Trends in Utilization of Whole-Body Computed Tomography in Blunt Trauma after MVC: Analysis of the Trauma Quality Improvement Program (TQIP) Database

Abstract

Background:

The use of whole-body computed tomography (WBCT) in awake, clinically stable injured patients is controversial. It is associated with unnecessary radiation exposure and increased cost. We evaluate use of CT imaging during the initial evaluation of injured patients at American College of Surgeons (ACS) Level I and II Trauma Centers (TCs) after blunt trauma.

Methods:

We identified adult blunt trauma patients after motor vehicle crash (MVC) from the ACS Trauma Quality Improvement Program (TQIP) database between 2007–2016 at Level I or II TCs. We defined awake clinically stable patients as those with SBP ≥ 100mmHg with a Glasgow Coma Scale score of 15. CT imaging had to have been performed within 2 hours of arrival. WBCT was defined as simultaneous CT of the head, chest and abdomen, and selective CT if only 1–2 aforementioned regions were imaged. Patients were stratified by injury severity score (ISS).

Results:

There were 217,870 records for analysis; 131,434 (60.3%) had selective CT, 86,436 (39.7%) had WBCT. Overall, there was an increasing trend in WBCT utilization over the study period (p<0.001). In patients with ISS < 10, WBCT was utilized more commonly at Level II vs Level I TCs in patients discharged from the emergency department (26.9% vs 18.3%, p <0.001), that had no surgical procedure(s) (81.4% vs 80.3%, p <0.001) and no injury of the head (53.7% vs 52.4%, p=0.008) or abdomen (83.8% vs 82.1%, p=0.001). The risk adjusted odds of WBCT was 2 times higher at Level II TC vs Level I (OR 1.88, 95% CI [1.82–1.94], p <0.001).

Conclusion:

WBCT utilization is increasing relative to selective CT. This increasing utilization is highest at Level II TCs in patients with low injury severity scores, and in patients without associated head or abdominal injury. The findings have implications for quality improvement and cost reduction.

Level of Evidence:

III

Background:

Recent advances in multidetector computed tomography (CT) technology has led to increasing implementation of CT protocols for rapid injury detection in the early management of trauma patients.^1–4 Performing whole-body CT (WBCT) imaging on unevaluable patients is accepted as the standard of care in many trauma centers (TCs). In contrast, for patients with minor injuries and a reliable exam, the role of CT imaging is typically left to the discretion of the treating physician. Some studies have suggested that the use of WBCT, when compared to selective use of CT imaging, facilitates timely and accurate diagnosis of occult injuries leading to a reduction in morbidity and mortality.^{1, 5, 6} Proponents of liberal WBCT use advocate that establishing the absence of injuries or diagnosis of minor injuries can allow for patients to be safely discharged home from the emergency department rather than being admitted to the hospital for ongoing observation. Increasing use of CT imaging, however, is also associated with increased risk of radiation exposure to the patient⁷ and higher cost of care.⁸

The American College of Surgeons (ACS) Verification Review Committee and Advanced Trauma Life Support programs both promulgate standards to optimize care of trauma patients. However, despite these standards, there continues to be variation in practice patterns between providers and institutions with the use of CT imaging. Despite multiple attempts to establish practice guidelines,^9–12 there remains a lack of general consensus regarding the appropriate use of CT imaging in the early evaluation of patients presenting after blunt injury.

To our knowledge, the utilization of CT imaging during the initial evaluation of trauma patients in U.S. TCs has not been studied at a national level. The purpose of this study was to evaluate the use of WBCT imaging relative to selective CT use during the initial evaluation of patients presenting with blunt trauma after MVC at ACS Level I or II trauma center designation.

Methods:

Patient Population

This study was deemed exempt by the Loyola University Chicago Institutional Review Board. The ACS Trauma Quality Improvement Program (TQIP) database is the largest national trauma database, containing high-quality incident-based data voluntarily submitted by participating TCs. We queried ACS-TQIP to identify adult patients aged 18–65 years with blunt trauma secondary to MVC between 2007 and 2016.

We only included patients who were hemodynamically stable and neurologically intact, defined as those with an initial systolic blood pressure (SBP) ≥ 100mmHg and Glasgow Coma Scale (GCS) score of 15. WBCT was defined as having had a CT of the head, chest and abdomen simultaneously performed within the first 2 hours of arrival using International Classification of Disease (ICD) 9 and 10 procedure codes and time to procedure. Selective CT was considered to be a CT of at least one but not all three of these same body regions.

