American Journal of Emergency Medicine 37 (2019) 1643–1648

Contents lists available at ScienceDirect

American Journal of Emergency Medicine

journal homepage: www. elsevier. com/ locate/ajem

Original Contribution

External validation of a clinical prediction rule for very low risk pediatric
blunt abdominal trauma☆
Elise Springer, MD a,1, S. Barron Frazier, MD a, Donald H. Arnold, MD, MPH a,b, Adam A. Vukovic, MD, MEd a,⁎
a
Department of Pediatrics, Division of Pediatric Emergency Medicine, Vanderbilt University Medical Center, 2200 Children's Way Suite 1025, Nashville, TN 37232, USA
b
Department of Pediatrics, Division of Pulmonary Medicine and the Center for Asthma Research, Vanderbilt University School of Medicine, 2200 Children's Way, Nashville, TN 37232, USA

a r t i c l e in f o
abstract
Article history:
Background: Computed tomography (CT) is frequently used to identify intra-abdominal injuries in children
Received 21 September 2018
Received in revised form 17 November with blunt abdominal trauma (BAT). The Pediatric Emergency Care Applied Research Network (PECARN)
2018 proposed a prediction rule to identify children with BAT who are at very low risk for clinically-important
Accepted 21 November 2018 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
Keywords: in our pediatric trauma registry.
Blunt abdominal trauma Methods: Retrospective review of our institutional trauma registry to identify patients with CIIAI. CIIAI
Computed tomography included cases resulting in death, therapeutic intervention at laparotomy, angiographic embolization of
Clinical prediction rule
intra-abdominal arterial bleeding, blood transfusion for intra-abdominal hemorrhage, and administration of
Pediatric
intravenous fluids for two or more nights for pancreatic or gastrointestinal injuries. Patients were identified
Emergency
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.
© 2018 Elsevier Inc. All rights reserved.

1. Introduction
Compared to adults, children have a higher predicted rate of
devel- oping radiation-induced fatal cancer, with an estimated
Unintentional injury is the leading cause of death in pediatric
lifetime attrib- utable risk of death from cancer up to 0.14% per
pa- tients, causing approximately 6000 deaths annually in children
abdominal CT scan [6,7]. Although pediatric institutions have
between 1 and 19 years old [1]. Injuries in children younger than 15
lower rates of CT utilization in children, most pediatric patients are
years of age account for over 7 million emergency department
initially evaluated in general emergency departments [8-11].
visits per year, with the majority of these visits for unintentional
Additionally, young children with mild to moderate injuries
falls, bicycle, and motor ve- hicle collisions (MVCs), which are
undergo more abdominal CTs than their severity-matched adult
frequently associated with blunt ab- dominal trauma (BAT) [2,3].
counterparts [12]. The benefit of identifying clinically important
Computed tomography (CT) is the criterion standard for
intra-abdominal injury (CIIAI) must be balanced with the risk of
identifying intra-abdominal injury in patients with BAT, and is
radiation-induced malignancy from abdominal CT. Sev- eral
thus frequently used in the evaluation of both adult and pediatric
prediction rules have been proposed to risk-stratify these patients
trauma patients [4,5].
and to identify low-risk children in whom abdominal CT can be
avoided [13-20].
☆ This research did not receive any specific grant from funding agencies in the In 2013, Holmes et al. published a prospective study through the
public, commercial, or not-for-profit sectors.
⁎ Corresponding author at: Division of Pediatric Emergency Medicine, Monroe Carell
Pe- diatric Emergency Care Applied Research Network (PECARN),
Jr. Children's Hospital at Vanderbilt, 2200 Children's Way, Suite 1025, Nashville, TN which of- fered a clinical prediction rule to identify a cohort of
37232, USA. children with BAT who are at very low-risk of CIIAI [21]. The rule
E-mail address: [email protected] (A.A. Vukovic). uses seven readily avail- able clinical variables based on history and
1
Pediatric Emergency Medicine Associates, Children's Hospital at Erlanger, 910 physical exam: evidence of abdominal wall trauma or seatbelt sign,
Blackford Street, Chattanooga, TN, 37403 (present address).
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/© 2018 Elsevier Inc. All rights reserved.
1644 E. Springer et al. / American Journal of Emergency Medicine 37 (2019) 1643–1648

