51

51

Research Forum Abstracts Conclusion: Using a single month as a cross-section of ED protocol, our findings show that a great number of head CTs are rou...

77KB Sizes 2 Downloads 44 Views

Research Forum Abstracts Conclusion: Using a single month as a cross-section of ED protocol, our findings show that a great number of head CTs are routinely performed with very few yielding significant findings. The negative predictive value of this diagnostic tool still remains an important weapon in the arsenal of the emergency department physician, but the fact that less than 4% of CTs came back positive suggests an unhealthy reliance on the CT. A historical review of CT rates shows a more than tripling of Ct rates in our hospitals in the past 10 years for ED patients. In order to reduce hospital costs, ED overcrowding and patient satisfaction, physicians may consider relying more on other diagnostic methods a more complete neurologic exam and improved history taking and less on these costly exams.

49

Cardiac-Gated Multidetector CT Coronary Angiography in Evaluation of Emergency Department Chest Pain Patients

Mahr CW, Flaherty Jr JJ, Mylonas I, Salanitri J/Evanston Northwestern Healthcare, Feinberg School of Medicine, Northwestern University, Evanston, IL

Study Objectives: Cardiac-gated multidetector computed tomography (MDCT) is an emerging diagnostic imaging modality to non-invasively assess for coronary artery disease (CAD) through coronary artery calcium scoring (CACS) and CT coronary angiography. With ECG gating, detailed visualization of the pulmonary arteries and thoracic aorta are possible as well. This study attempts to determine the utility of MDCT in emergency department (ED) patients presenting with chest pain. We hypothesized that ECG-gating would improve imaging quality and increase the sensitivity of detecting and excluding clinically significant coronary stenosis, pulmonary embolism (PE) and aortic aneurysm in patients presenting to the ED with chest pain. Methods: A prospective cross sectional study enrolled 20 ED patients with nondiagnostic ECG and troponin I levels of ⬍ 0.05 ng/dl at low or intermediate risk for acute coronary syndrome (ACS). Patients were studied using ECG-gated 16-slice MDCT CT scanner. A novel dual-head pressure contrast injection protocol was used for simultaneous opacification of both sides of the heart, the coronary arteries, the thoracic aorta, and pulmonary arteries. Axial 1.25 mm slices were acquired with curved multiplanar reconstruction and maximum intensity projection reformatted on a dedicated workstation and assessed for CACS, coronary artery stenosis, PE, and aortic aneurysm. Results: Detected pathologies included: significant coronary stenosis, (n⫽5), myocardial bridging (3), anomalous coronary artery origin (2), pericardial effusion (2), aneurismal interatrial septum (1), thoracic lymphadenopathy (1), hepatic mass (1), and hiatal hernia (1). No aortic dissection or PE was visualized. 11 patients had CACS of 0, and 5 patients had CACS ⬎ 400, 4 subsequently underwent cardiac catheterization. Coronary artery visualization was significantly improved at heart rates of 50-70 beats / minute. Extensive coronary artery calcifications made assessment of segments of coronary arterial lumen difficult. The presence of arrhythmias (PVC’s, others) precluded ECG-gating. Visualization of the pulmonary arterial tree was possible down to 4th - 5th order of branching segments, equal to standard PEprotocol-type scanning. Cardiac gating eliminated artifacts of the ascending thoracic aorta often seen on non-gated studies. Conclusion: In patients presenting to the ED with CP, 16 slice cardiac-gated MDCT assessed coronary anatomy and CACS, excluded clinically significant PE and aortic aneurysm, and detected some unusual causes of chest pain that may otherwise have gone undiagnosed. Visualization of the pulmonary vasculature was equal to standard PE protocol scans. Cardiac gating improved visualization of the thoracic aorta, eliminated motion-induced ‘pseudo-dissection artifacts’, and allowed for evaluation of clinically significant Type-A thoracic aortic dissections. The absence of CAD by CACS and CT coronary angiography may be a very sensitive negative predictor for ACS, clinically significant PE and thoracic aortic aneurysm as part of a “triple rule-out examination” and may expedite ED management and disposition of low- and intermediate-risk patients. Further investigation is warranted with larger sample sizes and longer follow-up, as well as further study with a 64 slice CT scanner.

