64-Row multi-dector computed tomography coronary image from a center with early experience: First illustration of learning curve

Sze Piaw Chin, Tiong Kiam Ong, Wei Ling Chan, Chee Khoon Liew, M. Tobias Seyfarth, Fong Yean Yip Alan, Houng Bang Liew, Kui Hian Sim

Research output: Contribution to journalArticlepeer-review

9 Scopus citations

Abstract

Background and objectives: The recent joint ACCF/AHA clinical competence statement on cardiac imaging with multi-detector computed tomography recommended a minimum of 6 months training and 300 contrast examinations, of which the candidate must be directly involved in at least 100 studies. Whether this is adequate to become proficient in interpretation of coronary computed tomography angiography (CTA) is not known. The aim of our study was to plot the 'learning curve' for CTA assessment of haemodynamically significant coronary stenosis in a center with 1 year's experience using a 64-row scanner. Methods: A total of 778 patients underwent contrast-enhanced CTA between January and December 2005. Out of these patients, 301 patients also underwent contrast-enhanced conventional coronary angiography (CCA). These patients were divided into 4 groups according to the time the examination was underwent. Group Q1: first quarter of the year (n=20), Group Q2: second quarter (n=128), Group Q3: third quarter (n=134), and Group Q4: fourth quarter (n=19). For Group Q4 patients we used a 'test-bolus' protocol instead of 'bolus-tracking' for contrast-enhancement. Results: The sensitivity, specificity, positive, and negative predictive values were Q1 - 64%, 89%, 49% and 94%, respectively; Q2 - 79%, 96%, 74% and 97%, respectively; Q3 - 78%, 96%, 74%, 97%, respectively, and Q4 - 100% for all. Conclusions: In a center with formal training and high caseload, our accuracy in CTA analysis reached a plateau after 6 months experience. Test-bolus protocols produce better image quality and can improve accuracy. New centers embarking on CTA will need to overcome an initial 6-month learning curve depending upon the caseload during which time they should consider correlation with CCA.

Original languageEnglish
Pages (from-to)29-34
Number of pages6
JournalJournal of Geriatric Cardiology
Volume3
Issue number1
StatePublished - Mar 2006
Externally publishedYes

Keywords

  • Angiography
  • Computed tomography
  • Coronary artery disease
  • Training

Fingerprint

Dive into the research topics of '64-Row multi-dector computed tomography coronary image from a center with early experience: First illustration of learning curve'. Together they form a unique fingerprint.

Cite this