| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||
ABSTRACT |
Tuesday, May 8, 11:20 AM–12:30 PM
Abstracts 091-097
Moderator(s): Michael Bruno and Diego Nunez
11:20 AM
091. A Survey of Radiology Practices' Use of After-hours Radiology Services
Kaye A.3*; Kapoor R.2; Forman H.3; Sunshine J.1 1. American College of Radiology, Washington, DC; 2. The George Washington University School of Medicine, Washington, DC; 3. Yale University School of Medicine, New Haven, CT
Address correspondence to A. Kaye (adam.kaye{at}yale.edu)
Objective: To identify the motives behind radiologists' use of `nighthawk' services; to provide a baseline for future research of nighthawk practices; and insight into a growing force in the economics and practice of diagnostic radiology.
Materials and Methods: 300 hospitals were contacted by phone, email, and mail, with an attempt made to speak to the chief of radiology. The hospitals were selected randomly from the 2005 American Hospital Association Directory of Hospitals. The survey consisted of 59 questions. We obtained 115 surveys, a response rate of 38.3%, including 63 practices that use an external nighthawk. The results were analyzed using descriptive statistical analyses. Questions related to motivation, demographics, and selection of nighthawk services, financial arrangements, and level of training.
Results: Most practices gave convenience as the most important reason they use a nighthawk. The second most important factor was the value for recruiting. The least important factor was efficiency. Excessive volume was also a common motivation. 51% of practices said they obtain between 1–5% of their reads from the nighthawk, and 22% of the practices said they receive less than 1%. 14% of practices received between 6–10% of their reads from nighthawks, and the remaining 13% received greater than 10%. 40% of practices were paying about 100% of what they collected for each study, and another 14% paid greater than 100%. About 17% did not know how much they were paying. 6% paid between 81–99%, 17% paid between 51–80%. Two practices paid between 26 and 50%, and only one practice paid less than 25%. Out of the 63 practices surveyed, 25 or about 40%, utilized a nighthawk located internationally. Of these, 40% did not know the proportion of foreigners or Americans reading films. Another 56% said that the radiologists reading internationally were either all Americans or mostly Americans, and one practice said it was about 50% Americans.
Conclusion: Most radiology groups are using nighthawks for convenience rather than necessity. Most practices send a very small percentage of their studies to the nighthawk, and the majority of those practices surveyed were not making profit on this arrangement. While the general public may have fears of outsourcing radiology, 40% of practices used international nighthawks. The majority of radiologists at these are American citizens, and probably American-trained physicians fully capable of reading studies suitably.
092. Ultra-Low-Dose MDCT of Abdomen and Pelvis in Patients with Acute Nonspecific Abdominal Pain: Impact on Patient Care and Management
Udayasankar U. K.1*; Li J.1; Kalra M. K.2; Baumgarten D.1; Small W. C.1 1. Emory University School of Medicine, Atlanta, GA; 2. Massachusetts General Hospital, Boston, MA
Address correspondence to U. Udayasankar (uudayas{at}emory.org)
Objective: The purpose of this study was to assess the impact of an ultra-lowdose multi-detector row CT (MDCT) of abdomen and pelvis on patient care and management, in patients presenting with acute specific abdominal pain.
Materials and Methods: Once hundred and sixty-three consecutive patients (mean age, 54; age range, 18–94; M:F, 53:110), who presented with acute nonspecific abdominal pain and who underwent ultra-low dose CT of abdomen and pelvis were included in this IRB-approved retrospective study. All patients were evaluated on a 64-slice MDCT, with scan parameters of 120 kV, 12.5–100 mAs range (automatic tube current modulation at noise indices of 25 or 35) at 5-mm interval. The effective radiation doses were calculated from the scan-generated dose-length products. Average period of stay in the hospital was estimated. The reports of the radiological studies were compared with those of further diagnostic work up, clinical and surgical findings and followup patient visits up to 6 months.