vomiting. 12,044 patients were enrolled from 20 participating

from the intra-abdominal injury, blood transfusion for anemia as a
pediatric emergency departments (PEDs). The prediction rule is
re- sult of hemorrhage from the intra-abdominal injury, or
97% sensitive (95% CI [93.7, 98.9]) in identifying patients at very
administration of intravenous (IV) fluids for two or more nights in
low-risk of having CIIAI [21]. In that study, only 0.1% of patients
patients with pan- creatic or gastrointestinal injuries. Therapeutic
who fulfilled the very low-risk criteria were found to have an intra-
laparotomy was defined as any surgical intervention to treat an
abdominal injury requir- ing acute intervention, resulting in a less
intra-abdominal injury. Thus, for this research we defined CIIAI as
favorable risk-benefit ratio in relation to radiation-induced
(1) Intra-abdominal injury that re- quired (2) Acute intervention,
malignancy. Our objective was to deter- mine the sensitivity of the
both defined in accordance with the PECARN study.
PECARN prediction rule by determining the proportion of patients
Spinal and pelvic injuries were not included in the PECARN
with CIIAI who failed the proposed very low- risk criteria and to
co- hort. Our study group elected to include both spinal and pelvic
identify patients who would not have been identified using the
inju- ries in addition to intra-abdominal injuries, to determine if
prediction rule.
any patients with these injuries required transfusion and met very
low- risk criteria. Specifically, the authors were interested in
2. Methods
patients with pelvic fractures and associated shear injuries causing
sufficient hemorrhage to require blood transfusion. For these
2.1. Study design
patients, ortho- pedic surgical intervention for these fractures was
not considered an acute intervention.
We performed a retrospective review of pediatric patients
Patients with penetrating trauma, injuries occurring over 24 h
evaluated and treated in a pediatric emergency department in an
prior to presentation, known pregnancy, and pre-existing
academic, ter- tiary care children's hospital and American College
neurologic disor- ders precluding reliable examination were excluded
of Surgeons (ACS) level 1 trauma center from January 2011 to
from the PECARN cohort. We had two notable differences in
August 2016. The pediatric trauma service maintains a registry of all
exclusion criteria from the original study. First, to increase our
patients treated for traumatic injuries, excluding minor extremity
sensitivity in identifying patients with CIIAI, patients with injuries
trauma. Patients are identified by International Classification of
occurring N24 h from time of presenta- tion, or with an unknown
Diseases (ICD) diagnosis and procedure codes, and trauma team
injury time or mechanism, were included. Second, children
members manually abstract data from the charts. This includes, but
transferred from a referring facility with a previous ab- dominal CT
is not limited to, date of injury, trauma level, initial vital signs,
or diagnostic peritoneal lavage were included. It is the stan- dard
transfusions required, length of hospital stay, and dis- position. We
practice of our institution to have all referring facility trauma CTs
used this registry to identify patients with CIIAI requiring acute
reinterpreted by our institution's radiologists. The CT
intervention (defined below) and reviewed these charts for the
reinterpretation was referenced for the presence and grading of all
presence of very low-risk criteria. This study was approved by the
intra-abdominal inju- ries. If a reinterpretation report was not
insti- tutional review board (Protocol #161912) prior to
available, the referring facility radiology report was used.
commencement.
Three independent reviewers abstracted information from pro-
vider documentation, radiology reports, trauma nursing
2.2. Study setting and population
