Joshua Kallen MD, Presenter: Nothing to Disclose

Bret Francis Coughlin MD, Abstract Co-Author: Nothing to Disclose

Michael Thomas O'Loughlin MD, Abstract Co-Author: Nothing to Disclose

Barry Stein MD, Abstract Co-Author: Research Consultant, General Electric Company

Jyoti Chhabra, Abstract Co-Author: Nothing to Disclose

To determine the diagnostic accuracy of reduced Z-axis coverage CTA compared to standard Z-axis coverage CTA in a retrospective cohort of patients undergoing evaluation for acute PE.

IRB approval was obtained. The RIS was retrospectively reviewed to identify 426 consecutive patients examined for acute PE using 64 detector row CTA. Exam reports were reviewed for the diagnosis of PE. Z-axis coverage and CT dose index were recorded. The CTA data sets from each patient were limited to the anatomic range from the top of the aortic arch through the heart on an independent workstation. Two board-certified radiologists blinded to the original report interpreted the reduced scans for PE. Z-axis coverage was measured. Effective dose was estimated using Z-axis coverage and CT dose index. Sensitivity, specificity, and accuracy of the reduced Z-axis scan were determined. Where the reduced scan yielded false-positive results, both the standard and reduced scans were reviewed in unblinded fashion and the diagnosis of PE was determined by consensus.

Patient age was 62.9 +/- 18.96 years. More patients were female (58.2%). Standard scans were positive for PE in 82/426 patients. Reduced scans were positive for PE in 75/426 patients. There were 59 true-positives, 16 false-positives, 23 false-negatives, and 328 true-negatives. Sensitivity, specificity, and accuracy of the reduced scan are 72%, 95%, and 91%. Unblinded consensus review of the false-negative studies was performed, confirming PE in 13/23. In all cases, the findings initially interpreted as positive for PE on the standard scan were present on the reduced scan. Mean Z-axis coverage was 24.8 +/- 32.1 cm for standard and 16.1 +/- 10.7 cm for reduced scans. Mean estimated effective dose was 11.47 +/- 3.68 mSv for standard and 7.40 +/- 2.25 mSv for reduced scans, a statistically significant difference (P<0.001). The reduction in dose is directly proportional to reduced Z-axis coverage.

The accuracy of reduced CTA for PE is 91% compared to standard CTA. In all false-positive cases, the findings discordant for PE were visible on the reduced scan. There were no cases in which the diagnosis of PE was missed due to the shorter Z-axis coverage.

Reduced Z-axis coverage CTA approaches the diagnostic accuracy of standard Z-axis coverage CTA while reducing effective dose by 35%.

Kallen, J, Coughlin, B, O'Loughlin, M, Stein, B, Chhabra, J, Reduced Z-Axis Coverage CTA vs Standard Z-Axis Coverage CTA for Pulmonary Embolism.  Radiological Society of North America 2009 Scientific Assembly and Annual Meeting, November 29 - December 4, 2009 ,Chicago IL. http://archive.rsna.org/2009/8005622.html