Results: 92% (150/163) of the studies were diagnostically acceptable. The mean radiation dose from this study was estimated to be 2.01 mSv (range of 0.67 to 6.64). Nineteen patients (11%) underwent contrast enhanced study of the abdomen and pelvis within 48 hours of the ultra-low-dose CT study. Mean duration of hospital admission was 2.4 days. 37.1% (61/163) of the patients were discharged the same day from the emergency department based on a negative CT study. Ultra-low-dose CT was shown to have a sensitivity of 92.8% with a negative predictive value of 92.6%.
Conclusion: Ultra low dose MDCT of abdomen and pelvis are useful in providing the relevant information and reducing hospital stay in patients with acute nonspecific abdominal pain.
093. Comparison of Automated CT Reconstructions from 64-MDCT with Manual Reformats from 16-MDCT in Emergency Radiology Setting: An Image Quality Analysis Based on Patient Body Weight and Workflow Considerations
Singh A. K.*; Sahani D. V.; Ouellette-Piazzo K.; Novelline R. A. Massachusetts General Hospital, Boston, MA
Address correspondence to A. Singh (dranandsingh{at}yahoo.com)
Objective: Our aim was to compare the diagnostic quality of coronal and sagittal reformations obtained manually from 16-MDCT and as automated reformats from 64-MDCT scanner console in emergency radiology and study the impact on the CT workflow.
Materials and Methods: In this study, 2 radiologists (R1, R2) evaluated coronal and sagittal reformats of patients who were scanned on either 16- or 64-MDCT scanners [GEHC] in the emergency radiology department (100 patients on each scanner–total of 200) for thorax, abdomen, or pelvis. Reformats at the 16-slice MDCT console was generated manually by the technologists at the console (GE Xtream) whereas they were available as automatically generated reconstructions during the scan at 64-slice MDCT console. Time for generation of images at both consoles was recorded. The patients were categorized in 3 different weight groups according to their body weight (A- <200, B- 200–250 and C- >250 lbs). Reformats were then graded for image quality (IQ) and diagnostic confidence (DC) on a 5-point scale.
Results: The IQ and DC of readers for reformats from 64-MDCT was significantly better than 16-slice as it gave high DC grading for heavy weight patients due to improved IQ [64-MDCT–mean IQ of 3 groups 4.5; 16-MDCT–mean IQ of 3 groups 3.3]. Mean DC of 64- and 16-MDCT patient groups were 4.7 and 3.8, respectively. These findings were statistically significant (p < 0.002). Average time for generation of the images was 3.8 minutes for 16-MDCT and 1.1 minute for 64-MDCT.
Conclusion: Automated coronal and sagittal reformations from 64-MDCT are consistently of superior quality even in heavy patients in comparison to the technologist-generated reformations from 16-MDCT. Instantaneous and automated generation of these reconstructions with the axial data sets at the 64-MDCT facilitates the CT workflow in an emergency setting.
094. Overnight Resident Interpretation of Conventional Radiographs in the Emergency Department: A Simple Method to Track Discrepancies from the Final Report for Quality Assurance and Education
Strub W. M.; Leach J. L.*; Moulton J. S.; Vagal A. University of Cincinnati, Cincinnati, OH
Address correspondence to J. Leach (james.leach{at}uc.edu)
Objective: With such a large volume of cases reviewed in the emergency department, an efficient way to track discrepancies from the preliminary report generated by a resident, rather than completing a traditional form that can easily be lost, must be created. The purpose of this study is to report our experience with Powerscribe Workstation (Nuance; Burlington, MA) voice recognition software as a means to track discrepancies on overnight resident preliminary reads on radiographs in the emergency department and use those results to enhance resident education.
Materials and Methods: From October 1, 2005 to July 31, 2006, radiographs were prospectively interpreted by residents at night at a Level I Trauma Center. Preliminary readings were communicated to the Emergency Department electronically after the residents report generated by Powerscribe Workstation (Nuance; Burlington, MA) voice recognition software directly entered the radiology information system and became a permanent part of the patient's medical record. The examinations were marked as resident preliminary dictations so they could be easily retrieved later for quality assurance purposes. The radiographs were reviewed between 6:00 a.m. and 7:00 a.m. the following morning by the body imaging staff. Discrepancies were documented in the final dictation of the report by using a standard phrase in the voice recognition software program.
