Injury, Int. J. Care Injured 45 (2014) 835–839

Contents lists available at ScienceDirect

Injury

journa l homepage: www.e lsevier .com/ locate / in jury

Consequences of increased use of computed tomography imaging for

trauma patients in rural referring hospitals prior to transfer to aregional trauma centre

Timothy J. Berkseth a, Michelle A. Mathiason b, Mary Ellen Jafari c, Thomas H. Cogbill d,*,Nirav Y. Patel d,1

a Department of Medical Education, Gundersen Medical Foundation, La Crosse, WI, USAb Department of Medical Research, Gundersen Medical Foundation, La Crosse, WI, USAc Department of Diagnostic Physics, Gundersen Health System, La Crosse, WI, USAd Department of General and Vascular Surgery, Gundersen Health System, La Crosse, WI, USA

A R T I C L E I N F O

Article history:

Accepted 4 January 2014

Key words:

Computed tomography

Diagnostic imaging

Critical care

Referral hospital

Trauma centres

Radiation exposure

A B S T R A C T

Background: Computed tomography (CT) plays an integral role in the evaluation and management of

trauma patients. As the number of referring hospital (RH)-based CT scanners increased, so has their

utilization in trauma patients before transfer. We hypothesized that this has resulted in increased time at

RH, image duplication, and radiation dose.

Methods: A retrospective chart review was completed for trauma activations transferred to an ACS-

verified Level II Trauma Centre (TC) during two time periods: 2002–2004 (Group 1) and 2006–2008

(Group 2). 2005 data were excluded as this marked the transition period for acquisition of hospital-based

CT scanners in RH. Statistical analysis included t test and x2 analysis. P < 0.05 was considered significant.

Results: 1017 patients met study criteria: 503 in group 1 and 514 in group 2. Mean age was greater in

group 2 compared to group 1 (40.3 versus 37.4, respectively; P = 0.028). There were 115 patients in group

1 versus 202 patients in group 2 who underwent CT imaging at RH (P < 0.001). Conversely, 326 patients

in group 1 had CT scans performed at the TC versus 258 patients in group 2 (P < 0.001). Mean time at the

RH was similar between the groups (117.1 and 112.3 min for group 1 and 2, respectively; P = 0.561).

However, when comparing patients with and without a pretransfer CT at the RH, the median time at RH

was 140 versus 67 min, respectively (P < 0.001). The number of patients with duplicate CT imaging

(n = 34 in group 1 and n = 42 in group 2) was not significantly different between the two time periods

(P = 0.392). Head CTs comprised the majority of duplicate CT imaging in both time periods (82.4% in

group 1 and 90.5% in group 2). Mean total estimated radiation dose per patient was not significantly

different between the two groups (group 1 = 8.4 mSv versus group 2 = 7.8 mSv; P = 0.192).

Conclusions: A significant increase in CT imaging at the RH prior to transfer to the TC was observed over

the study periods. No associated increases in mean time at the RH, image duplication at TC, total

estimated radiation dose per patient, and mortality rate were observed.

� 2014 Elsevier Ltd. All rights reserved.

Introduction

Computed tomography (CT) plays an integral role in theevaluation and management of trauma patients. The initial careof the trauma patient, according to Advanced Trauma LifeSupport (ATLS) guidelines, emphasizes early stabilization and

* Corresponding author. Tel.: +1 608 775 2310; fax: +1 608 775 4460.

E-mail address: thcogbil@gundersenhealth.org (T.H. Cogbill).1 Dr. Patel’s current affiliation is Banner Good Samaritan Medical Center;

Phoenix, AZ.

0020–1383/$ – see front matter � 2014 Elsevier Ltd. All rights reserved.

http://dx.doi.org/10.1016/j.injury.2014.01.002

expedited transfer to a definitive trauma centre based onseverity of injury. The American College of Surgeons Committeeon Trauma has guidelines on criteria for the immediate transferof moderately to severely injured patients to Level I/II TraumaCentres [1]. Criteria for immediate transfer of trauma patients toa higher level of care include altered mental status, respiratoryfailure requiring mechanical ventilation, hemodynamic insta-bility, and penetrating trauma. These guidelines rely oninformation primarily obtained from history, physical examina-tion, chest X-ray and pelvis X-ray. Furthermore, these guidelinesadvocate against the acquisition of pretransfer CT in this patientpopulation.

