ANAHEIM, Calif. -- Diabetic patients with chest pain who
have more than one other common risk factor for heart attack
should be considered for direct admission for a complete
cardiac work-up, bypassing a period of Chest Pain Unit (CPU)
observation, according to a new analysis by Duke University
Medical Center researchers.

Since it is known that diabetic patients have higher rates
of heart disease, have worse outcomes and expend more health
care resources than heart patients without diabetes,
cardiologists have long debated the most effective and
efficient way of treating them when they first come to the
hospital with chest pain, especially those who arrive without
obvious evidence of a heart attack.

Duke University Medical Center researchers believe they are
getting closer to an answer that not only has implications for
how these patients ultimately fare, but also for the
expenditure of health care resources.

"We found that diabetic patients assigned to chest pain
units for observation who have more than one other traditional
heart disease risk factor may be better served by being
admitted directly to the hospital for a full cardiac work-up or
care, bypassing the period of observation and testing in a
chest pain unit," said Carlos Sanchez, a fourth-year Duke
medical student working at the Duke Clinical Research Institute
(DCRI).

"More than 8 million Americans come to emergency rooms with
chest pain, but only 10 to 15 percent are actually having a
heart attack," Sanchez continued. "We are always looking to
improve our ability to evaluate the risk for the other chest
pain patients, and the diabetic patients within that group have
always been a difficult group to ?risk stratify.'"

Sanchez prepared the results of the Duke analysis for
presentation at the 74th annual scientific sessions of the
American Heart Association in Anaheim, Calif.

The Duke researchers studied the results of 1,005 patients
in the CHECKMATE trial, the results of which were first
presented in 2000. The trial compared the prognostic abilities
of three different biochemical markers of heart muscle damage
in low- to moderate-risk patients in CPUs at six different
institutions.

For the current investigation, the researchers analyzed the
course of treatment and outcomes of the 722 patients who were
deemed to be low risk. Of those, 109 (15 percent) had
diabetes.

The analysis revealed that diabetic patients had more than
twice the risk of short-term death or heart attack (8.3 percent
vs. 3.2 percent); were admitted to the hospital from the CPU
earlier (after 8.9 hours versus 13.3 hours) and twice as often
(40.4 percent vs. 22.7 percent) as non-diabetic patients, and
were twice as likely to have more than one coronary artery
afflicted with disease (44.4 percent vs. 18.4 percent).

"Our analysis suggests that because of the likely adverse
outcomes these patients suffer and their higher resource use,
less may be gained by the traditional observation and
monitoring that occurs in chest pain units," said Duke
cardiologist Dr. Kristin Newby, the senior member of the
research team.

In addition to diabetes, other important predictors of worse
outcome for heart patients include the status of a major
biochemical marker known as troponin; being a male over age 55,
or a female over 65; abnormal electrocardiograms (EKG) or
previous history of cardiac event.

"It appears that if a diabetic patient has only one or none
of those other risk factors, they have about the same
short-term risk as a similar non-diabetic patient," Sanchez
said. "However, with two or more of the factors, these patients
are clearly at higher risk and should receive more aggressive
attention."

Chest pain units are usually located within emergency
departments or on cardiology units of hospitals. They are
supported 24 hours a day by emergency physicians or
cardiologists who can quickly conduct the necessary tests
needed to differentiate patients who are truly having a heart
attack from those who are not.

Earlier Duke studies have demonstrated the value of CPUs in
speeding the diagnosis of chest pain patients and in better
determining the potential for future heart attacks. These CPUs
can save health care dollars, the researchers say, because up
to 85 percent of chest pain patients are not having a heart
attack. As a precaution, many hospitals without CPUs routinely
admit these chest pain patients for days of tests they may not
need.

However, after several hours of monitoring in a CPU, doctors
often know whether a patient needs hospitalization or can
safely be sent home.

The CHECKMATE trial involved CPUs in the following centers:
Duke; the University of Cincinnati; Carolinas Medical Center,
Charlotte, N.C.; St Luke's Medical Center, Milwaukee; Stanford
University Medical Center, Palo Alto, Calif.; and St. Luke's
Roosevelt Hospital, N.Y.

The Duke analysis of the CHECKMATE data was supported by the
DCRI.

####