We excluded patients who had an ICD diagnosis code indicating pregnancy at the time of injury. Records with missing data for any of the variables used to define our study population and for any variables indicating the year of admission/injury and Injury Severity Score (ISS) were also excluded. We then grouped patients by the ACS TC designation (Level I or II). Trends in WBCT utilization were first examined within our overall study population and then stratified according to ISS. Given the marked difference in WBCT use in patients with less severe injury we chose to focus the remainder of our analysis to only those patients with an ISS ranging from 0 to 9.

Variable Definitions

Procedures were classified using HCUP Clinical Classifications Software for Services and Procedures.¹³ Surgical procedures were defined according to the Medicare Severity Diagnosis Related Group (MS-DRG) classifications¹⁴ as any diagnostic/therapeutic operating room procedure. Extremity procedures were defined as any procedure performed for fracture of any of the long or short bones of the upper or lower extremities.

We determined the severity of injury according to Abbreviated Injury Scale (AIS) by body region. A severe injury in patients with ISS < 10 was considered to be any injury with AIS of 3. To control for errors inherent in large data collection, ISS was manually calculated from AIS as the sum of squares for the highest values in each of the three most severely injured body regions¹⁵ and compared to the recorded ISS.

Statistical Analysis

To account for variation in the number of patients entering the dataset each year, we calculated yearly based percentages of WBCT relative to selective CT for ACS Level I and II TC designation for an ISS <10, 10 to 15, 16 to 24 and > 24. Unadjusted annual rates were plotted over time for comparison of trend. Multivariable Poisson regression was used to evaluate for significant increase or decrease between scanning modality over time adjusting for age, sex, vital signs on presentation, TC designation and highest AIS by body region.

We calculated general descriptive statistics as frequency and percentages. Data are reported using χ² test for categorical variables and Student’s t-tests for continuous variables for comparison between cohorts as appropriate. We used multivariable regression to assess the overall risk-adjusted odds of a WBCT being performed relative to selective CT. The extent to which the effect of variables used in risk-adjusted models on TC designation varied according to the severity of injury was examined and interaction terms were incorporated where indicated.

One-to-one fixed ratio nearest neighbor propensity score matching was used to create matched cohorts between Level I and Level II trauma designation. Cohorts were matched for age, sex, initial presenting vital signs and severity of injury by body region without replacement and caliper distance of 0.001. Predictors entered into regression models for risk-adjusted analysis were selected a priori from those available in the dataset as those felt by the practicing physicians on the research team to be most likely to be associated with the odds of receiving WBCT vs selective CT. Variance inflation factor analysis was performed to assess for multicollinearity of models. Model fit was evaluated using c-statistic and Hosmer-Lemeshow goodness-of-fit testing. All statistical tests were two sided and significance was defined as P < 0.05. All data acquisition and analysis were performed in R (Version 3.6.1, The R Foundation for Statistical Computing).

Results:

Of the 7,722,056 records in TQIP from 2007 to 2016, there were 2,045,835 which were indicated as having CT imaging of the head, chest and/or abdomen within the first two hours of arrival. Of these, a total of 217,870 records were identified as meeting inclusion criteria. There were 131,434 (60.3%) patients who had a selective CT and 86,436 (39.7%) had a WBCT.

Overall Trends in CT Utilization

Figure 1 shows the unadjusted annual percentage of patients who had a WBCT relative to selective CT at Level I vs Level II TCs. From 2007 to 2016, there was increased utilization of WBCT imaging at both Level I and II TCs while the selective use of CT scans decreased at both facilities. Overall, the changes in use of CT imaging during the study period demonstrate a convergence in practice patterns at Level I and II TCs with the greatest change in WBCT use occurring at Level I TCs. The results of our multivariable Poisson regression demonstrated a significant difference in the risk-adjusted utilization rate of WBCT relative to selective CT over the 10-year study period (p < 0.001).

Figure 1.