flowsheets, lab results, scanned referring facility records, and
discharge summa- ries. Prior to the start of data collection, two sets
All pediatric trauma patients b16 years of age are evaluated and of 10 charts were co- reviewed to ensure agreement in the method
treated in the children's hospital, as are those from 16 to 18 years of of data abstraction and to modify the data collection instrument as
age who do not meet adult level one or two trauma activation needed. A data source hi- erarchy was created to improve inter-
criteria. The institutional trauma registry includes pediatric trauma rater agreement when the chart contained conflicting
patients treated in and admitted to the pediatric hospital, as information. Patient demographics and details of injury
defined by ICD codes in Appendix A. During the study period, the mechanism and mode of arrival were obtained from a trauma
institution transitioned from ICD-9 to ICD-10 coding. flowsheet, a standard form filled out for all patient en- counters
with a trauma designation. Emergency physician documen- tation
2.3. Study protocol and measurements was used for subjective and physical examination findings,
supplemented by the trauma surgeon's documentation when data
We initially sought to identify any patient within the trauma registry elements were not present. Laboratory data were obtained from
who sustained any intra-abdominal, spinal or pelvic injury after sustain- the electronic medical record, and surgical interventions from the
ing blunt abdominal trauma. To accomplish this, we narrowed our list operative note. A random number generator was used to select
to all patients in the registry with ICD codes in Appendix B. We then 10% of charts for co-review in order to assess inter-rater
sought to specifically identify those patients who sustained CIIAI as de- reliability. Co- reviewers determined the presence or absence of all
fined below. To achieve this, we further narrowed our patient list to low-risk criteria and a qualifying acute intervention. While
only those with applicable ICD procedure codes, administration of reviewers were not blinded to the objectives of the study, they
blood products, hospitalization for two or more nights in the hospital, were blinded to the data abstracted from co-reviewed charts.
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 2.4. Data analysis
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 Descriptive statistics were calculated to report patient
using criteria from the original PECARN study [21]. Intra-abdominal in- demographics, injury patterns, and interventions required for intra-
jury included any radiographically- or surgically-apparent injury to the abdominal injuries. The PECARN blunt abdominal trauma
following structures: spleen, liver, urinary tract (from the kidney to the prediction rule, as well as other validated and widely implemented
urinary bladder), gastrointestinal tract (including the bowel and associ- clinical prediction rules, have high sensitivity to identify patients at
ated mesentery from the stomach to the sigmoid colon), pancreas, gall- very low risk of injuries requiring inter- vention [22]. We followed
bladder, adrenal gland, intra-abdominal vascular structure, or traumatic this rationale and calculated sensitivity as the proportion of patients
fascial defect (traumatic abdominal wall hernia). Acute intervention with CIIAI who failed one or more very low risk criteria. Sensitivity
was defined by an intra-abdominal injury associated with any of the fol- of the prediction rule and Cohen's kappa statistic to determine
lowing: death caused by the intra-abdominal injury, a therapeutic inter- inter-rater agreement were generated using Stata Version
vention at laparotomy, angiographic embolization to treat bleeding 14.1 (StataCorp, College Station, TX).
 E. Springer et al. / American Journal of Emergency Medicine 37 (2019) 1643–1648 1645

