1) Have a clear understanding of the newer clinical criteria for imaging in pediatric head trauma based on the more recent large multicenter studies. 2) Optimization of the imaging protocols to enhancing the diagnostic performance. 3) The importance of integrating the pretest probability (clinical criteria) and the diagnostic test in order to have the highest posttest probability (probability after the imaging study).

Although magnetic resonance imaging is routinely utilized in cases of suspected non accidental head trauma, little data exists regarding the use of imaging to evaluate for associated cervical spinal ligamentous injury. Furthermore, the association between ligamentous cervical injury and intracranial abnormalities on MRI has not been documented. Through retrospective review of MRI brain examinations, we aim to establish the value of fast spin-echo inversion-recovery (FSE-IR) in assessing for cervical spinal ligamentous injury in cases of suspected abusive head trauma.

MRI brain examinations performed in all cases of suspected non accidental head trauma between 2010 and 2013 were retrospectively reviewed. First, the fast spin-echo inversion-recovery (FSE-IR) sequence was examinated on each study to evaluate for hyperintense signal in the apical, anterior longitudinal, posterior longitudinal and interspinous ligaments. Subsequently, each positive study was evaluated for abnormal signal intensity on diffusion-weighted imaging, susceptibility-weighted imaging and on T2*.

A total of 60 patients with non accidental head trauma received MRI brain examinations in our institution between January 2010 and December 2013. Of these patients, 17 (29%) were found to have ligamentous injury on FSE-IR. Additional findings of severe trauma were also present on other MR sequences in all patients. Hypoxic ischemic injury, detected on diffusion-weighted imaging, was present in 10 patients (59%). Retinal hemorrhages, seen on the T2* sequence, were identified in 8 patients (47%) with concomitant ligamentous injury. Cortical venous thrombosis, detected on either susceptibility-weighted imaging or T2*, was present in 16 patients (94%).

The fast spin-echo inversion-recovery (FSE- IR) sequence detects cervical ligamentous injury in patients with non accidental head trauma and is associated with significant intracranial injuries including hypoxic-ischemic injury, thrombosed cortical veins and retinal hemorrhages. FSE-IR should be performed routinely in all cases of suspected abusive head trauma.

Fast spin-echo inversion-recovery detects ligamentous cervical spinal injury and should be routinely used whenever non accidental trauma is suspected.

To identify the incidence and clinical predictors of facial fracture in the setting of whole-body multi-detector computed tomography (MDCT) for trauma.

500 consecutive patients who received dedicated maxillofacial CT as part of whole-body MDCT for trauma were studied. Patients younger than 18 and those who received initial evaluation at an outside facility were excluded. Fracture incidence and clinical parameters were obtained from the electronic medical record. Clinical and demographic variables were compared between patients who had an acute fracture and those who did not. Two sample t-tests were used to compare continuous variables, and the Fisher's exact tests were used to compare categorical variables.

A total of 221 (44.2%) patients had acute fracture demonstrated on the maxillofacial CT. In all, 470 (94.0%) patients had documented positive facial physical exam findings at presentation. Of the 30 patients without exam findings, 29 (negative predictive value = 96.7%) did not have a facial fracture. Orbital fractures were most common overall, seen in 52.5% of positive cases. Nasal fractures were the most common isolated fracture (18.6% of all fractures; 42.3% of isolated fractures). Statistically significant difference was found between positive and negative cases of facial fracture in GCS score ≤ 8 (p <0.0001), intubated at presentation (p<0.0001), Injury Severity Score (ISS) of ≥ 16 (p<0.0001), positive facial physical exam (p<0.0001), and loss of consciousness (p = 0.0299). By history, the highest fracture rates were seen in falls from elevation or standing height and open-vehicle accidents (80.0%, 58.9%, and 55.2%, respectively).

The absence of physical exam findings reliably excludes facial fractures. Clinical variables that positively associate with facial fracture include: GCS ≤ 8, ISS ≥ 16, intubated status, positive loss of consciousness, and presence of facial physical exam findings.

These data can support clinical decision-making by identifying those at greatest risk for facial fracture and those who are less likely to have a fracture based on the initial clinical survey.

1) Understand the evidence for best practices in cervical spine imaging of trauma. 2) Develop an evidence based approach to selection of appropriate imaging in cervical spine trauma.

There is abundant evidence on when it is appropriate to image the cervical spine in trauma victims, and with which imaging modality. However, controversies persist. This session will focus on the evidence supporting the roles of CT, MRI, and radiography in cervical spine imaging. We will discuss special populations, including children, the elderly, obtunded patients, patients with neurological deficits, and patients with spinal fusion. Included will be a discussion of accuracy of imaging, cost effectiveness analysis, and use of clinical prediction rules to risk-stratify subjects.