Unadjusted annual rates of utilization of WBCT vs selective CT between Level I vs Level II TCs

Trends in CT Utilization by Injury Severity

Figure 2 depicts the unadjusted annual rates of WBCT utilization relative to selective CT for each of the levels of injury severity at Level I vs Level II TCs. In patients with ISS <10, utilization of selective CT imaging was higher than WBCT throughout the study period at both Level I and Level II TCs. However, the difference in rates of WBCT and selective CT utilization between TCs decreased throughout the 10-year study period. In the first year of this study (2007), the rate of selective CT use was nearly 3 times higher than WBCT (73.5% vs 26.5%) at Level I TCs. In contrast, there was only a 24.0% absolute difference in the frequency of the two imaging practices at Level II TCs (62% vs 38%). By 2016, this difference decreased and the practices of using WBCT and selective CT were more similar at both Level I and II TCs. The use of selective CT in patients with an ISS <10 was 21.8% higher than WBCT at Level I TCs (60.9% vs 39.1%) and 18.1% at Level II TCs (59.1% vs 41.0%).

Figure 2.

Annual rates of utilization of WBCT vs selective CT by severity of injury between Level I vs Level II TCs

As expected, the differences between TC designation and utilization of selective and WBCT scanning was less as the burden of injury (ISS) increased. Despite WBCT and selective CT use becoming more similar, selective CT continued to be utilized more frequently in patients with ISS of 10–24 at Level I TCs throughout the 10-year study period. Conversely, Level II TCs started utilizing WBCT more frequently than selective CT in patients with ISS of 10–15 and ISS of 16–24 beginning in 2009 until 2016 (Figure 2).

WBCT Utilization in Patients with ISS <10

The greatest difference in utilization of selective and WBCT by level of TC designation was observed in the least injured patients (ISS <10). Thus, we focused the remainder of our analysis on this subgroup to identify differences in practices between TCs. Of the 77,407 patients with ISS 0–9 treated at Level I TCs from 2007–2016, 25,221 (32.6%) had a WBCT performed. In contrast, significantly (p <0.001) more patients seen at a Level II TC underwent WBCT scanning (43.3%).

Univariate Analysis

Univariate analysis comparing the rates of WBCT by TC designation (Table 1) revealed that WBCT was performed more commonly at centers with Level II vs Level I designation in patients who were discharged home from the ED (26.9% vs 18.3%, p <0.001) and in those who were discharged within 4 hours of the CT scan (23.5% vs 10.7%, p <0.001) or had a hospital stay of less than 24 hours (47.6% vs 44.9%, p <0.001). Of all patients who had surgical procedure(s) performed, 43.1% were for extremity fracture. There was no significant difference in time to procedure between Level I designation (mean 23.35 hours [SD 32.28 hours]) vs Level II designation (mean 24.21 hours, [SD 161.18 hours], p=0.718). More than 50% of the patients who received a WBCT at either TC designation, Level I or II, had no injury of the head, chest or abdomen when assessed by individual body region. There was no significant difference between Level I and Level II designation regarding the percentage of WBCTs performed in patients who did not have any injury (AIS=0) to the head, chest or abdomen combined (35.3% vs 35.2%, p=0.702).

Table 1.

Univariate analysis evaluating associations between WBCT use at Level I TC compared to Level II TC in patients with ISS < 10

	WBCT, Level I	WBCT, Level II	p
n	25221	16500
Age, years, mean± SD	36.77± 13.50	37.13± 13.69	0.009
Male (%)	15631 (62.0)	10521 (63.8)	<0.001
Race (%)			<0.001
White	15459 (62.2)	11657 (72.8)
Black	5135 (20.7)	1542 (9.6)
Other	4243 (17.1)	2805 (17.5)
Hispanic or Latino (%)	7118 (18.3)	6908 (20.9)	<0.001
Insurance (%)			<0.001
Private	7993 (33.5)	5255 (34.3)
Other/Not Billed	6045 (25.3)	5032 (32.9)
Uninsured	6182 (25.9)	2564 (16.8)
Government	4346 (18.2)	2966 (19.4)
ISS, mean± SD	4.55± 2.55	4.50± 2.52	0.022
Underwent surgery during admission (%)	4973 (19.7)	3065 (18.6)	0.004
Time to first surgical procedure, in hours (mean (SD))	23.35(32.28)	24.21(161.18)	0.718
No head injury (%)	13180 (52.4)	8837 (53.7)	0.008
No chest injury (%)	19661 (78.1)	12771 (77.6)	0.213
No abdominal injury (%)	20666 (82.1)	13720 (83.3)	0.001
No spine injury (%)	18765 (74.5)	12058 (73.2)	0.003
No head, chest or abdominal injury (%)	8913 (35.3)	5800 (35.2)	0.702
Discharged within 4 hours of CT (%)	2690 (10.7)	3716 (23.5)	<0.001
Hospital LOS1, in days (mean (SD))	2.73 (3.84)	2.60 (4.40)	0.002
Hospital LOS1 < 24 hours (%)	11281 (44.9)	7535 (47.6)	<0.001
ED discharge disposition (%)			<0.001
Admitted, floor2	14849 (58.9)	8380 (50.8)
Home	4611 (18.3)	4433 (26.9)
Admitted, observation unit	2575 (10.2)	944 (5.7)
Admitted, ICU3	1447 (5.7)	1151 (7.0)
Operating room	1494 (5.9)	985 (6.0)
Left against medical advice	104 (0.4)	504 (3.1)
Other (jail/institutional facility/transferred)	141 (0.6)	103 (0.6)
Teaching status
University	18771 (74.4)	5351 (32.4)
Community	6411 (25.4)	8477 (51.4)
Non-teaching	39 (0.2)	2672 (16.2)