3. Results An additional three patients met very low risk criteria with pelvic
or spinal injuries requiring acute intervention, excluding operative
Of 5743 total patients in the registry, 671 (11%) had ICD codes fixation of their fractures. These patients had no additional intra-
indi- cating possible intra-abdominal injuries (Fig. 1). 148 (2.6%) of abdominal inju- ries. A 13-year-old female in a rollover all-terrain
these pa- tients did not require acute intervention and were vehicle (ATV) incident with neurologic deficits on exam was found
therefore excluded. The remaining 523 charts (9.1%) were reviewed to have an L2 fracture- dislocation that required operative fixation
and abstracted. 81 pa- tients (1.4%) had penetrating trauma and and intra-operative blood transfusion. However, her associated spinal
were not included in the final data set. The Cohen's kappa values cord injury prevented reli- able abdominal examination, excluding
for inter-rater agreement comparing very-low risk criteria her from the low-risk group. Following an ATV collision, a 4-year-
fulfillment and acute intervention requirement were 0.81, 95% CI old male required a blood transfu- sion after sustaining a fracture of
[0.76, 0.86] and 0.92, 95% CI [0.86, 0.97], respectively. The composite the anterior superior iliac spine, in ad- dition to a large groin
kappa for both observations was 0.85, 95% CI [0.82, 0.89], indicating laceration which required operative laceration repair and repair of
strong inter-rater agreement [23]. the femoral artery by vascular surgery. A 4-week- old female with
133 patients (2.3%) were identified with CIIAI requiring acute somnolence and possible seizure activity was found to have skull
inter- vention (Table 1). 39 patients (29%) were transferred to our fractures, subdural and subarachnoid hemorrhages, and pa-
PED from referring hospitals. 80 patients (60%) were male, with an renchymal contusions on head CT. A skeletal survey additionally
average age of 8 (SD 4.4) years. Children 3 years of age or less noted pelvic fractures and abdominal ultrasound showed no intra-
comprised 18.8% of the total study sample (n = 25). The most abdominal injury or hemoperitoneum. She required blood
commonly injured organs were bowel or mesentery, liver (median transfusion for her inju- ries. These two latter patients required blood
grade 3 [IQR 3,4]), and spleen (median grade 3 [IQR 2,4]). Motor transfusion but had signif- icant alternate sources of blood loss, and
vehicle collisions were the most fre- quent mechanism of injury. neither had hemoperitoneum suggestive of pelvic shear injury.
For six patients (4.5%), non-accidental trauma was ultimately
determined to be the mechanism of injury. The most common 4. Discussion
intervention was blood transfusion, followed by two or more
nights of IV fluids, and therapeutic intervention at laparotomy. No In this study, we sought to validate the sensitivity of the PECARN
patients required angiographic embolization, and one patient died. blunt abdominal trauma clinical prediction rule in a cohort of patients
Four patients with CIIAI met low-risk criteria during initial chart with CIIAI by identifying those who screened as very low risk for the
ab- straction and triggered further review (Table 2). Three of these outcome. Our results indicate that the PECARN blunt abdominal
patients were excluded, as noted in Table 2, resulting in only one trauma clinical prediction rule has high sensitivity to identify patients
case (patient at very low risk of CIIAI who are thus unlikely to benefit from CT
1) in which a patient with CIIAI would have been identified as imaging and should not be exposed to the radiation dose of this
very- low risk by the PECARN prediction rule. This resulted in a procedure.
clinical predic- tion rule sensitivity of 99%, 95% CI [95.9, 100]. A Several other sets of low-risk criteria have been proposed to de-
sensitivity analysis was performed in which patients transferred crease unnecessary CT use in pediatric BAT. Hynick, et al.
from referring hospitals were not included in our estimate, retrospectively reviewed data from a single trauma center and
resulting in a sensitivity of 99%, 95% CI [94.2, 100]. found that hematuria,