Blunt carotid and vertebral artery injuries (BCVI) can cause devastating ischemic neurologic events. The Denver criteria are often used to guide BCVI screening and include all patients with C2 fractures (fxs). We hypothesize that patients with ground level falls (GLF) and isolated dens fxs (IDF) have a very low risk of BCVI and do not require vascular imaging.

All patients with C2 fxs in the hospital trauma registry from 2006-2012 were retrospectively reviewed. Age, sex, injury mechanism (GLF or non-GLF, a higher risk mechanism), C2 fracture type (IDF or other C2 fracture (OthC2F)), vascular imaging type, and Biffl injury grade were evaluated.

Of 789 subjects with C2 fxs, 176 (22%) had IDF and 613 (78%) had OthC2F. 538 of 789 (68%) subjects underwent vascular imaging, and 141 (26%) had BCVI. 76 of 176 (43%) patients with IDF underwent vascular imaging and 6 (8%) had BCVI. Of the 31 subjects with type 1 or 2 IDF and vascular imaging, 1 (3%) had BCVI compared to 5/45 (11%) with type 3 IDF. Of the 462 patients with OthC2F and vascular imaging, 135 (29%) had BCVI. There was a significantly decreased prevalence of BCVI in IDF compared to OthC2F (p<0.001). In 31 patients with IDF after GLF who had vascular imaging, only 1 (3%) patient with a type 3 IDF had a BCVI, compared to 35/144 (24%) with OthC2F. Compared to patients with OthC2F and non-GLF, there was an odds ratio of 0.11 for vascular injury in patients with IDF and GLF. There was a significantly decreased risk of BCVI in patients with IDF (p=0.0002) and GLF (p=0.02) compared to patients with OthC2F and non-GLF.

In patients with vascular imaging, only 8% with IDF had BCVI compared to 29% of those with OthC2F. The rate of BCVI in IDF sustained after GLF is low (1/31), and no patients with type 2 IDF after GLF had BCVI. Thus, these patients may not require routine screening, suggesting the need for further evaluation of the Denver criteria to decrease unnecessary imaging utilization. The rate of BCVI in OthC2F is higher (24-29%) and these patients should be screened regardless of injury mechanism.

Patients with type 2 isolated dens fractures resulting from ground level falls may not require screening for BCVI. Patients with other C2 fractures regardless of mechanism should be screened.

To assess the value of multidetector computed tomography (MDCT) in patients with acute pancreatitis and suspected pancreatic necrosis with regard to both lab tests (C-reactive protein, lipase, creatinine) and histopathology.

102 consecutive patients with acute pancreatitis and suspected pancreatic necrosis underwent contrast-enhanced MDCT. Two blinded readers assigned patients into one of three groups (GR). Patients in GR1 showed edematous organ swelling, peripancreatic fluid collection, and pseudocysts; patients in GR2 showed necrotic collection and a lack of pancreatic parenchymal contrast-enhancement; and patients in GR3 had no evidence of pancreatitis. Findings were correlated with results from pancreatic surgery and guided fine-needle aspiration (FNA). Mann-Whitney's U test was used to evaluate significant differences in lab findings between the groups. Cut-off values were calculated using ROC curve analysis.

Using MDCT, 54/102 patients (52.9%) were classified as GR1, 17/102 patients (16.7%) as GR2, and 31/102 patients (30.4%) as GR3. 13/17 patients (76.5%) in GR2 underwent either surgery (n=6, 46.2%) or FNA (n=7, 53.8%) and pancreatic necrosis was confirmed histopathologically in all of them. Statistical analysis showed significant CRP differences between GR2 vs. GR3 (p=0.001; cut-off point: 82mg/L; AUC 0.76) as well as between GR1 vs. GR3 (p<0.001; cut-off point: 98mg/L; AUC 0.84). The comparison between GR2 vs. GR3 and GR1 vs. GR3 revealed no significantly different lipase (p=0.35; AUC 0.58/p=0.85; AUC 0.52) or creatinine levels (p=0.96; AUC 0.5/p=0.24; AUC 0.6).

In patients with acute pancreatitis, MDCT may help when CRP values are highly elevated to rule out complications such as pancreatic necrosis. In contrast, lipase and creatinine are poor predictors.

Patients with clinically suspected pancreatic necrosis and mild to moderate elevated lab parameters could be saved from unnecessary MDCT examinations.

