Skip to main content

News & Media

News & Media Front Page

Study Finds Success, Not Number of Procedures, is Better Predictor of Competence

Study Finds Success, Not Number of Procedures, is Better Predictor of Competence
Study Finds Success, Not Number of Procedures, is Better Predictor of Competence


Duke Health News Duke Health News

DURHAM, N.C. -- Success rates, not just the number of
procedures performed, are better predictors of a physician's
competence in one of the most technically challenging
endoscopic procedures, Duke University Medical Center
researchers have concluded.

The procedure studied is the endoscopic retrograde
cholangiopancreatography (ERCP), a minimally invasive procedure
used to diagnose and treat ailments of the bile ducts and
pancreas. Not only is it a difficult procedure to perform, but
it is one in which mistakes can cause serious complications,
and even death, in patients.

While it had generally been felt that physicians who have
performed between 50 and 100 ERCPs were technically competent,
the Duke researchers found that by using success rates, it can
take as many as 180 procedures before a physician can be
considered adequately trained.

"The use of success rates is ideal because it takes into
account the learning curves of individual physicians," said Dr.
Paul Jowell, Duke gastroenterologist and lead author of the
study. "Our study showed that 180 procedures is a rough
threshold -- some physicians achieved competence with fewer,
some with more.

"Since the risks of complication are high for ERCPs, it is
very important that we ensure those who perform them have the
necessary skills," Jowell said.

In an ERCP procedure, a physician must snake a camera-tipped
endoscope down a patient's throat, through the stomach and into
the duodenum. They must then locate the tiny openings where the
bile duct and pancreas empty into the duodenum. From there,
thin catheters are inserted into either the bile duct or
pancreas, guided by moving X-ray images on a television

The results of the Duke study are published Dec. 15 in the
Annals of Internal Medicine. Over a two-year period, the
researchers graded the abilities of 17 Duke gastroenterology
fellows as they performed 1,450 ERCP's. Fellows are physicians
who have already completed residencies in internal medicine and
are undergoing subspecialty training in gastroenterology. ERCP
is one of the procedures fellows may learn during their

Each step in the ERCP represents a distinct skill that a
fellow can be deemed to have successfully mastered. At each
stage of a procedure, the attending faculty gastroenterologist
graded the fellow's success in achieving each objective.

If, for whatever reason, a fellow was unable to successfully
complete a particular aspect of the ERCP, the supervising
faculty member took over, so patients were not adversely
affected, Jowell said.

"For the purposes of the study, we defined competence as an
80 percent probability of fully completing each stage of the
procedure," Jowell said. "As a comparison, the attending
physicians here at Duke have a 95 percent success rate."

Once the study was underway, the American Society of
Gastrointestinal Endoscopy independently recommended the 80
percent threshold as an appropriate measure of success, Jowell

While many ERCPs are performed in major medical centers,
both non-gastroenterologists and gastroenterologists may
perform the procedure in hospital endoscopy suites. There are
no binding national certification or credentialing standards
for ERCP; individual hospitals or institutions set their own

"One of the reasons we conducted this study was that we have
observed that a fair number of the procedures had failed out in
the community, and the patients then came to Duke," Jowell
said. "We wondered how much of this was due to training, so we
started looking at how we train our gastroenterology

The Duke gastroenterologists said the methods they developed
for judging competency for their specialty could also be used
in other specialties.

"We feel that this system of basing competence on the
success of the intervention could be used in other areas,
ranging from lumbar punctures to cardiac catheterizations,"
Jowell said. "This type of assessment is an effective way to
combat assumed competence -- just because a physician is
trained in a diagnostic procedure does not necessarily mean he
can competently perform a more complicated therapeutic

As a result of the study, Duke's division of
gastroenterology plans to reduce the number of fellows who will
receive advanced training in ERCP in the future.

"Since we do 800 to 900 ERCPs a year, we are decreasing the
number of fellows who will receive advanced training in ERCP
from three or four each year to one or two per year," Jowell
said. "This will allow us to concentrate the experience -- a
smaller number of fellows will perform a greater number of

News & Media Front Page