Results: 13,213 conventional radiographic examinations were initially interpreted by a resident radiologist during the study period. 124 discrepancies were identified which yielded a disagreement rate of 0.938%. Radiographs of the chest and spine accounted for the greatest number of studies performed and the greatest number of discrepancies. The overall discrepancy rate was shown to be constant over time. The preliminary reporting process and recording of discrepancies was well accepted by the residents and the faculty of the radiology and emergency medicine departments. The discrepancies were reviewed with the residents each quarter as an ongoing part of their educational process.
Conclusion: We observed a low discrepancy rate on radiographs with regards to the overnight preliminary interpretation by the resident and that of the final report. The use of voice recognition software coding provides a very efficient way to track any discrepancies for quality assurance and resident education purposes, and could be applied to any location that uses voice recognition software.
095. CT Test Ordering Behavior of College-Based and ER Physicians and their Correspondence to ACR Appropriateness Criteria Guidelines
Baker S. R.2*; Pan L.2; Susman P.1; Wolaver A.1; Vidgop Y.2 1. Bucknell University, Lewisburg, PA; 2. UMDNJ-New Jersey Medical School, Newark, NJ
Address correspondence to S. Baker (bakersr{at}umdnj.edu)
Objective: Computed tomography (CT) has increased in use for a range of acute abdominal conditions. But access to CT has also engendered questions of overutilization. Appropriateness Criteria Guidelines, developed by the ACR, are an attempt to align technology with clinical presentation so that the best rather than the most extensive battery of tests are performed. Yet what is suggested by radiologists may not correspond to what is ordered by referring physicians. This study sought to evaluate CT-ordering behavior of emergency and college-based physicians throughout the United States with regard to two acute conditions. In addition to charting the extent of reliance on CT, we endeavored to assess various factors influencing the decision to request or refrain from choosing CT.
Materials and Methods: A survey was sent to all college health and emergency physicians seeking anonymous responses to a series of questions related to CT utilization in young patients who presented with either one of two clinical scenarios. 1. Right lower quadrant pain, fever, and leukocytosis. 2. Diffuse abdominal pain, no guarding, and normal white count. Respondents were asked to provide information related to factors which influenced their decision to seek a CT study.
Results: The response rates were 23.1% and 20.8% for college and emergency physicians, respectively. For patients of both genders presenting with right lower quadrant pain, tenderness, fever, and leukocytosis, 76% of both physician groups would request CT. The differences in this category were significant for female patients for which ER physicians were more likely to request CT than college physicians. In patients with diffuse abdominal pain, no guarding, and a normal white count, 36% of emergency physicians would request CT, but only 11% of college physicians would seek that test. When asked to assess the influence of factors, distinguished as very, some, little, or of no importance, the two respondent groups agreed that physical examination was very important. But a relationship with a radiologist was deemed not important by more than 50% of physicians and very important by less than 6% in both groups.
Conclusion: The data suggest that the decision to choose CT in these two acute conditions is based upon other factors than radiologist's input suggesting that ACR Appropriateness Criteria Guidelines may not be heeded or even considered by referring physicians.
096. Process Modification and Emergency Department Radiology Service
DeFlorio R.*; Coughlin B.; Santoro J.; Akey B.; Favreau M. Baystate Medical Center, Springfield, MA
Address correspondence to R. DeFlorio (rdeflori{at}yahoo.com)
Objective: To demonstrate the impact of changes in technology, staffing, and departmental processes on service levels in Emergency Department (ED) radiology.
