECARN study does not address all potential ab- dominal injuries sustained via blunt abdominal trauma, including those that, while not defined as clinically-important, might still be impactful to both the treating team, patient and family.

Conclusions

This study supports the use of the PECARN clinical prediction rule to decrease abdominal CT use in pediatric patients at very low-risk for CIIAI. Only one of 130 total patients with intra-abdominal injury requir- ing acute intervention met very low-risk criteria, and this patient had extra-abdominal arterial bleeding, offering an alternate etiology of blood loss requiring transfusion. No patients with pelvic or spinal frac- tures with associated shear injuries met the proposed criteria. A pro- spective, multicenter study is necessary to evaluate additional test characteristics of this clinical prediction rule and ensure a large enough patient volume for the outcome of interest.

Grants

Meetings

Poster presentation at Pediatric Academic Societies annual meeting, Toronto, Canada, May 2018.

Conflicts of interest

The authors have no conflicts of interest to disclose.

Author contributions

ES, AAV, and DHA conceived and designed the study; ES, SBF, and AAV performed data collection; ES and AV analyzed and interpreted the data; ES and AAV drafted the manuscript; all au- thors critically reviewed the manuscript; and DHA provided statis- tical expertise.

Appendix A. ICD-9 and ICD-10 codes used to generate institutional trauma registry

ICD-9

800-959 Injury and Poisoning, EXCLUDING

905-909: Late effects of injuries, poisoning, toxic effects, and other external causes

930-939: Effects of foreign body entering through body orifice

940-949: Burns without trauma

Patients who are evaluated, resuscitated, treated, and released

Patients who are treated and released from the ambulatory surgical center

Patients who are obstetrics patients with no defined trauma injury

Patients who are admitted whose injury is second- ary to a primary medical diagnosis which caused the event

Patients whose injuries were known to be sustained

N48 h prior to admission

ICD-10

S00-S99 Injuries to specific body parts - initial encounter (with 7th character modifiers of A, B, or C only)

EXCLUDING:

S00: superficial injuries of the head S10: superficial injuries of the neck S20: superficial injuries of the thorax

S30: superficial injuries of the abdomen, pelvis, lower back and external genitals

S40: superficial injuries of the shoulder and upper arm

S50: superficial injuries of the elbow and forearm S60: superficial injuries of the wrist, hand, and fingers

S70: superficial injuries of the hip and thigh

S80: superficial injuries of the knee and lower leg S90: superficial injuries of the ankle, foot, and toes

T07 Unspecified multiple injuries

T14 Injury of unspecified body region

This project was completed without any internal or external finan- cial support.

T79.

A1-79. A9

Traumatic compartment syndrome - initial encounter (with 7th character modifier of A only)

Appendix B. ICD-9 and ICD-10 Diagnosis codes used as initial screen for patient inclusion in study

ICD-9

805.2-805.9 Fracture of vertebral column without mention of spinal cord injury (excluding cervical vertebra)

806.2-806.9 Fracture of vertebral column with spinal cord injury (excluding cervical vertebra)

808 Fracture of pelvis

809 Ill-defined fractures of bones of trunk

863 Injury to gastrointestinal tract

864 Injury to liver

865 Injury to spleen

866 Injury to kidney

867 Injury to pelvic organs

868 Injury to other intra-abdominal organs

869 Injury to other intra-abdominal organs

902 Injury to blood vessels of abdomen and pelvis 926 Crushing injury of trunk

929 Crushing injury of multiple and unspecified sites

959.12 Other injury of abdomen ICD-10

S31 Open wound of abdomen, lower back, pelvis and external genitals S32 Fracture of lumbar spine and pelvis

S35 Injury of blood vessels at abdomen, lower back and pelvis level S36 Injury of intra-abdominal organs

S37 Injury of urinary and pelvic organs

S38 Crushing injury and traumatic amputation of abdomen, lower back, pelvis and external genitals

S39 Other and unspecified injuries of abdomen, lower back, pelvis and external genitals

T07 Unspecified multiple injuries

T14 Injury of unspecified body region

Appendix C. ICD-9 and ICD-10 procedure codes used as secondary screen for patient inclusion in study

ICD-9

00 Procedures And Interventions, Not Elsewhere Classified

17 Other MiscellaneoUS procedures

34.8 Operations On Diaphragm

38 Incision, Excision, And Occlusion Of Vessels

39 Other Operations On Vessels

41 Operations On Bone Marrow And Spleen 42-54 Operations on the Digestive System

55-59 Operations On The Urinary System

83 Operations On Muscle, Tendon, Fascia, And Bursa, Except Hand

84 Other Procedures On musculoskeletal system

99.0 Transfusion Of Blood And Blood components ICD-10

04 Medical and surgical procedures of the lower arteries 06 Medical and surgical procedures of the lower veins 07-P Medical and surgical procedures of the spleen