Fig. 1. Patient inclusion flow chart.

 1646 E. Springer et al. / American Journal of Emergency Medicine 37 (2019) 1643–1648

Table 1
Demographics, injury type and mechanism, and interventions for 133 patients with CIIAI experience of only two trauma centers, limiting its external validity. Ad-
evaluated for proposed low-risk criteria. ditionally, the study excluded injuries from motorized vehicles, which
Age (years, SD) 8.0, 4.4
comprise the majority of pediatric BAT injuries.
Male n (%) 80 (60) Holmes, et al. prospectively identified 6 findings associated with IAI:
Total number of injuries 314 low systolic blood pressure, abdominal tenderness, femur fracture, as-
Injuries per patient (Median [IQR]) 2 [1,3] partate aminotransferase (AST) N 200 U/L or ALT N 125 U/L,
Injury types (n [%]) with N5 red blood cells per high-powered field, and initial hematocrit
urinalysis
Bowel or mesentery 74 (24%)
Liver 39 (12%)
Spleen 37 (12%) b30% [17]. The criteria are 98% sensitive and 49% specific. However, pa-
Spinal fracture 32 (10%) tients were enrolled at a single center, limiting generalizability. Karam
Pelvis fracture 30 (10%) et al. proposed the BATiC score, which is calculated using abnormal ab-
Pancreas 20 (6%)
dominal Doppler ultrasound, abdominal pain, peritoneal irritation, he-
Abdominal hematoma 19 (6%)
Kidney 16 (5%) modynamic instability, AST N 60 IU/L, ALT N 25 IU/L, white blood
Adrenal gland 15 (5%) cell count N 9.5 g/L, LDH N 330, lipase N 30 IU/L, and creatinine N 50
Other 11 (4%) μg/L [20]. A BATiC score ≤ 7 is 91% sensitive and 84% specific to
Vascular 10 (3%)
identify chil- dren with IAI. However, the study was small, with only
Bladder or other urinary tract 6 (2%)
Traumatic fascial defect 5 (2%) 31 patients with IAI identified during the study period.
Gallbladder 0 (0%) Streck et al. prospectively derived a prediction rule including readily
Trauma mechanism (n [%]) available history and physical exam findings, in addition to lab and
Motor vehicle collision 69 (52%) chest x-ray data. Patients were enrolled at fourteen trauma centers,
ATV or other non-street motorized vehicle 20 (15%)
improv- ing generalizability from single center studies. The resulting
Pedestrian vs. automobile 14 (11%)
Other 9 (7%) prediction rule is 98.4% and 100% sensitive for identifying IAI and IAI
Fall 7 (5%) requiring in- tervention, respectively [14]. However, the rule is less
Bicycle, skateboard, or rollerblade 6 (5%) specific (38% and 34.7%) than the PECARN study, and depends on the
Crush 5 (4%) availability of rapid laboratory results, which may not be available at all
Horse-related 2 (2%)
facilities.
Assault 1 (1%)
Total number of interventions 211 There are some notable differences between this study and the
Interventions per patient (Median [IQR]) 2 [1,3] initial PECARN study. Patients transferred from referring
Interventions (n [%]) institutions with prior abdominal CT or diagnostic peritoneal
Blood transfusion 83 (39%) lavage were ex- cluded from the initial prospective study. To
≥2 nights of IV fluids 65 (31%)
increase the sample size, the authors elected to include these
Therapeutic intervention at laparotomy 62 (29%)
Death 1 (0.5%) patients in this study, de- spite the introduction of provider bias.
Angiographic embolization 0 (0%) 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
elevated serum alanine aminotransferase (ALT), and clinically design, the authors chose to include both medical and surgical
concerning abdominal exam findings were significant predictors of physician documentation to increase the likelihood of
a clinically important intra-abdominal injury (IAI) [16]. Acker, et al. documentation of low-risk variables.
also proposed low-risk criteria comprised of GCS 15, normal While four patients meeting very low-risk criteria with possible
pediatric age- adjusted shock index (heart rate/systolic blood CIIAI were initially identified, on further chart review three patients
pressure), and injury at- tributable to a single, non-motorized, blunt met one or more exclusion criteria from the original PECARN study.
force to the abdomen [13]. The criteria included readily available This resulted in a prediction rule sensitivity of 99%.
information from history and vital signs, and thus did not depend The remaining patient had significant poly-trauma, including
on rapid laboratory availability or ultrasonography expertise. The femo- ral arterial bleeding requiring emergent operative repair, in
authors reported that of 101 low-risk children, only one required addition to liver and adrenal lacerations. As this patient had a clear
blood transfusion (1%) and no child died or required laparotomy source of extra- abdominal large-volume blood loss, it is difficult to
or discharge to a rehabilitation facility. However, the study definitively attribute his anemia and transfusion requirement to his
included a relatively small sample size and the 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 Additional injuries Criteria for exclusion from low- Notes
intervention risk cohort
for IAI
1 15
Adrenal Rollover ATV Blood Femur fracture, superficial None Arterial injury required
years
laceration, grade incident transfusion femoral artery damage immediate operative
3 liver laceration revascularization
2 26 Grade 5 liver Presumed Blood Unknown injury time, possible No specific traumatic
months laceration non-accidental transfusion abdominal tenderness, faint abdominal injury ever reported by
trauma bruising noted by consulting child caregivers
abuse physician
3 13 Rectal laceration Fall onto Operative Penetrating trauma, injury to rectum
months bathtub laceration not classified as intra-abdominal by
stopper repair, IV PECARN study
fluids
4 13 Rollover Blood Spinal and scapular fractures, Patient insensate below umbilicus in HPTX required
years ATV incident transfusion thoracic spine epidural trauma bay secondary to thoracic spine chest tube
hematoma, large-volume hematoma, precluding reliable exam placement and
hemopneumothorax (HPTX) blood transfusion in
trauma bay
 E. Springer et al. / American Journal of Emergency Medicine 37 (2019) 1643–1648 1647