¹Length of stay

²Includes in-patient floor/telemetry/step-down

³Intensive care unit

Trends in CT Utilization and Injury Incidence

To further evaluate the trend in injury incidence in relation to the rate of WBCT utilization over time, we plotted the relative annual percentage of patients with no injury of the head, chest or abdomen combined in relation to the rate of WBCT at Level I and II trauma designation in patients with ISS < 10 (Figure 3). Despite an increase in WBCT utilization at both Level I and II TCs, the relative percentage of patients with no combined head, chest or abdominal injury decreased over time. Comparing Level I vs Level II designation of TCs, the rate of WBCT utilization increased more over time relative to the percentage of patients with no injuries detected at Level II designated TCs when compared to Level I designated TCs.

Figure 3.

Annual rates of WBCT utilization relative to percentage of patients with no injury detected in Level I vs Level II TCs

Multivariable and Propensity Match Analysis

On multivariable analysis of patients with ISS of <10 (Table 2), adjusted for age, sex, insurance, initial heart rate, teaching status and severity of injury, the risk adjusted odds of WBCT at centers with a Level II trauma designation was 2 times higher as compared to Level I trauma designation (OR 1.88, 95% CI [1.82–1.94], p <0.001). The association between the odds of WBCT and TC designation in patients with ISS < 10 was further evaluated using one-to-one propensity match analysis where cohorts were matched between the two levels of trauma designation adjusting for age, sex, initial SBP, pulse, respiratory rate and highest AIS of head, chest, abdomen, spine, upper extremity and lower extremity. Of the 115,535 eligible records, 36,938 were matched to TCs with a Level I designation and 36,938 matched to TCs with Level II designation with no significant differences observed between cohorts for the aforementioned variables (p-values ≥ 0.25).The results of the propensity matched analysis demonstrated that centers with a Level II trauma designation were more likely to utilize WBCT when compared to Level I trauma designation (56.9% vs 43.1%, p <0.001).

Table 2.

Multivariable analysis evaluating risk adjusted odds of WBCT compared to selective CT in patients with ISS < 10

	OR1	95% CI2		p-value
Age, in years	1.01	1.00	0.98	<0.001
Female	0.98	0.95	1.00	0.06
Insurance
Private	-	-	-
Government	1.10	1.06	1.15	<0.001
Other/Not Billed	1.06	1.03	1.10	<0.001
Uninsured	1.18	1.14	1.23	<0.001
Heart rate, in bpm3	1.00	1.00	1.00	<0.001
Severe head injury4	0.49	0.39	0.60	<0.001
Severe chest injury4	1.05	0.93	1.18	0.41
Severe abdominal injury4	1.00	0.71	1.39	1.00
Severe spine injury4	0.67	0.57	0.78	<0.001
Severe lower extremity injury4	0.92	0.75	1.11	0.37
Teaching status
University	-	-	-
Community	0.58	0.56	0.760	<0.001
Non-teaching	0.53	0.50	0.57	<0.001
Trauma Center
Level I	-	-	-
Level II	1.88	1.82	1.94	<0.001

¹Odds Ratio

²Confidence Interval

³beats per minute

⁴in patients with ISS > 10 “severe injury” refers to patients with AIS = 3

Hosmer-Lemeshow statistic of 19.334, p > 0.05

Discussion:

In this first, contemporary, national study on the utilization of WBCT in adults who sustain blunt trauma secondary to MVC and arrive with a SBP ≥ 100mmHg a GCS of 15, we demonstrate that 1) overall WBCT utilization is increasing in the United States TCs; 2) the increase is most apparent in centers with ACS Level II trauma designation, especially in patients with ISS less than 10; and 3) TCs with ACS Level II designation utilized WBCT without a head or abdominal injury more often than Level I designated centers did, but also discharged more patients within 4 hours of admission to the ED. The results of our study would suggest that there is increasing utilization of WBCT in patients without associated clinically significant injury burden, at both ACS Level I and II trauma designations.