50

Multicenter Trial Assessing the Impact of an Overnight International “Nighthawk” Teleradiology System on CT Radiology Re-Interpretation Rates

Venieris PY, Chan TC, Killeen J/University of California, San Diego, San Diego, CA; Univerisity of California, San Diego, San Diego, CA

Study Objectives: We sought to assess whether overseas attending radiologists reading nighttime abdominal and chest CT scans (also known as “nighthawk” or NH) at 2 academic medical center Emergency Departments (ED)s, improves

S16 Annals of Emergency Medicine

accuracy and reduces the number of significant overreadings that occur the following day compared to a radiology resident’s reading alone. Specifically, we determined the impact of NH on the rate of morning call backs to the ED from the on-site radiology attending reporting a significant change in radiology interpretation from the overnight radiology resident’s reading. Methods: We conducted a before-after comparison trial at an urban academic medical center (ED census 40,000) and suburban community hospital (ED census 21,000) over a 13 month period. During the first 6-months (pre-NH), all overnight chest and abdominal CT scans were read by a radiology resident and “over-read” the following morning by an on-site attending radiologist. Significant discrepancies were called to the ED and documented in the patients’ electronic ED record the following morning. Following a 1-month NH initiation period, a comparative 6 month post-NH period was also studied. During the post-NH period, all overnight radiology resident readings were overread at that time by an attending radiologist in Australia who received the study images over the Internet. As with the pre-NH period, scans read by the resident and NH were overread by an onsite attending the following morning with discrepancies reported to the ED. We reviewed the electronic ED records of all patients who had chest or abdominal CT scans performed during overnight hours (22:00-05:00) for both time periods comparing the rate of CT re-readings between the pre-NH and post-NH periods. We analyzed this data to derive rates of radiology morning call-backs, the significance of the change in reading, and whether any action or notification was relayed to the patient. Data between the 2 time periods was compared using a statistical software package (STATA) with p ⬍ 0.05 considered significant. Results: During the pre-NH period from 11/1/04- 4/30/05, 564 CT scans of the chest and abdomen were performed with only resident overnight radiology interpretation. Of these, 13 scans (2.3%) were re-interpreted with a different reading by the morning radiology attending. During the post-NH period, 635 chest and abdominal CT scans were performed. With the addition of NH readings at night along with the resident reading, there were 16 call backs by the morning onsite radiology attending for changes in the initial interpretation (2.5%). There were no statistically significant differences in morning re-interpretation and callback rates preand post-NH. In both periods, these changes in interpretation resulted in a patient notification and hospital admission for one patient during each time period. Both of these patients had uneventful hospital courses and were discharged. Conclusions: The addition of an international NH system at 2 EDs for overnight resident radiology readings did not significantly change the morning re-interpretation rate for chest and abdominal CT scans at our institutions.

51

A Study for Diagnostic Clue of Pro-ARDS in the Mice’s Lung Induced by Ischemia-Reperfusion Injuries by Using Phase-Contrast X-Ray Microscopy

June YC, Choi H, Ko Y, Hong H, Seol S, Je J/Kyung Hee Medical Center, Seoul, Republic of Korea; Postech, Pohang, Republic of Korea

Study Objectives: We have recently succeeded in observing the lungs of normaland acute respiratory distress syndrome (ARDS) induced- mice in live state with 40x⬃200x magnifying powers using phase-contrast x-ray microscopy (PCXM). The PCXM findings of the ARDS-induced mice’s lung differ from the normal mice’s PCXM findings in the initial stage; however, there is no particular guide for reading. Therefore, the objective of this study was to make appropriate diagnostic clues of proARDS by using computer imaging program and PCXM. Methods: The authors used 15 male nude mice, and were divided into two groups; 5 mice in the control group without surgical procedures, and 10 mice in the study group that was injured by ischemia-reperfusion of superior mesenteric arteries. The images of every experiment animal were taken at every 24 hours using MDCT and PCXM. Upon completion of shooting, two of the mice from the experiment group and one from the control group were killed and their images were taken by microtomography and then the mice were sacrificed to visualize the pathologic findings of lung tissue. The rest were bred until ARDS was observed in MDCT. The images were compared in mean values measured on the basis of the number of bright objects and the size of the area divided into six portions fixed to pixel size 300x300 using imaging program. ANOVA was conducted using SPSS package 11.0 for any statistical work and the significant statistical difference was in any case of p-value ⬍ 0.05. Results: In microtomogram the average thickness of alveolar walls in control group was 11.3 ␮m and in PCXM findings the average number of bright objects was 95, with 775 pixels for the average size of the area. The experiment group was divided into 3 subgroups according to the thickness of alveolar walls. Those with 12.5⬃15 ␮m as the average thickness of alveolar walls had 86 as the average number of bright objects and 664 pixels for the average area, whereas those with the average thickness