While the prediction rule proposed by the PECARN study was Meetings
devel- oped to identify patients with solely intra-abdominal
injuries, we elected to also include pelvic and spinal fractures to Poster presentation at Pediatric Academic Societies annual
determine if any pa- tients with associated shear injuries required meeting, Toronto, Canada, May 2018.
transfusion and met very low-risk criteria. No such patient was
identified. Two patients with pel- vic fractures without Conflicts of interest
hemoperitoneum suggestive of pelvic shear injury required blood
transfusion, but both had significant alternate sources of blood loss. The authors have no conflicts of interest to disclose.
Thus, providers should be cautious in applying these criteria to rule
out pelvic and/or spinal injuries, as the prediction rule was not Author contributions
derived to identify these injuries and they may be present despite
fulfill- ment of the proposed very low-risk criteria. ES, AAV, and DHA conceived and designed the study; ES, SBF,
and AAV performed data collection; ES and AV analyzed and
5. LiThitations interpreted the data; ES and AAV drafted the manuscript; all au- thors
critically reviewed the manuscript; and DHA provided statis- tical
This study is a retrospective chart review at a single institution expertise.
and thus has several inherent limitations. Blunt abdominal injuries
in pedi- atric patients infrequently require procedural intervention, Appendix A. ICD-9 and ICD-10 codes used to generate institutional
making ad- equate study samples difficult to achieve in single trauTha registry
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 ICD-9
years. 800–959 Injury and Poisoning, EXCLUDING
Ascertainment of low risk criteria was based on initial treating
phy- sician documentation, which occasionally was incomplete.
905–909: Late effects of injuries, poisoning, toxic
Physical exam findings often conflicted both between medical and
effects, and other external causes
surgical teams and between trainee and attending physicians.
Additionally, pa- tients are identified for inclusion in the trauma 930–939: Effects of foreign body entering through
registry by ICD codes in- dicating known injuries. Because patients body orifice
evaluated for BAT but without identified injuries would not be 940–949: Burns without trauma
included in the registry, the authors were unable to calculate Patients who are evaluated, resuscitated, treated,
specificity and further test characteristics. Fur- thermore, the registry and released
is generated by manual data entry and thus data entry errors may Patients who are treated and released from the
have occurred. ambulatory surgical center
There were a number of differences in patient inclusion from the Patients who are obstetrics patients with no defined
ini- tial PECARN study. Patients transferred with previous
trauma injury
abdominal CT were included in this study, which introduces bias
Patients who are admitted whose injury is second-
toward documenta- tion of such elements as abdominal pain and
tenderness on examina- tion. However, time of documentation was ary to a primary medical diagnosis which caused the
not compared to time of availability of imaging results for non- event
transferred patients, and thus providers who complete Patients whose injuries were known to be sustained
documentation following the patient encoun- ter are theoretically N48 h prior to admission
subject to the same bias. Finally, poly-trauma re- quiring blood ICD-10
S00–S99 Injuries to specific body parts – initial encounter (with 7th
transfusion accounted for a large number of the total patients, and character modifiers of A, B, or C only)
the etiology of blood loss is difficult to definitively attri- bute in EXCLUDING:
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 S00: superficial injuries of the head
potential ab- dominal injuries sustained via blunt abdominal S10: superficial injuries of the neck
trauma, including those that, while not defined as clinically- S20: superficial injuries of the thorax
important, might still be impactful to both the treating team, S30: superficial injuries of the abdomen, pelvis,
patient and family. lower back and external genitals
S40: superficial injuries of the shoulder and upper
6. Conclusions arm
S50: superficial injuries of the elbow and forearm
This study supports the use of the PECARN clinical prediction
S60: superficial injuries of the wrist, hand, and
rule to decrease abdominal CT use in pediatric patients at very
fingers
low-risk for CIIAI. Only one of 130 total patients with intra-
S70: superficial injuries of the hip and thigh
abdominal injury requir- ing acute intervention met very low-risk
criteria, and this patient had extra-abdominal arterial bleeding, S80: superficial injuries of the knee and lower leg
offering an alternate etiology of blood loss requiring transfusion. S90: superficial injuries of the ankle, foot, and toes
T07 Unspecified multiple injuries
No patients with pelvic or spinal frac- tures with associated shear
T14 Injury of unspecified body region
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

