Accuracy of Emergency Medicine Ultrasound in the Evaluation of Abdominal Aortic Aneurysm Constantino et al. J Emerg Med. 2005 Nov;29(4):455-60
The objective of this study was to determine the efficacy and feasibility of assessing abdominal aortic aneurysms (AAA’s) in patients using emergency bedside ultrasound administered by residents as the operators. The study— published in 2005— provides the rationale for addressing this problem that in 1990 in the United States there were approximately 11,000 ruptured AAA/year with a mortality rate as high as 80-95%, with early diagnosis being identified as a significant positive prognostic indicator1. Thus, having accurate assessments able to be made by residents quickly in the ED setting would be highly valuable. Fortunately, since Constantino et al. published this study, mortality from AAA’s has steadily decreased progressively over the past two decades overlapping beginning at roughly this study’s enrollment period. The rate of mortality associated with AAA has decreased 62.1%, partially due to improvements in screening and early detection.
In the present study, 238 subjects were enrolled in two urban hospital centers. These patients who had met inclusion criteria were older than 55 with ≥ 1 of the following: abdominal/back/flank pain, hypotension , or other clinical suggestion of AAA. The enrollment period lasted one year, between 1999-2000. All patients received a bedside ultrasound performed by a PGY-III, who had been trained on ultrasound, with secondary radiological assessment, unless emergent surgical intervention was required. This enabled direct comparison between the accuracy of EM resident AAA measurements v. those made by CT, MRI, angiography (conventional), or intraoperative measurements. The study found a 4.4 mm difference between radiological and EUS is consistent with past findings of differences between radiologist administered US1. EUS had a sensitivity of 0.94 (CI = 0.86-1) & specificity of 1 (CI = 0.98-1); mean diameter for EUS AAA was 5.43 vs 5.35 for radiological assessments for identification of AAA.
The study listed potential limitations due to the residents in these programs possibly having above average amounts of training, making results non-generalizable. Additionally, US evaluations were discussed with attendings by residents, which was not specifically controlled for. However— not cited by the study— on additional potential limitation was the lack of blinding in the study, which could have been improved upon to more closely reflect best research practices by blinding resident operators to the indications for the exams. However, in spite of these minor limitations, the study fairly unequivocally demonstrated that resident EUS could very successfully be used to screen for AAA.
Post by Rob Leger MS4