1) Review the evidence behind the use of CTA for acute pulmonary embolism (PE) 2) Discuss the concept of overdiagnosis 3) Review ACR appropriateness criteria for the disgnosis of PE

Pulmonary embolism (PE) is a relative common, potentially fatal disease which remains a challenge in the daily clinical practice. Computed tomography pulmonary angiography (CTAP) has become the gold-standard non-invasive test in patients with suspected PE. Several studies have shown that due to the introduction of Multi-detectors-Computed-Tomography (MDCT) the sensitivity of CTAP increased significantly compared to single-detector-CT without changing the PE mortality rate. In July 2012, the MDCT-scanner in our ED was changed from a 16-row to a 128-row single source MDCT scanner. The aim of this retrospective study was to evaluate if the number of diagnoses of possibly clinically irrelevant solitary subsegmental PE (SPE) has increased after changing from a 16-row to a 128-row-MDCT-scanner.

We included all CT-scans requested by the ED between January 1st, 2005 and December 31st, 2013. All scans before July 16th, 2012 were performed with 16-row-MDCT-scanner (Siemens Somatom Sensation 16), and thereafter, with a 128-row-MDCT- scanner (Siemens Somatom Edge). All examinations were performed with standard protocols for pulmonary embolism, triple-rule-out or poly-trauma.

3,533 examinations were included in this retrospective analysis, 2,661 with the 16-row (mean age: 49.2 years; male: 56.4%) and 872 with the 128-row scanner (mean age: 60.7 years; male: 60.2%). There were no significant differences in the number of PE or SPE diagnoses before and after change of the scanner. PE was diagnosed in 388/2,661 cases (14,6%) with the 16-row-scanner and in 118/872 cases (13,5%) with the 128-row-scanner (p=.44), SPE was diagnosed in 69/2,261 (2,6%) and in 24/872 cases (2.8%), respectively (p=.80).

Changing from a 16-row- to a 128-rowMDCT-scanner will not increase the number of possibly clinically irrelevant SPE and, therefore no further increase in unnecessary thrombolytic therapies based on radiological diagnoses has to be expected.

Although the detection rate of possibly clinically irrelevant SPE increased significantly after the introduction of MDCT compared to single-detector-CT, it seems that there is no further increase in the detection rate changing from an 16-row- to a 128-row scanner.

Compared to original trials which derived pre-test probability systems for suspected pulmonary embolism (PE), the prevalence of PE at each given level of pre-test probability has decreased. Consequently, higher values of d-dimer may safely exclude PE in suspected cases. We therefore examined the implications of increasing the d-dimer threshold for patients with decreasing clinical pretest probability.

Consecutive CT pulmonary angiography (CTPA) exams performed for suspected PE over a 14 month period were retrospectively identified and final interpretations were recorded. Data to calculate the Revised Geneva Score (RGS) for each encounter were extracted from the electronic medical record by electronic means and manual review, and d-dimer values were collected. All patient encounters for which pretest probability was calculated as low (RGS 0-3) or intermediate (RGS 4-10) and for which d-dimer testing was performed were included in the study. The prevalence of PE for low and intermediate probability patients with d-dimer values below adjusted thresholds was then determined.

Of 3500 CTPA exams performed, 1745 involved encounters for patients with low or intermediate probability and d-dimer testing performed. The remainder included 167 with high probability, and 1588 with low to intermediate probability and no d-dimer testing performed. Intermediate probability patients had a slightly higher mean age (53.2 vs. 50.1 years, p=0.001), but there was no significant difference in the prevalence of PE for low and intermediate probability patients at d-dimer levels below 1000 (3.7% vs. 2.5%, p=0.29). For both groups combined, prevalence of PE remained below 2% with a threshold of 700 (1.8%, 95% CI 1.1-3.1%), which accounted for 41% of the CTPA exams.

Prevalence of PE is not significantly different between patients with low and intermediate pretest probability at d-dimer levels below 1000. Prevalence of PE remains below 2% for all low and intermediate probability patients below 700, and 41% of the CTPA exams could be avoided if this level was used to exclude PE. Prospective management studies to select the optimal adjustment of d-dimer are necessary before clinical implementation may occur.

CTPA utilization could be substantially reduced if d-dimer thresholds were increased for exclusion of PE for patients with both low and intermediate pretest probability.

The purpose of the study was to validate the hypothesis that discontinuing the use of oral contrast (OC) for MDCT will not affect the detection of acute abdominal abnormalities in emergency room (ER) patients.

We conducted a retrospective study to assess the effect of eliminating OC use for 64MDCT scans of the abdomen and pelvis (AP) for patients presenting with acute abdominal pain to ER and BMI greater than 25. Patients with BMI less than 25 continued to receive OC. Only patients who underwent AP 64MDCT imaging in the portal venous phase without OC were included. The study was approved by the REB. Informed consent was waived. The electronic medical records were reviewed to determine the rate of repeat imaging within seven days from initial CT scan, as well as delayed or missed diagnoses related to the lack of OC.