Volume , .  : October 

Research Forum Abstracts of 15⬃17.5 ␮m showed 72 and 558, respectively. In addition, those with the average thickness of 17.5⬃20 ␮m showed 63 in number and 469 in pixels, exhibiting a significant variation among the four groups. The findings of light microscopy and the dead mouse’s microtomogram showed a notable correlation. When ARDS is diagnosed by hazy density on MDCT, the average thickness of the alveolar walls was greater than 20 ␮m. Conclusion: The average thickness of alveolar walls is significantly related to the average number of bright objects and the average size of the area measured using the imaging program after images taken from PCXM. ARDS was developed in most of the mice that exhibited this type of change, which in turn suggests that such a change may exhibit possible findings for preceding stages of ARDS. If further studies are conducted on various diagnostic clues, it is with no doubt many breakthrough researches on unknown pathogeneses of diverse pulmonary diseases can be successfully accomplished.

completed their evaluation in the ED while awaiting an inpatient bed. An additional 33 (19%) would have been admitted despite MDCT findings: 19 (11%) with an elevated biomarker, 10 (6%) with an ECG suggesting ACS, and 4(2%) with hemodynamical instability. The remaining 114 (66 %; 95% CI 59-73%) were considered patients who might have benefited from an ED MDCT. Among them, 9 (8%) were ineligible because they were in atrial fibrillation, 22 (19%) had serum creatinines over 1.3 ng/ml, 3 (3%) required another study requiring contrast, and one patient (1%) was taking metformin. Hence 69% (95 % CI 60-78%) of ED patients who might have benefited from an MDCT had no contraindications. Conclusion: Based on single-center pilot data, a minority of ED patients with a planned admission for chest pain would not receive an MDCT because of consent issues or because it would not help in the admission decision. Among those who might benefit from this technology as a triage tool, over two-thirds are without contraindications.

52

54

Computed Tomagraphy of the Abdomen Versus Right Upper Quadrant Ultrasonagraphy for Diagnosis of Acute Gallbladder Disease

Hsu L, Rozeski JE, Aubin C, Peterson CM, Menias CO, Lewis LM/Washington University of St. Louis, Saint Louis, MO

Background: Access to ultrasonagraphy (US) to diagnosis acute gallbladder (GB) disease is limited during off peak hours. It would be useful to the Emergency Physician (EP) to know if Computed Tomography (CT) is adequate to rule out acute GB disease. Study Objectives: We hypothesized that CT of the abdomen is as sensitive as US to rule out acute GB disease. Methods: Medical charts of all patients (October 2003 to March 2005) who presented with abdominal pain and received both US and CT within 48hrs of one another were retrospectively reviewed. Patients who had CT or US for procedures (biopsies or paracentesis), or isolated liver/kidney evaluation were excluded. Sensitivity (SEN), specificity (SPE), and negative predictive value (NPV) were calculated using standard formulae and compared between CT and US using 95% confidence intervals. Inter-test agreement was calculated using the kappa statistic. Results: 234 patients received both US and CT within 48 hours of each other during the study period. 27 of these patients were identified to have acute GB disease (22 acute cholecystitis, 2 cholangitis, and 3 choledocholithiasis): 13 confirmed by pathology, 6 by percutaneous cholecystostomy tube placement, 7 by endoscopic retrograde cholangiopancreatography, and 1 by autopsy. CT identified 22 of 27 cases of acute GB disease with 2 false positives: SEN ⫽ 81.5% (95% CI 66.9% to 96.1%), SPE 98.8% (95% CI 97.1% to 100%), NPV ⫽ 97.1% (95% CI 94.5% to 99.6%). US identified 23 patients with acute gallbladder disease with no false positives: SEN 85.2% (95% CI 72% to 99%), SPE 100%, NPV ⫽ 97.7% (95% CI 95% to 100%). There was good agreement between CT and US in diagnosing acute GB disease (kappa ⫽ 0.83). Conclusions: CT of the abdomen is as sensitive as US for identifying acute GB disease and may be used to screen for acute GB disease when US is not readily available or there are other strong considerations in the differential diagnosis.

53

How Often Do Emergency Department Patients with Chest Pain Meet Eligibility Criteria for Multi-Detector Computerized Tomography of the Coronary Arteries?

