|
Interpreting Clinical
Prediction Rule Scores
The authors of the
chest pain rule illustrated in this chapter developed it using
patients referred for coronary arteriography and then applied
it to chest pain patients from several primary care populations,
in which the overall prevalence of coronary disease is much lower
than in patients hospitalized for an arteriogram.
The figure below shows
the prevalence of coronary artery disease (CAD) for different
chest pain scores in the different populations.
| Figure
1.5.1: Prevalence of Coronary Artery Disease
|
|---|
|
|
|
For your estimate of our case study patient's probability of coronary
artery disease, you need to consider his chest pain score and
the overall prevalence of coronary artery disease among chest
pain patients in your practice.
Let's suppose that
you have a mature practice with many older patients. About 35%
of your patients with chest pain have a coronary artery disease
diagnosis, which is equivalent to the practice called "self-referral-1"
in the figure.
The case study patient's
chest pain score is 12. Self-referred population-1 had a 0.33
over-all prevalence of CAD and is the closest analogue to your
office practice in which the prevalence of CAD among chest pain
patients is 35%. The open bar denotes self-referred population-1.
The prevalence of CAD was 0.25 when the chest pain score was 10-14.
Based on the case study
patient's presenting symptoms and history, his pre-test probability
of coronary artery disease is about 25%. Of 100 people just like
him, 25 would have coronary artery disease severe enough to cause
chest pain, and 75 would not have severe coronary artery disease.
|