Achievements in public health efforts over the last two decades have led to a significant reduction in the total number of vehicular crashes and MVC related deaths in the U.S.¹⁶ The results presented here (Figure 3) suggest that the overall burden of injury for those involved in MVC in the U.S. is decreasing as well. Despite decreases in total crashes, MVC fatalities and severe injury, the estimated economic cost of trauma care from vehicular crashes in the U.S. has increased by over $10 billion dollars.^{17, 18} We demonstrate that use of CT imaging for the early management of patients presenting after MVC at both ACS Level I and II TCs is rising and may contribute to this increase in health care cost.

It is likely that this increase in use of CT imaging is driven in part by technologic advancements and increased availability of multi-detector CT technology over the last two decades. There is evidence to suggest that WBCT use in the evaluation of patients injured by blunt force may significantly improve survival over patients who receive selective CT imaging.^{5, 6, 19} and may also allow for prompt and safe discharge from the ED.^{11, 20, 21} Our study demonstrates there has been increasing utilization of CT imaging for rapid, detailed injury detection in the early management of injured patients.^1–4 Inappropriate use of WBCT may, however, also lead to increased risk of unnecessary radiation exposure,^{7, 22} detection of clinically insignificant findings, longer lengths of stay and increased cost.^{7, 22–24}

In examining the difference in CT utilization between TCs, overall, the rates of WBCT use at both Level I and II TCs have converged over time. The greatest increase in WBCT use occurred at Level I TCs. Despite this change at Level I TCs, the utilization of WBCT remained higher at Level II TCs. While we are unable to comment on reasons for these differences in practice between the two trauma designations, changes in criteria for verification regarding the evaluation of less injured patients during the study period may be a contributing factor. Additionally, the TQIP dataset does not include information on the level of activation or whether the initial evaluation was performed by a trauma surgeon or an emergency medicine physician.

Advocates of more liberal use of CT during the initial evaluation of blunt trauma patients have argued that the imaging allows for documentation that there has been no injury and thus more prompt identification of patients who can be discharged from the ED. Historically, a 23-hour observational period was recommended following moderate-risk blunt trauma.²⁵ Livingston et al²¹ and Salim et al¹¹ demonstrated that a normal head²¹ and abdominal^{11, 20} CT scan in the absence of clinical findings allowed for safe discharge of patients from the ED. Whether or not WBCT as opposed to selective CT²⁶ should be utilized to assure early, safe discharge, and in what clinical scenarios, remains uncertain.

In this study, we evaluated the association between patients who received a WBCT within 2 hours of arrival and the proportion of patients who were discharged within 4 hours of CT. Our results demonstrate that significantly more patients with ISS < 10 who received WBCT at TCs with Level II designation were discharged within 4 hours of CT as compared to those managed in a similar fashion at centers with Level I designation (23.5% vs 10.7%, p <0.001). While there are numerous other medical and social factors that contribute to timely discharge and without knowing the intent or indication for the scan, our findings suggest that Level II trauma designation may be associated with performing WBCT scan more often to facilitate prompt discharge. We are unable to discern cost/charge information from the available data in this study, however, it is possible that while WBCT may incur additional upfront cost, prompt discharge as opposed to admission may provide an overall cost-savings. Additional studies are needed in order to determine whether it is appropriate to use WBCT for this purpose versus relying on clinical examination and more selective CT scanning and how such practices may be associated with overall hospital cost and patient safety.

We found that TCs with both Level I and II designation, in patients with minimal injury burden (ISS < 10) who arrived hemodynamically stable, neurologically intact and had WBCT performed, the majority did not have injury of the body regions scanned. Specifically, over 50% did not have any head injuries, over 70% did not have any chest injuries, over 80% did not have any abdominal injuries when evaluating each independent body region imaged by WBCT. In an attempt to evaluate the incidence of a “negative” WBCT scan, defined here as the combination of no head, chest or abdominal injury, we found that 35% of patients with ISS < 10 who received WBCT at centers with both Level I and II trauma designation did not have any injury to the torso or head. Furthermore, only 6% of patients with ISS < 10 who received a WBCT went to the operating room from the ED with over half of cases being performed for extremity fracture. One explanation for these findings may be that the use of WBCT in cases of isolated extremity fracture may reassure the clinician that there are no occult injuries and allow for early discharge and planned readmission at a later time for operative fixation of the fracture.