Materials and Methods: ED radiology performance was evaluated
before and after modifications of processes integral to the interpretation of
ED imaging at a tertiary care medical center with approximately 110,000 ED
visits per year. PACS, voice recognition (VR), staffing, physical site, work
flow, and administrative modifications were undertaken over a period of 
2
years. Serial metrics such as time interval between exam completion and time
dictated and time signed were reviewed. Metrics were obtained at 3 points in
time: after PACS, prior to VR; and after VR, increased professional staffing,
physical site movement of the ED radiology reading room, and dictations of all
exams including preliminary interpretations on exams without immediately
available prior studies; and after sharing data from the initial review with
all of the staff via staff meetings, education of staff on expectations for
interpretation of ED radiology, and implementation of administrative sanctions
for noncompliance. Secondary metrics were also reviewed, including; accuracy
of VR, relevance of reports to the clinical question, and documentation of
direct communication of urgent findings. Similar data pre- and post-additional
coverage from 5 pm–8 pm were also obtained.
Results: Average time interval from exam completion to dictation
and from exam completion to report signature was 152 min (SD 218) and 5,184
min (SD 1,858), respectively, before implementation of VR and other
modifications of ED radiology processes. Post initial modifications it was 150
min (SD 169) for both time intervals, and 157 min (SD 215) for both intervals
post additional modifications. For the three evaluated points in time, the %
of signed written reports available in 
60 min was 0%, 27%, and 40%,
respectively. For the time period of 5 pm–8 pm, the average time from
exam completion to signed report pre and post increased staff coverage was 99
minutes (SD 115) versus 143 minutes (SD 843), respectively. Excluding 1 report
that took >5 days, the average time was 57 minutes (SD 53) during the 5
pm–8 pm period with additional staff coverage. The percentage of signed
written reports available in 60 minutes was 49% versus 65%, respectively
before and after additional coverage.
Conclusion: Ongoing improvements are needed to increase service levels for ED radiology. Further improvement will require collaboration and adjustment with ongoing assessment of metrics.
097. Survey of Current Imaging Trends in the Emergency Departments Within the United States
Thomas J.*; Rideau A.; Paulson E.; Bisset G. Duke University Medical Center, Durham, NC
Address correspondence to J. Thomas (thoma120{at}mc.duke.edu)
Objective: To assess the current imaging practices in the emergency departments across the United States.
Materials and Methods: An online survey was created with Views Flash (Cogix®, Monterey, CA). E-mail addresses of former fellows and random e-mail addresses from the Radiological Society of North America address book were chosen. Approximately 30% of radiology groups responded to our survey.
Results: Seventy-four percent of the respondents were from university teaching hospitals, 26% were from private practice radiology groups. Seventy-seven percent of the departments provided level 1 trauma service. Only 28% of the groups had a designated emergency radiology division. Of all the groups polled, only 21% provided in-house 24/7 faculty coverage. Only 34% of the groups are using CT coronary angiograms in the work-up of chest-pain in the ED. Coronal and sagittal reconstruction of CT images were routinely performed in 11% of chest and abdominal CTs. Fifty-six percent of the groups used reformatted CT images instead of conventional radiographs in the workup of cervical and thoraco-lumbar spine trauma. Thirteen percent of the groups use off-site nighthawk services. In most EDs radiology still performs the 1st trimester obstetrical ultrasound (64%). The preferred modality to rule out acute appendicitis in the setting of pregnancy was CT (41%). In general, 87% of the groups did not get a written consent if a non-contrasted CT scan or MRI scan was ordered. However, when intravenous contrast was administered 62% of the groups obtained written consent. Eighty-three percent of radiology groups had radiology equipment physically located within the ED. 41% of which was CT, 29% ultrasound. The majority of the groups still communicated unexpected findings via telephone (52%). Only 32% of the practices had an interdisciplinary conference between ED and radiology.
Conclusion: A survey of current radiology practices within the emergency departments in the United States is presented. Of note only 34% of groups use coronary CTAs in the assessment of chest pain. At least half the groups surveyed use reformatted images instead of conventional radiographs for assessment of cervical and thoracolumbar spine trauma. However, only 11% routinely use coronal and sagittal reconstructions of the abdomen. CT appears to be the first choice (41%) for ruling out appendicitis in the setting of pregnancy. Unexpected findings and discrepancies in preliminary reports are still communicated via telephone in over half the groups surveyed.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