Table 1Effective radiation dose calculations by exam type.

Study group Effective dose (mSv)

Group 1

Head CT 2.0a

Chest CT 9.1a

Abdomen/pelvis CT 12.1a

Group 2

Head CT 2.0b

Chest CT 7.0b

Abdomen/pelvis CT 10.0b

CT = computed tomography.a CRCPD 2000 NEXT CT Survey, Tables 1.3, 1.24, and 1.302.b NCRP Report No 160, Table 4.23.

Table 2Patient characteristics.

Variable Group 1 Group 2 P value

(2002–2004) (2006–2008)

N 503 514 –

Age, mean (SD); years 37.4 (20.5) 40.3 (21.4) 0.028

Sex, n (%) 0.081

Female 156 (31) 133 (26)

Male 347 (69) 381 (74)

Mechanism of injury, n (%) 0.175

Blunt 483 (96) 483 (94)

Penetrating 20 (4) 31 (6)

ISS, median (IQR) 14 (19–25) 11 (6–19) <0.001

AIS �3, n (%)

Face 11 (2) 4 (1) 0.109

Head 177 (35) 166 (32) 0.363

Chest 157 (31) 120 (23) 0.006

Abdomen 57 (11) 47 (9) 0.295

Extremity 151 (30) 98 (19) <0.001

Any region 363 (72) 344 (67) 0.081

30-Day mortality, n (%) 28 (5.6) 36 (7.0) 0.415

SD = standard deviation; ISS = injury severity score; IQR = interquartile range

(25th–75th percentile); AIS = abbreviated injury score.

T.J. Berkseth et al. / Injury, Int. J. Care Injured 45 (2014) 835–839836

Prior to 2005, most of the referring hospitals (RH) in our areahad only limited access to mobile-based CT scanners. However,after 2005, the number of hospital-based CT scanners at thesefacilities increased dramatically. We hypothesized that thewidespread acquisition of hospital-based CT scanners at RHresulted in increased utilization of CT prior to transfer, increasedtime at the RH, more duplicate CT imaging, and increased overallradiation dose per patient.

Materials and methods

Our health system is an integrated healthcare organizationserving 19 counties in western Wisconsin, southeast Minnesota,and northeast Iowa. The institution includes an American Collegeof Surgeons (ACS) verified Level II Trauma Centre (TC). Aretrospective review of our TC’s entries in the National TraumaRegistry of the American College of Surgeons (NTRACS) wasperformed to identify all trauma patients admitted to the TC fromJanuary 1, 2002 to December 31, 2008. All patients meeting traumaactivation criteria who were transferred from the 35 RH in ourreferral area to the TC during two defined study periods wereincluded. Group 1 consisted of those patients transferred fromJanuary 1, 2002 to December 31, 2004. Group 2 consisted of thosepatients transferred from January 1, 2006 to December 31, 2008.The 2005 data were excluded as this marked the transition periodfor the widespread acquisition of hospital-based CT scanners at RHin our region. By the beginning of 2006, all of the RH in our regionwere using hospital-based CT scanners. During the early years ofthe study period, CT images were frequently sent on a compact discwith the patient. After 2006, electronic transfer of images wasincreasingly used. Transfer agreements between our institutionand the RH remained the same throughout the entire study period.

Variables reviewed included demographics (age and gender),mechanism of injury, injury severity score (ISS), and abbreviatedinjury scale (AIS). We reviewed the number of patients whounderwent CT at a RH, at the TC, and overall. CT scans weresubcategorized into the total number of CTs by anatomic location(head, chest, abdomen/pelvis) at a RH, at the TC, and overall.Finally, we reviewed the time at the RH, based on arrival anddischarge times at the RH, estimated radiation dose per patient,length of hospital stay at the TC (LOS), and 30-day mortality.

