December 18, 2007Annals of Internal Medicine
An Externally Validated Model for Predicting Long-Term Survival after Exercise Treadmill Testing in Patients with Suspected Coronary Artery Disease and a Normal Electrocardiogram
Key result
A nomogram-illustrated model based on pretest and exercise variables showed better discrimination for mortality than the Duke treadmill score (concordance index 0.83 vs. 0.73, P<0.001).
Why the study?
Does a novel nomogram-illustrated multivariable model improve mortality prediction compared to the Duke treadmill score in adults with suspected coronary artery disease and normal electrocardiograms?
Population
39,089 adults with suspected coronary artery disease and normal electrocardiograms referred for exercise…
Comparison
A novel nomogram-illustrated multivariable… vs Duke treadmill score.
Design
Cohort
Follow-up
median 6.2 years (derivation set)
Authors
Discussion
Loading...
Member takes
Implication
May refine mortality stratification in suspected CAD; hypothesis-generating pending prospective validation.
Study Design
Type
Cohort (n=39,089)
Multicenter
Yes
Structured PICO
Does a novel nomogram-illustrated multivariable model improve mortality prediction compared to the Duke treadmill score in adults with suspected coronary artery disease and normal electrocardiograms?
P
Population
39,089 adults (33,268 in derivation set, 5,821 in validation set) with suspected coronary artery disease and normal electrocardiograms referred for exercise treadmill testing.
I
Intervention
A novel nomogram-illustrated multivariable mortality prediction model based on pretest and exercise test variables (age, sex, smoking, hypertension, diabetes, typical angina, functional capacity, ST-segment changes, symptoms, heart rate recovery, and frequent ventricular ectopy).
C
Comparator
Duke treadmill score.
O
Outcome
All-cause mortality (long-term survival) at median 6.2 years follow-up.hard clinical
Main Result
Effect estimate: Concordance index 0.83 vs 0.73
p-value: p=<0.001
A novel nomogram incorporating clinical and exercise variables provides superior long-term mortality risk stratification compared to the standard Duke treadmill score in patients with suspected CAD and normal ECGs.
Limitations
- Blood test-based measures or left ventricular ejection fraction were not included.
- The nomogram can be applied only to patients with a normal electrocardiogram.
- Clinical utility remains to be tested.
- Blood test-based measures or left ventricular ejection fraction were not included
- The nomogram can be applied only to patients with a normal electrocardiogram
- Clinical utility remains to be tested
Cite This Study
Lauer et al. (2007) conducted a cohort in Suspected coronary artery disease with normal electrocardiograms (n=39,089). Nomogram-illustrated multivariable mortality prediction rule vs. Duke treadmill score was evaluated on All-cause mortality (Concordance index 0.83 vs 0.73, p=<0.001). A nomogram-illustrated model based on pretest and exercise variables showed better discrimination for mortality than the Duke treadmill score (concordance index 0.83 vs. 0.73, P<0.001).