1378 patients had an AP 64MDCT between November 1, 2012 and October 31, 2013. 375 patients met the inclusion criteria (174 males and 201 females, mean age 57, range 18-97). 7/375(1.9%) patients had repeat CT examination with OC within 7 days. Of these 7 patients, none had a change in the course of their management due to the utilization of OC. No delayed or missed diagnoses related to the lack of OC were identified.

Omitting OC for imaging patients with BMI greater than 25 presenting with acute abdominal pain in an ER setting resulted in no delayed or missed diagnoses. The benefits of prompt imaging diagnosis outweighs the minimal potential need for repeat imaging.

64MDCT evaluation of ER patients with acute abdominal pain can be safely performed without oral contrast.

1) Review pertinent osseous landmarks of the hip joint. 2) Translate evidence based knowledge to the observed imaging findings. 3) Discuss important features that the surgeons needs to know. 4) Employ ACR appropriateness criteria for CT and MRI utilization.

To determine if there is a difference in outcomes for patients with suspected fragility fractures at the femoral neck evaluated using CT in comparison to those evaluated with MRI.

This study was an observational retrospective cohort design. Imaging studies ordered through the emergency department at Milton S. Hershey Medical Center were searched to identify participants. Due to higher prevalence of radiographic occult osteoporotic fractures of the hip, the study was limited to females over the age of 65 who had experienced a fall and were suspected of having a fracture. Patients were included if initial radiographic evaluation of the pelvis or hip was negative for fracture and followed by either a non-contrast CT or MRI of the hip or pelvis. Records were grouped based on whether evaluation for occult fracture was made with CT or MRI. Their electronic medical records were searched for hospitalizations in the year following their index evaluation in the emergency department. Our primary analysis was to compare outcomes thought to be directly related to delayed diagnosis or treatment of suspected fragility fractures. These outcomes include death, secondary displacement, avascular necrosis, rapid progression of osteoarthritis, delayed diagnosis, and malunion. These were determined based on a literature search conducted prior to the collection of data.

A database search from the dates of 1/1/05 to 12/31/12 yielded 926 records. After review, 119 of these records met the inclusion criteria; 21 were evaluated with MRI and 98 with CT. Statistical analysis showed no significant difference in clinical outcomes during the year following initial evaluation. 8.3% of patients evaluated with MRI experienced adverse outcomes directly related to hip fracture within one year in comparison to 10.6% of patients evaluated with CT (p=1.0). The most common of these adverse events in both groups was death within the following year.

In contrast to current guidelines that recommend MRI, our results indicate there is no significant difference in clinical outcomes between elderly female patients evaluated with CT or MRI for suspected fragility fractures of the hip following fall.

Appropriateness criteria recommending MRI for suspected occult fragility fractures are based on studies of diagnostic accuracy; however, our results suggest no difference in patient centered outcome.

To determine factors that increase the likelihood of new or progressive lumbar MRI findings in patients with a prior MRI.

Retrospective review was performed on ED patients with a lumbar MRI who had a prior MRI within 6 years. Demographics and 'red flag" symptoms (malignancy, infection, fracture) were recorded. Lumbar MRIs were reviewed for acute findings (infection, new tumor, fracture, disc herniation, cord compression). Degenerative change was considered predominately facet joint arthropathy (FJA), degenerative disc disease (DDD), or both (BFD), and categorized as single or multi-level change. Images were compared with the prior MRI to assess change in canal stenosis. Odds Ratio analyses evaluated likelihood of worsening canal stenosis for single vs multi-level change, FJA vs DDD, and either FJA or DDD vs BFD.

285 lumbar MRIs were performed on patients with prior MRIs within 6 years. 7 cases were excluded. 136 patients (49%) had a 'red flag'. There were 66 cases with acute findings, 34 of which were patients with malignancy. Among 212 cases without an acute finding, 44 had more than 1 repeat exam and 16 had no degenerative change. As a result, 152 cases were evaluated for change in spinal canal stenosis. 42 (28%) had single level degenerative change and 110 (72%) had multilevel change. More patients with multilevel changes exhibited worsening spinal canal stenosis over time than patients with single level degeneration (OR 8.95, CI 2.0-39.2). There was no significant difference in the change in canal stenosis between patients with predominately FJA or DDD. More patients with BFD had worsening canal stenosis over time than patients with FJA or DDD (OR 2.9, CI 1.33-6.29).

ED lumbar spine MRIs are commonly performed when prior MRIs exist. Clinical "red flags" increase the likelihood of acute findings, consistent with previously published data. Patients with single level degeneration and no acute finding are less likely to have progressive spinal canal stenosis. Progression is more likely in patients with both FJA and DDD, than in those with one or the other.

This research will help clarify the role of repeat lumbar MRIs in the ED for patients with various low back pain presentations and will allow for more prudent use of a limited imaging resource. Additionally we aim to explore which lumbar degenerative risk factors predispose to worsening spinal canal stenosis over time.