Actual radiation dose per patient was difficult to accuratelydetermine. Individual patient volume CT dose index (CTDIvol) anddose length product (DLP) were not routinely recorded for all CTexams at the TC until 2007 and not at the RH until 2008–2010.Effective radiation dose for trauma patients in group 1 was derivedfrom a Food & Drug Administration report of a Conference ofRadiation Control Program Directors (CRCPD) Nationwide Evalua-tion of X-Ray Trends (NEXT) survey publication [2]. Effectiveradiation dose for trauma patients in group 2 was derived fromNational Council on Radiation Protection and Measurements(NCRP) Report number 160 [3]. These derivations are summarizedin Table 1. The reason for using two different sources for estimationof effective dose was because significant advances in CT technologyaffecting patient dose occurred during the study periods. Higherpower generators and more robust X-ray tubes capable of scanningat greater kVp and mAs values were implemented. Use of volumescanning with spiral CT became standard, as did wider detectorsfor multi-row, multi-slice scanning, enabling faster scanning andacquisition of thinner slices. Cone beam CT, dual source CT, anddose reduction strategies such as tube current modulation weredeveloped [4]. Effective dose values from the CRCPD NEXTpublication and NCRP Report No. 160 were from the years 2000and 2006, respectively, and were thus appropriate for technologyin use during each study period.

Statistical analysis consisted of t test, Wilcoxon Rank Sum, andchi square analysis. Continuity adjustment was used in 2 � 2tables. Statistical significance was defined as a P value <0.05.

Results

One thousand seventeen patients met study inclusion criteria;503 patients in group 1 and 514 patients in group 2. Demographiccomparison revealed a significant difference in mean age betweenthe two groups with group 1 being younger (Table 2). There was nosignificant difference in gender between the two groups. Mechanismof injury was primarily blunt (Table 2). The overwhelming majorityof blunt trauma involved motor vehicle collisions and falls.

Median injury severity score (ISS) was 14 in group 1 and 11 ingroup 2 (P < 0.001). Overall, the number of patients withabbreviated injury scale (AIS) �3 was not significantly differentbetween the groups. However, when stratified by anatomiclocation, chest AIS �3 (n = 157 for group 1 and n = 120 for group2; P = 0.006) and extremity AIS �3 (n = 151 for group 1 and n = 98for group 2; P < 0.001) were more frequent in group 1. Nosignificant differences were found for face, head, or abdomen AIS�3 between the two groups.

As depicted in Table 3, there were 115 patients in group 1 versus202 patients in group 2 who underwent CT imaging at RH(P < 0.001). Conversely, 326 patients in group 1 had CT scansperformed at the TC versus 258 patients in group 2 (P < 0.001). Theoverall number of patients who had CT scans performed was not

Table 3Number of patients who underwent CT scans.

Location of CT Group 1 Group 2 P value

(2002–2004) N = 503 (2006–2008) N = 514

N (%)

At RH 115 (22.9) 202 (39.3) <0.001At TC 326 (64.8) 258 (50.2) <0.001At TC or RH 394 (78.3) 399 (77.6) 0.787

CT = computed tomography; TC = level 2 trauma centre; RH = referring hospital.

T.J. Berkseth et al. / Injury, Int. J. Care Injured 45 (2014) 835–839 837

different in the two time periods (n = 394 in group 1 and n = 399 ingroup 2).

Table 4 provides a breakdown of the number of CTs performed,according to anatomic location, in group 1 and group 2 at both theRH and TC. There were 63 patients in group 1 and 127 patients ingroup 2 who underwent head CT imaging at RH. There was aconcomitant decrease in the number of patients undergoing headCTs at the TC during the two time periods, with 229 patients ingroup 1 and 166 patients in group 2 (P < 0.001). There was asimilar trend in chest CT imaging at RH and TC during the two timeperiods; 6 patients had chest CTs at RH in group 1 and 35 patientshad chest CTs at RH in group 2 versus 81 patients with chest CTs atthe TC in group 1 and 65 patients with chest CTs at TC in group 2(P < 0.001). Finally, the number of patients who had abdomen/pelvis CTs obtained at RH increased from 24 in group 1 to 40 ingroup 2. At the TC the number of abdomen/pelvis CTs obtaineddecreased from 155 patients in group 1 to 119 patients in group 2(P = 0.013). The total number of CT scans performed at RH was 127in group 1 versus 244 in group 2. The total number of CT scansperformed at the TC was 499 in group 1 versus 392 in group 2. Thetotal number of CT scans performed in each time period wassimilar, with 626 in group 1 and 636 in group 2.