Nagurney JT, Moselewski F, Pena AJ, Nichols JH, Parry BA, Butler J, Manini AF, Brown DF, Siebert U, Hoffmann U/Massachusetts General Hospital, Boston, MA

Study Objective: To determine, among ED patients being admitted for acute chest pain after a non-diagnostic workup, the percent who would be eligible for a multi-detector CT scan (MDCT) of the coronary arteries. Methods: Type of study: A descriptive single-site pilot study. Setting: A 78,000visit academic ED. The intake period was 24/7 from 5/21/2005 to 6/4/2005. Subjects: Adult ED pts with chest pain and a planned admission for a “rule-out MI” protocol. Observations: Patients were determined to be eligible for MDCT if they lacked four exclusion criteria (outlined below). The potential pool of patients was derived by combining an ED log and a billing database. A researcher interviewed patients and reviewed all medical records. Analysis: Simple descriptive statistics including 95% CIs. Results: Over this period, 174 patients had in-patient beds requested. The median age of this population was 58 years (range 25-92; IQRs 47-72); 114 (66 %) were men. Overall, 16 patients (9%) were unable to consent and another 11 (6%)

Volume , .  : October 

Prevalence and Demographics of MRSA in Culturable Skin and Soft Tissue Infections in an Urban Emergency Department

Jacobus CH, Leach SD, Lindsell CJ, Kressel AB, Fermann GJ, Rue LE/University Hospital, Cincinnati, OH; Department of Emergency Medicine, University of Cincinnati, Cincinnati, OH; Department of Medicine, Division of Infectious Diseases, University of Cincinnati, Cincinnati, OH; University of Kentucky College of Medicine, Lexington, KY

The rising incidence of methicillin resistant Staph. Aureus skin and soft tissue infections (MRSSTIs) is a concern for emergency practitioners. While many studies have examined MRSA in inpatients, few have focused on ED populations. Study Objectives: To describe the demographics and predictors of MRSSTIs in an ED population. Methods: A convenience sample of 182 patients with a culturable skin infection presenting to an urban ED was enrolled. Patients provided demographic and risk factor information by structured interview. Infections were cultured for MRSA in the hospital lab. The predictive value of each risk factor for MRSA was tested using univariable logistic regression, and a multivariable model was developed. Results: Overall prevalence of MRSA was 58% (95%CI 50 - 65%). Variables associated with MRSA infection were age (OR 0.95; 95%CI 0.93 - 0.97), BMI (OR 0.99; 95%CI 0.91 - 0.99), sexual contact in the past month (OR 2.81; 95%CI 1.49 5.30), presence of a discrete collection of purulence (OR 4.54; 95%CI 1.96 - 10.5), spontaneous infection (OR 2.53; 95%CI 1.34 - 4.78), and incarceration or close contact with an inmate in the past year (OR 2.00; 95%CI 1.06 - 3.77). The best-fit multivariable model had a C-statistic of 0.73 with four significant variables: age (OR 0.97; 95%CI 0.945 - 0.997), group home living (OR 4.71; 95%CI 1.17 - 18.9), presence of a discrete collection of purulence (OR 3.08; 95%CI 1.23 - 7.75), and sexual contact within one month (OR 2.15; 95%CI 1.05 - 4.42). Conclusion: In this population of ED patients, MRSSTI was related to decreased age, recent sexual contact, presence of abscess, low BMI, spontaneity of infection, incarceration or contact with an inmate, and group home living. Other traditional risk factors were found not to be significant. Some of the risk factors shown here have not previously been described.

55

The Prevalence of Community-Acquired MethicillinResistant Staphyloccus Aureus in a Northeast Urban Level 1 Trauma Center

Dalley MT, Schneid S, Gernsheimer J, Gernsheimer J, Hosford K/Lincoln Medical Center, Bronx, NY

Study Objectives: To determine the prevalence of methicillin resistant Staphylococcus aureus (MRSA) among ED patients who presented with furuncles. Characterize MRSA by antimicrobial sensitivity. Trend treatment outcomes (Abx Tx, I&D, both) with our follow-up clinic. Methods: Obtain cultures from the next 100 skin abscess or furuncles presenting to the ED. Demographic information obtained from a brief questionnaire along with antibiotic Rx at discharge. Patients included in the study were asked to fullow up in our follow-up clinic in 24 to 48 hours at this time; outcomes such as worsening infection, admission to the hospital, and reevaluation in the ED were collected. Results: Ongoing. Currently we have 38 patients included in our study. Of the 38 infection site cultures 11 cultures grew MRSA (28%), 7 MSSA, Coag neg staph 3, group A strep 3, Enterococcus 1, Strep viridians 1, Kleb PNA 1, Ecoli 1 no growth 8,

Annals of Emergency Medicine S17