This project was completed without any internal or external finan- cial support.
T79. T raumatic compartment syndrome – initial encounter (with 7th
A1–79. A9 character modifier of A only)
1648 E. Springer et al. / American Journal of Emergency Medicine 37 (2019) 1643–1648

Appendix B. ICD-9 and ICD-10 diagnosis codes used as initial 0F Medical and surgical procedures of the hepatobiliary system and
screen for patient inclusion in study pancreas 0T Medical and surgical procedures of the urinary system
0W Medical and surgical procedures of anatomical regions, general

References
ICD-9
805.2–805.9 Fracture of vertebral column without mention of spinal cord injury [1] Heron M. Deaths: leading causes for 2014. Natl Vital Stat Rep 2016;65(5):1–96.
(excluding cervical vertebra) [2] National Center for Injury Prevention and Control. Web-based Injury Statistics
806.2–806.9 Fracture of vertebral column with spinal cord injury (excluding Query and Reporting System (WISQARS). https://www.cdc.gov/injury/wisqars/;
cervical vertebra) 2017, Accessed date: 27 August 2017.
808 Fracture of pelvis [3] Rui P, Kang K, Albert M. National hospital ambulatory medical care survey:
2013 emergency department summary tables; 2013.
809 Ill-defined fractures of bones of trunk
[4] Broder J, Warshauer DM. Increasing utilization of computed tomography in the
863 Injury to gastrointestinal tract
adult emergency department, 2000–2005. Emerg Radiol 2006;13(1):25–30.
864 Injury to liver
[5] Broder J, Fordham LA, Warshauer DM. Increasing utilization of computed
865 Injury to spleen tomogra- phy in the pediatric emergency department, 2000–2006. Emerg
866 Injury to kidney Radiol 2007;14 (4):227–32.
867 Injury to pelvic organs [6] Brenner D, Elliston C, Hall E, Berdon W. Estimated risks of radiation-induced
868 Injury to other intra-abdominal organs fatal cancer from pediatric CT. AJR Am J Roentgenol 2001;176(2):289–96.
869 Injury to other intra-abdominal organs [7] Brenner DJ, Hall EJ. Computed tomography—an increasing source of radiation
902 Injury to blood vessels of abdomen and expo- sure. N Engl J Med 2007;357(22):2277–84.
pelvis 926 Crushing injury of trunk [8] Scaife ER, Rollins MD. Managing radiation risk in the evaluation of the pediatric
929 Crushing injury of multiple and unspecified sites trauma patient. Semin Pediatr Surg 2010;19(4):252–6.
[9] Singh S, Kalra MK, Moore MA, et al. Dose reduction and compliance with
959.12 Other injury of abdomen
pediatric CT protocols adapted to patient size, clinical indication, and number of
ICD-10
prior studies. Radiology 2009;252(1):200–8.
S31 Open wound of abdomen, lower back, pelvis and external
[10] Murtagh Kurowski E, Byczkowski T, Grupp-Phelan JM. Comparison of emergency
genitals S32 Fracture of lumbar spine and pelvis care delivered to children and young adults with complex chronic conditions be-
S35 Injury of blood vessels at abdomen, lower back and pelvis tween pediatric and general emergency departments. Acad Emerg Med 2014;21
level S36 Injury of intra-abdominal organs (7):778–84.
S37 Injury of urinary and pelvic organs [11] Schappert SM, Bhuiya F. Availability of pediatric services and equipment in