The number of patients with duplicate CT imaging (n = 34 ingroup 1 and n = 42 in group 2) was not significantly differentbetween the two time periods (P = 0.392). Head CT comprised themajority of duplicate CT imaging in both time periods (82.4% ingroup 1 and 90.5% in group 2). The number of duplicate CT chestand CT abdomen/pelvis studies stayed relatively constant betweenthe two time periods.

Table 4Types of CTs performed by location.

Type of CT Group 1 Group 2 P value

(2002–2004) (2006–2008)

N (%)

Head CT <0.001None 183 (36.4) 183 (35.6)

RH only 63 (12.5) 127 (24.7)

TC only 229 (45.5) 166 (32.3)

Duplicate (RH and TC) 28 (5.6) 38 (7.4)

Chest CT <0.001None 415 (82.5) 413 (80.5)

RH only 6 (1.2) 35 (6.8)

TC only 81 (16.1) 65 (12.8)

Duplicate (RH and TC) 1 (0.2) 1 (0.2)

Abdomen/pelvis CT 0.013None 319 (63.4) 352 (68.5)

RH only 24 (4.8) 40 (7.8)

TC only 155 (30.8) 119 (23.2)

Duplicate (RH and TC) 5 (1.0) 3 (0.6)

Total number of studies

RH 127 244

TC 499 392

Overall 626 636

CT = computed tomography; TC = level 2 trauma centre; RH = referring hospital.

Mean time at the RH was 117.1 min in group 1 patients and112.3 min in group 2 patients (P = 0.561). When comparingpatients with and without a pretransfer CT, regardless of group,the median time at RH was 140 versus 67 min (P < 0.001). Themedian ISS for those with versus without a pretransfer CT was 14versus 13 (P = 0.828). Mean estimated radiation dose per patientscanned was not significantly different between the two groups(group 1 = 8.4 mSv versus group 2 = 7.8 mSv; P = 0.192). The 30-day mortality rate in group 1 was 5.6% (n = 28) and group 2 was 7%(n = 36) (P = 0.415). Median TC LOS was 3 days for patients withand without a pretransfer CT. The 75th percentile was 5 days forthose with a pretransfer CT and 8 days for those without apretransfer CT (P = 0.008). Mortality among patients with andwithout a pretransfer CT were 4.4% and 7.2%, respectively(P = 0.088).

Discussion

Data from a study published by Brenner and Hall [5], confirmedan explosion in the use of CT from 3 million in 1980 to 62 millioncurrently. There has been increased reliance on imaging in trauma,with the increased emphasis on non-operative management forboth blunt and penetrating trauma. In a study published by Inabaet al. [6], trauma patients admitted during a two month period in2002 were compared to the same two month period in 2007.Despite similar patients and outcomes, the mean number of CTsper blunt trauma patient increased significantly (2.1 � 1.6 versus3.2 � 2.0; P < 0.001). There was also a significant consequent increasein radiation dose per patient from 2002 to 2007 (11.5 � 11.3 mSvversus 20.7 � 14.9 mSv; P < 0.05). Similar results were reported byAhmadinia et al. [7], in an analysis of 500 trauma patients each in theyears 2002, 2005, and 2008. The mean number of CT scans forcategory 1 (highest acuity) trauma patients in 2002, 2005, and 2008was 1.5, 3.1, and 4.5, respectively (P = 0.01). A similar trend wasobserved in category 2 trauma patients: 2.0, 3.5, and 5.1, respectively(P < 0.01). This resulted in increased total radiation dose to category 1and 2 trauma patients over 2002, 2005, and 2008: 12.0 mSv,23.6 mSv, and 33.6 mSv (P = 0.02); and 17.5 mSv, 24.1 mSv, and37.5 mSv (P < 0.001), respectively. No differences in injury severity ormortality were demonstrated.

