Parameters used in logistic regression model:
(1) age of the patient in years
(2) typical anginal symptoms
(3) prevalence of left main coronary artery disease in the general population
(4) ST segment depression during exercise testing
|
Parameter
|
Finding
|
Points
|
|
typical anginal symptoms
|
absent
|
-0.611
|
|
|
present
|
+0.455
|
|
ST segment depression
|
< 1 mm
|
-1.406
|
|
|
1.0 - 1.9 mm
|
-0.887
|
|
|
2.0 - 2.9 mm
|
+0.632
|
|
|
>= 3.0 mm
|
+1.097
|
X =
= (0.057 * ((age in years) - 65)) + (points for typical anginal symptoms) + (points for ST segment depression during exercise testing) + LN((local prevalence of left main coronary artery disease) / (1 - (local prevalence of left main coronary artery disease)))
probability of left main coronary artery disease in the patient =
= EXP(X) / (1 + EXP(X))
where:
• Based on the example in the Appendix, "log" is interpreted as natural logarithm.
• The probability equation above is a rearrangement for the equation:
LN((probability) / (1 - (probability)))
• Patients with no angina or atypical anginal pain appear to be lumped together.
