VC researchers urge more training

Oct 16, 2005

Research suggests that adequate training is required to properly interpret virtual colonoscopy data, though the amount of training needed remains unknown.

In a recent editorial in Radiology, leading VC researchers stated that even though the precise relationship between training and performance has yet to be spelled out -- and that more studies will be needed to do so -- training must be made a priority in the meantime.

"As (virtual colonoscopy) continues to evolve and improve, its use is shifting from highly specialized academic centers to community hospitals and nonacademic radiology practices," wrote authors Drs. Jorge Soto, Matthew Barish, and Judy Yee (Radiology, October 2005, Vol. 237:1, pp. 26-27).

Soto is from the Boston Medical Center in Massachusetts. Barish is from Brigham and Women's Hospital, also in Boston, and Yee is from the San Francisco Veteran's Affairs Medical Center in California.

Radiologists are under increasing pressure from clinical colleagues to offer virtual colonoscopy as a routine service, according to the authors. "However, CT colonography is not just an extension of CT performed for other indications; therefore, as with the training required before any new technique is applied, radiologists who are currently in practice should not be expected to be able to perform CT colonography or interpret the resultant findings without additional training," the authors wrote.

Based on their experience and the available literature, members of a survey group from the Working Group on Virtual Colonoscopy came to the consensus that specific training is needed to interpret VC findings. The best way of accomplishing that goal was in intensive, hands-on courses conducted at workstations by experienced radiologists, using only cases with optical colonoscopy proof, Soto and colleagues wrote.

Each reader should interpret 40-50 cases that include a wide range of findings and morphology, they stated. Formal lectures discussing related issues such as technical parameters, interpretation techniques, and issues unrelated to interpretation are advisable, they added.

There is also emerging consensus that virtual colonoscopy reading is subject to a learning curve that improves the performance of most, but not all, readers with the addition of greater numbers of cases.

"There is no guarantee that a given reader will achieve an adequate performance level by completing training sessions," the group wrote. "Even if the desired performance level is to be reached, the slope of the curve may vary considerably between readers." The starting point of the learning curve may also vary widely, they noted, and those who are more familiar with abdominal CT, for example, may advance further and more quickly than others.

Recent trials have suggested the importance of training in reader performance, though the precise relationship is unclear due to numerous factors -- such as technical and hardware considerations, the quality of the CT data, the quality of the prep, and whether fecal tagging was used successfully -- that may affect results significantly.

In a multicenter trial by Cotton et al in which technical parameters among the centers were similar, the center that recruited the highest number of participants (n = 184, 29%) also had the highest performance: a mean sensitivity of 82% for detecting colonic lesions 6 mm and larger. In contrast, the mean sensitivity achieved by the eight remaining centers, in which readers needed only to have completed 10 VC cases, was 24% (with an average of 54 patients at each center) (Journal of the American Medical Association, April 2004, Vol. 291:14, pp. 1713-1719).

Conversely, Soto and colleagues noted, a multicenter trial with required VC training had a better outcome. The 2003 multicenter trial by Pickhardt et al (New England Journal of Medicine, December 4, 2003, Vol. 349:23, pp. 2191-2200), which required training for each reader on at least 25 cases, had sensitivities of greater than 90% for lesions 6 mm and larger, the authors wrote, while acknowledging significant technical differences between the two trials.

Further studies are needed to clarify the relationship between training and performance, Soto and colleagues stated. Unfortunately, the widely discrepant results of large trials "have generated doubts about the true diagnostic capabilities of CT colonography, and thus, the technique is now the target of careful scrutiny," they wrote.

New multicenter studies under way, including the Special Interest Group in Gastrointestinal and Abdominal Radiology (SIGGAR) trial in the U.K. and the American College of Radiology Imaging Network (ACRIN) II trial in the U.S., are requiring demonstrated proficiency in VC interpretation for all readers.

"We believe that CT colonography training should be incorporated into the curricula of diagnostic radiology residency and abdominal imaging fellowship programs," the authors concluded. "In the meantime, practicing radiologists must take specific training courses and/or participate in minifellowships before interpreting CT colonographic findings. Performing continued consistent interpretations after undergoing training sessions is probably important for sustaining adequate reader performance."

By Eric Barnes
AuntMinnie.com staff writer
October 17, 2005

Training key to VC performance, but how much is anyone's guess, February 8, 2005

CAD for VC improves reader performance, sensitivity for larger polyps, May 18, 2004

VC researchers push for study quality, consistency, March 17, 2004

Experience sharpens role of IV contrast-enhanced VC, January 19, 2004