There is increasing concern about the long-term health effectsof radiation dose from CT scans, especially in younger patients.Compared to 1991, when an estimated 0.4% of all cancers in theUnited States could potentially be attributed to radiation from CTscans, today the estimate is as high as 1.5% to 2.0% [5]. The reasonsfor this increase are multifactorial but are partly due to thesignificant increase in the use of CT scans during this time period.In a study to determine overall radiation dose from imaging, Tienet al. [8] placed dosimeters on 172 injured patients. The meaneffective dose was 22.7 mSv, which is potentially associated withan estimated 190 additional cancer deaths for every 100,000patients sustaining this exposure. The thyroid dose was 58.5 mSv,which could be responsible for a potential estimated 11.7additional fatal thyroid cancers for every 100,000 patientsexposed. Mahoney et al. [9] achieved a 33% reduction in theestimated total radiation dose per trauma patient by implementingan evidence-based guideline approach as opposed to a liberal CTscan protocol. The average number of CT scans per patientdecreased from 1.9 to 1.2. The estimated total radiation dose perpatient in the evidence-based group (n = 611) was 36.7 mGyversus 53.3 mGy in the conventional group (n = 612). Estimatedradiation dose values per patient were lower in our seriescompared to other reports, despite the use of similar referencesfor calculating doses. The reason for this difference appears to bethat the number of scans per patient in our study was lower than inthe other studies and consequently dose per patient was lower.

T.J. Berkseth et al. / Injury, Int. J. Care Injured 45 (2014) 835–839838

In an effort to identify factors affecting the decision to obtainimaging studies prior to transfer, Lee et al. surveyed 218physicians referring trauma patients to a Level 1 TC [10]. One-third of those responding identified perceived expectations of thereceiving TC as justification for obtaining imaging, independentof patient acuity and 28% of respondents obtained imagingbecause of liability concerns, even if that meant delays intransfer. Another study evaluating the issue of CT scanning priorto transfer, found that 53.2% of CT scans performed at localhospitals were repeated at the TC [11]. Increased age, higher ISS,and longer delays before transfer were factors associated withrepeat CT imaging.

There are several reasons why CT scans may need to berepeated. Gupta el al. [12] found that 75% of trauma patientsunderwent CT imaging prior to transfer to their regional TC; 58% ofpatients underwent repeat CT imaging at the TC. Head CT was themost commonly repeated scan and the most common reason wasclinical indication. Cervical, thoracic, and lumbar spine CT wererepeated predominately due to inadequate technique secondary tolack of reconstructions. The most frequent reason to repeat chestand abdomen CT was inappropriate use of contrast. In addition, 7%of CT scans were repeated as they were not sent with the patientand 13% had software incompatible with the imaging displaysystem at the TC. For patients with moderate or severe headtrauma, repeat head CT scans are routinely obtained at many TCs at24 h or for any clinical signs of neurologic deterioration.

Duplicate CT imaging increases patient cost and radiation doseand consumes valuable hospital resources. There also may bereluctance from TC radiologists to provide formal interpretationsof outside CT images due to medical-legal concerns. This places theburden on the trauma surgeon to base management decisionseither on their own interpretation or to repeat imaging. Developingsystem-wide integrated radiologic computer networks wouldminimize the need to repeat CT imaging due to images not sentwith the patient or software incompatibility. Mahoney et al. [9]have demonstrated a drop in duplicate CT imaging from 75%overall to 41% for facilities affiliated with their institution. Thishighlights the advantage of a regional trauma network. Adoptingstandardized protocols for trauma imaging at RH is another wayfor a trauma system to improve in this area of inefficiency. Sincethere can be wide variation among hospitals in CT manufacturer,model, and available techniques, protocol standardization can becomplex and assistance from a medical physicist is recommended.Over the past 5 years, our TC has collaborated with all of the RH inour region to assure that software and data network compat-ibilities allow physicians at the TC to electronically access all CTimages obtained at the RH. This has allowed the TC traumasurgeons to review CT images, usually in advance of patient arrivalat the TC. Also, the need to repeat CT scans because the RH imagesare not available is now rare.

