Objective: To report the prevalence of abnormal
treadmill test responses and their association with mortality in a
large consecutive series of patients referred for standard exercise
tests, with testing performed and reported in a standardized
Background: Exercise testing is widely
performed, but few databases exist of large numbers of consecutive
tests performed on patients referred for routine clinical purposes
using standardized methods. Even fewer of the available databases have
information regarding all-cause mortality as an outcome.
Methods: All patients referred for evaluation at two
university-affiliated Veterans Affairs medical centers who underwent
exercise treadmill testing for clinical indications between 1987 and
2000 were determined to be dead or alive using the Social Security
death index after a mean 6.2 years (median, 7 years) of follow-up.
Clinical and exercise test variables were collected prospectively
according to standard definitions; testing and data management were
performed in a standardized fashion using a computer-assisted protocol.
All-cause mortality was utilized as the end point for follow-up.
Standard survival analysis was performed, including Kaplan-Meier curves
and a Cox hazard model.
Results: There were 6,213 male
patients (mean ± SD age, 59 ± 11 years) who underwent standard
exercise ECG treadmill testing over the study period with a mean
follow-up duration of 6.2 ± 3.7 years. There were no complications
of testing in this clinically referred population, 78% of whom were
referred for chest pain, or risk factors or signs or symptoms of
ischemic heart disease. Overlapping thirds had typical angina or
history of myocardial infarction (MI). Five hundred seventy-nine
patients had prior coronary artery bypass surgery, and 522 patients had
a history of congestive heart failure (CHF). Indications for testing
were in accordance with published guidelines. Twenty percent died over
the follow-up period, for an average annual mortality rate of 2.6%.
Cox hazard function chose the following variables in rank order as
independently and significantly associated with time to death: exercise
capacity (metabolic equivalents < 5, age > 65 years, history of
CHF, and history of MI. A score based on these variables
(summing up the four variables [if yes = 1 point]) classified
patients into low-risk, medium-risk, and high-risk groups. The
high-risk group (score ≥ 3) has a hazard ratio of 5.0 (95%
confidence interval, 4.7 to 5.3) and a 5-year mortality rate of
Conclusion: This comprehensive analysis provides
rates of various abnormal responses that can be expected in patients
referred for exercise testing at a typical medical center. Four simple
variables combined as a score powerfully stratified patients according