0D Medical and surgical procedures of the gastrointestinal system

0F Medical and surgical procedures of the hepatobiliary system and pancreas 0T Medical and surgical procedures of the urinary system

0W Medical and surgical procedures of anatomical regions, general

References

1. Heron M. Deaths: leading causes for 2014. Natl Vital Stat Rep 2016;65(5):1-96.
2. National Center for Injury Prevention and Control. Web-based Injury Statistics Query and Reporting System (WISQARS). https://www.cdc.gov/injury/wisqars/; 2017, Accessed date: 27 August 2017.
3. Rui P, Kang K, Albert M. National hospital ambulatory medical care survey: 2013 emergency department summary tables; 2013.
4. Broder J, Warshauer DM. Increasing utilization of computed tomography in the adult

   emergency department, 2000-2005. Emerg Radiol 2006;13(1):25-30.

   Broder J, Fordham LA, Warshauer DM. Increasing utilization of computed tomogra- phy in the pediatric emergency department, 2000-2006. Emerg Radiol 2007;14 (4):227-32.

5. Brenner D, Elliston C, Hall E, Berdon W. Estimated risks of radiation-induced fatal cancer from pediatric CT. AJR Am J Roentgenol 2001;176(2):289-96.
6. Brenner DJ, Hall EJ. Computed tomography-an increasing source of radiation expo- sure. N Engl J Med 2007;357(22):2277-84.
7. Scaife ER, Rollins MD. Managing radiation risk in the evaluation of the pediatric trauma patient. Semin Pediatr Surg 2010;19(4):252-6.
8. Singh S, Kalra MK, Moore MA, et al. Dose reduction and compliance with pediatric CT protocols adapted to patient size, clinical indication, and number of prior studies. Radiology 2009;252(1):200-8.
9. Murtagh Kurowski E, Byczkowski T, Grupp-Phelan JM. Comparison of emergency care delivered to children and young adults with complex chronic conditions be- tween pediatric and general emergency departments. Acad Emerg Med 2014;21 (7):778-84.
10. Schappert SM, Bhuiya F. Availability of pediatric services and equipment in emer- gency departments: United States, 2006. Natl Health Stat Rep 2012;47:1-21.
11. Jindal A, Velmahos GC, Rofougaran R. Computed tomography for evaluation of mild to moderate pediatric trauma: are we overusing it? World J Surg 2002;26(1):13-6.
12. Acker SN, Stewart CL, Roosevelt GE, Partrick DA, Moore EE, Bensard DD. When is it safe to forgo abdominal CT in blunt-injured children? Surgery 2015;158(2):408-12.
13. Streck CJ, Vogel AM, Zhang J, et al. Identifying children at very low risk for blunt intra-abdominal injury in whom CT of the abdomen can be avoided safely. J Am Coll Surg 2017;224(4):449-458.e443.
14. Streck CJ, Jewett BM, Wahlquist AH, Gutierrez PS, Russell WS. Evaluation for intra- abdominal injury in children after blunt torso trauma: can we reduce unnecessary Abdominal computed tomography by utilizing a clinical prediction model? J Trauma Acute Care Surg 2012;73(2):371-6 [discussion 376].
15. Hynick NH, Brennan M, Schmit P, Noseworthy S, Yanchar NL. Identification of blunt abdominal injuries in children. J Trauma Acute Care Surg 2014;76(1):95-100.
16. Holmes JF, Sokolove PE, Brant WE, et al. Identification of children with intra- abdominal injuries after blunt trauma. Ann Emerg Med 2002;39(5):500-9.
17. Holmes JF, Mao A, Awasthi S, McGahan JP, Wisner DH, Kuppermann N. Validation of a prediction rule for the identification of children with intra-abdominal injuries after blunt torso trauma. Ann Emerg Med 2009;54(4):528-33.
18. Leeper CM, Nasr I, Koff A, McKenna C, Gaines BA. Implementation of clinical effec- tiveness guidelines for solid organ injury after trauma: 10-year experience at a level 1 pediatric trauma center. J Pediatr Surg 2017;53(4):775-9.
19. Karam O, Sanchez O, Chardot C, La Scala G. Blunt abdominal trauma in children: a score to predict the absence of organ injury. J Pediatr 2009;154(6):912-7.
20. Holmes JF, Lillis K, Monroe D, et al. Identifying children at very low risk of clinically important blunt abdominal injuries. Ann Emerg Med 2013;62(2):107-116.e102.
21. Kuppermann N, Holmes JF, Dayan PS, et al. Identification of children at very low risk of clinically-important brain injuries after head trauma: a prospective cohort study. Lancet 2009;374(9696):1160-70.
22. McHugh ML. Interrater reliability: the kappa statistic. Biochem Med 2012;22(3): 276-82.
23. Arbra CA, Vogel AM, Zhang J, et al. Acute procedural interventions following pediat- ric blunt abdominal trauma: a prospective multicenter evaluation. J Trauma Acute Care Surg 2017;83(4):597-602.