Previous studies have shown that prolonged times at RH anddelays to transfer to definitive care are associated with worsepatient outcomes [13,14]. Although our study did not demonstratean increased mean time at RH despite increased use of CT duringthe second time period, it is plausible that the time at RH wouldhave actually decreased if the additional CT scans had not beenperformed at RH. Our significant finding for the entire study periodthat the median time at the RH for patients with a pretransfer CTwas 140 min versus 67 min for patients without a pretransfer CTprovides support for this conclusion. Expedited transfer to the TC isconsistent with ATLS guidelines. When comparing patients whounderwent a pretransfer CT to those who did not have a pretransferCT, regardless of group, a longer LOS at the TC was observed amongpatients without a pretransfer CT. This is probably due to the factthat patients who were transferred immediately to the TC withoutpretransfer CTs were more likely to be less stable and have more

severe injuries. However, the median ISS was not different whencomparing patients with versus those without a pretransfer CT.

Increased volume of CT scans in trauma patients is associatedwith increased hospital charges. Ahmadinia et al. [7] retrospec-tively reviewed trauma patients in two different acuity categoriesin 2002, 2005, and 2008. The mean charge (technical andprofessional fees) for diagnostic imaging in the first 24 h perpatient was $2933, $4656, and $6677 for 2002, 2005, and 2008respectively (P < 0.01). Based on this institution’s annual traumavolume, they estimated that >$13 million in additional chargeswere accrued for CT scans during the final year of study. Therecertainly are financial incentives for small, RH to perform CTimaging in trauma patients. However, any resultant decrease in CTscans performed at the accepting TC may be associated with asignificant TC revenue loss.

What is the appropriate role for CT imaging at RH? If a transferto a higher level of care will be necessary based on the injuriesidentified in the primary survey, an argument can be made tostabilize the patient and expedite transfer without obtaining anyCT scans. However, if the results of a CT are likely to affect thedecision whether or not to transfer to a higher level of care,imaging at the RH is certainly justified. For example, a hemody-namically stable patient sustaining blunt trauma with isolated leftlower rib fractures alone, who requires IV analgesia and aggressivepulmonary care, may be a reasonable candidate for admission to aRH. However, that same patient, if found to have an associatedsplenic laceration underlying the rib fractures on CT imaging, maynot be a good candidate for admission to a RH. Any CT imaging at aRH that will not dictate a change in patient care before transfershould be avoided in an effort to expedite transfer to definitivecare. Since this study was completed, our recommended policy oflimiting CT scans performed at the RH to only those which areanticipated to yield results that may affect the decision to transferthe patient to the TC has become an area of educational emphasis.This point is clearly articulated at the three ATLS Courses and threeto four Rural Trauma Team Development Courses sponsored by ourTC each year. Additionally, this issue is a focus of our regionaltrauma performance improvement process.

There are several limitations of our study. The use of estimatedeffective radiation doses from national studies is less accurate thanpatient specific dose measurements. Until 2007 when CT radiationdose first received widespread national attention [5], few facilitiesrecorded patient CT dose indicators. Our health system implemen-ted routine recording of CTDIvol and DLP values for each patientexam in 2007, followed by the RH in 2008–2010; thus, completedose indicator data were not available for the study periods. A moreaccurate method of patient dose computation applies patient sizespecific conversion factors to the displayed CTDIvol value [15].Another limitation of our study is the inherent shortcoming ofretrospective study design. It is difficult in retrospect to determinethe appropriateness of CT imaging performed at RH as well as thereasons for repeat CT scanning. Finally, we did not have access tocharge or cost data for comparison.