While WBCT may afford clinicians with more detailed and rapid information, it is not without associated risk to the patient. The leading cause of death in patients aged 1– 44 years in the United States is traumatic injury.¹⁶ Equally important, it is this same age demographic which is most susceptible to the associated risk of malignancy seen with increasing levels of ionizing radiation exposure.²² In this study we found that the average age of patients with minimal to no injury burden in our cohort who received CT imaging was approximately 37 years old. Studies have reported that a single WBCT scan is associated with enough radiation to have the potential to induce malignancy in 1/1000 patients under 40 years of age.²² While this risk is additive over time and thus greatest in children, patients under the age of 30 are at approximately a 0.05% lifetime attributable risk of developing a malignancy from a single abdominal CT scan alone.⁷ Trauma surgeons should be cognizant of this risk when deciding to include CT scan imaging in the evaluation of injured patients.

Between 2008 and 2012 we found that the use of WBCT at Level I TCs increased from 20.4% to 40.5% of patients relative to selective CT. Census reports demonstrate that by 2007 over half of the CT scanners in the U.S. had 64-slice multidetector capability.²⁷ Survey data from the same year reports that 91% of emergency physicians significantly underestimated radiation dose from CT scans.²⁷ It was not until 2010 that the Food and Drug Administration launched an Initiative to Reduce Unnecessary Radiation Exposure from Medical Imaging, bringing increasing attention to the negative consequences from ionizing radiation exposure associated with CT imaging. It is therefore possible that these findings are in part due to a lag between increasing availability of advanced CT technology and awareness and understanding regarding the health risk and patient safety associated with use of ionizing radiation.

This study has several limitations. First, we recognize that the definition of WBCT typically includes a CT of the cervical spine in addition to the head, chest and abdomen/pelvis. However, there are no ICD-9 codes for a CT of the cervical spine and in order to include data prior to the introduction of ICD-10 codes in 2015, we had to omit CT of the cervical spine as a part of our definition of WBCT. We included injury of the spine in our analysis and as a part of our risk-adjusted model but not in our definition of a “negative” CT scan with combined head, chest and abdominal injury as the absence of cervical spine imaging in our definition may introduce bias to this composite metric. Additionally, the data contained within TQIP represents a convenience sample of voluntarily submitted data, thus our results are not nationally representative and do not include care for injured patients provided at non-trauma centers. It is possible there is more participation of Level I TCs than Level II TCs for reasons we cannot elucidate. Further, because the data contained in TQIP is de-identified in regards to patient and treating facility, we are unable to control for clustering effects by facility which may introduce bias in our results. We are also unable determine the level of trauma activation associated with record. However, it is reasonable to assume that these hemodynamically stable patients with a GCS of 15 were not a level I activation. Thus, the initial evaluation may more commonly be performed by an emergency medicine physician rather than a trauma surgeon. This difference in practice pattern may explain the increased utilization of WBCT scan that we observed in both levels of TCs for patients with minimal injury. Additionally, we recognize that AIS/ISS are not available at the time of decision to obtain CT is made. Ideally we would be able to stratify and perform our risk-adjusted analysis using the initial diagnosis made in the ED, however, this data was not available for use in the present study and as other groups have demonstrated,²⁸ inclusion of severity of injury is critical for risk-adjusted modeling in trauma care. Finally, while we were able to study a large patient population, there are limitations to the degree which we can draw conclusions based on reported statistical significance alone. Our results should be interpreted with the understanding that larger sample sizes demand smaller treatment effects in order to reach equivalent levels of statistical significance and odds ratios may overestimate risk ratios in such instances²⁹ We have chosen, in this study, to focus on, report and discuss statistically significant differences that also appear, by our best judgement, to be clinically meaningful. Notwithstanding these limitations, we believe the results to be relevant and timely to emergency medicine physicians and surgeons, in order to maximize effective use of technology and resources.

In conclusion, the results of this study show that there has been a significant change in the utilization of WBCT performed relative to selective CT over the 10 year study period and that there may be an overutilization of WBCT in trauma patients, particularly those with minimal injury and those presenting to Level II TCs. The findings have important implications and raise patient safety concerns that should be considered in order to optimize care to the injured patient while appropriately using resources and technology.