S38 Crushing injury and traumatic amputation of abdomen, lower emer- gency departments: United States, 2006. Natl Health Stat Rep 2012;47:1–
back, pelvis and external genitals 21.
S39 Other and unspecified injuries of abdomen, lower back, pelvis [12] Jindal A, Velmahos GC, Rofougaran R. Computed tomography for evaluation of
and external genitals mild to moderate pediatric trauma: are we overusing it? World J Surg
2002;26(1):13–6.
T07 Unspecified multiple injuries
[13] Acker SN, Stewart CL, Roosevelt GE, Partrick DA, Moore EE, Bensard DD. When
T14 Injury of unspecified body region
is it safe to forgo abdominal CT in blunt-injured children? Surgery
2015;158(2):408–12.
[14] Streck CJ, Vogel AM, Zhang J, et al. Identifying children at very low risk for
Appendix C. ICD-9 and ICD-10 procedure codes used as secondary blunt intra-abdominal injury in whom CT of the abdomen can be avoided safely.
screen for patient inclusion in study J Am Coll Surg 2017;224(4):449–458.e443.
[15] 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].
[16] Hynick NH, Brennan M, Schmit P, Noseworthy S, Yanchar NL. Identification of
ICD-9
blunt abdominal injuries in children. J Trauma Acute Care Surg 2014;76(1):95–
00 Procedures And Interventions, Not Elsewhere Classified 100.
17 Other Miscellaneous Procedures [17] Holmes JF, Sokolove PE, Brant WE, et al. Identification of children with intra-
34.8 Operations On Diaphragm abdominal injuries after blunt trauma. Ann Emerg Med 2002;39(5):500–9.
38 Incision, Excision, And Occlusion Of Vessels [18] Holmes JF, Mao A, Awasthi S, McGahan JP, Wisner DH, Kuppermann N. Validation
39 Other Operations On Vessels of a prediction rule for the identification of children with intra-abdominal
41 Operations On Bone Marrow And injuries after blunt torso trauma. Ann Emerg Med 2009;54(4):528–33.
Spleen 42–54 Operations on the Digestive [19] Leeper CM, Nasr I, Koff A, McKenna C, Gaines BA. Implementation of clinical
System 55–59 Operations On The Urinary effec- tiveness guidelines for solid organ injury after trauma: 10-year experience
System at a level 1 pediatric trauma center. J Pediatr Surg 2017;53(4):775–9.
[20] Karam O, Sanchez O, Chardot C, La Scala G. Blunt abdominal trauma in
83 Operations On Muscle, Tendon, Fascia, And Bursa, Except Hand
children: a score to predict the absence of organ injury. J Pediatr 2009;154(6):912–
84 Other Procedures On Musculoskeletal System
7.
99.0 Transfusion Of Blood And Blood Components [21] Holmes JF, Lillis K, Monroe D, et al. Identifying children at very low risk of
ICD-10 clinically important blunt abdominal injuries. Ann Emerg Med 2013;62(2):107–
04 Medical and surgical procedures of the lower 116.e102.
arteries 06 Medical and surgical procedures of the [22] Kuppermann N, Holmes JF, Dayan PS, et al. Identification of children at very low
lower veins 07-P Medical and surgical procedures of the risk of clinically-important brain injuries after head trauma: a prospective cohort
spleen study. Lancet 2009;374(9696):1160–70.
0D Medical and surgical procedures of the gastrointestinal system [23] McHugh ML. Interrater reliability: the kappa statistic. Biochem Med 2012;22(3):
276–82.
[24] 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.