Conclusions

The transition from mobile-based CT scanners to the acquisitionof hospital-based CT scanners at RH coincided with a significantincrease in the utilization of CT imaging in trauma patients prior totransfer to the regional TC. Interestingly, despite this increase inutilization of CT at RH, this was not associated with an increase inmean time at the RH. However, the median time for all patientsundergoing a pretransfer CT at the RH was significantly greaterthan those who did not undergo a pretransfer CT. Because fewerpatients underwent CT imaging at the TC from group 1 to group 2,both the overall number of patients who had CT imaging

T.J. Berkseth et al. / Injury, Int. J. Care Injured 45 (2014) 835–839 839

performed and the total number of CT scans performed during thetwo time periods remained essentially the same. The overall resultwas a significant shift in the location where the CT scans wereperformed, with more patients being scanned at the RH and aconcomitant decrease in the number of CT scans at the TC duringthe recent time period. This likely resulted in increased RH imagingcharges and less potential revenue for the TC. We did notdemonstrate an increase in duplicate CT imaging despite theincreased number of CT scans performed at the RH. There was nodifference in mortality rate when comparing patients with orwithout a pretransfer CT. Based on these findings, we continue tosupport a policy of limiting CT scans performed in RH to only thosewhich are anticipated to provide information which aids stabili-zation or transfer.

References

[1] American College of Surgeons. Resources for the optimal care of the injuredpatient. Chicago, IL: American College of Surgeons; 2006.

[2] Stern SH. CRCPD. Nationwide evaluation of X-ray trends (NEXT), tabulationand graphical summary of 2000 survey of computed tomography. Frankfort,KY: Conference of Radiation Control Program Directors, Inc; 2007.

[3] National Council on Radiation Protection and Measurements. Ionizing radia-tion exposure of the population of the United States. In: NCRP Report No 160.Bethesda, MD: National Council on Radiation Protection and Measurements;2009.

[4] Kalender W. Computed tomography fundamentals, system technology, imagequality, applications. 3rd ed. Erlangen, Germany: Publicis Publishing; 2011.

[5] Brenner DJ, Hall EJ. Computed tomography—an increasing source of radiationexposure. N Engl J Med 2007;357(22):2277–84.

[6] Inaba K, Branco BC, Lim G, Russell K, Teixeira PG, Lee K, et al. The increasingburden of radiation exposure in the management of trauma patients. J Trauma2011;70(6):1366–70.

[7] Ahmadinia K, Smucker JB, Nash CL, Vallier HA. Radiation exposure has in-creased in trauma patients over time. J Trauma Acute Care Surg 2012;72(2):410–5.

[8] Tien HC, Tremblay LN, Rizoli SB, Gelberg J, Spencer F, Caldwell C, et al.Radiation exposure from diagnostic imaging in severely injured traumapatients. J Trauma 2007;62(1):151–6.

[9] Mahoney E, Agarwal S, Li B, Dechert T, Abbensetts J, Glantz A, et al. Evidence-based guidelines are equivalent to a liberal computed tomography scanprotocol for initial patient evaluation but are associated with decreasedcomputed tomography scan use, cost, and radiation exposure. J Trauma AcuteCare Surg 2012;73(3):573–8.

[10] Lee CY, Bernard AC, Fryman L, Coughenour J, Costich J, Boulanger B, et al.Imaging may delay transfer of rural trauma victims: a survey of referringphysicians. J Trauma 2008;65(6):1359–63.

[11] Haley T, Ghaemmaghami V, Loftus T, Gerkin RD, Sterrett R, Ferrara JJ. Trauma:the impact of repeat imaging. Am J Surg 2009;198(6):858–62.

[12] Gupta R, Greer SE, Martin ED. Inefficiencies in a rural trauma system: theburden of repeat imaging in interfacility transfers. Trauma 2010;69(2):253–5.

[13] Sampalis JS, Denis R, Lavoie A, Frechette P, Boukas S, Nikolis A, et al. Traumacare regionalization: a process-outcome evaluation. J Trauma 1999;46(4):565–79.

[14] Harrington DT, Connolly M, Biffl WL, Majercik SD, Cioffi WG. Transfer times todefinitive care facilities are too long: a consequence of an immature traumasystem. Ann Surg 2005;241(6):961–6.

[15] American Association of Physicists in Medicine. Size-specific dose estimates(SSDE) in pediatric and adult body CT examinations. In: Report of AAPM TaskGroup 204. College Park, MD: American Association of Physicists in Medicine;2011.