Logistic Prediction Model of Morise et al from 1996 for Probability of Coronary Artery Disease

 

Significant coronary artery disease was defined as the presence of 1 or more lesions in the coronary arteries with >= 50% occlusion on angiography

 

Components to model:

(1) pretest probability

(2) post-exercise test probability

 

Parameters for pretest probability:

(1) age

(2) gender

(3) chest pain symptoms

(4) smoking

(5) diabetes mellitus

(6) cholesterol

Parameter	Finding	Points
age in years		(age in years)
gender	male	0
	female	1
chest pain symptoms	no angina pain	1
	nonanginal pain	2
	atypical angina	3
	typical angina	4
current smoking	no	0
	yes	1
diabetes mellitus	no	0
	yes	1
cholesterol in mg/dL		(cholesterol)

 

If cholesterol data is available:

 

X =

= (0.0894 * (age)) - (1.522 * (points for gender)) + (0.476 * (points for symptoms)) + (0.447 * (points for smoking)) + (0.631 * (points for diabetes)) + (0.004 * (serum cholesterol)) – 7.123

 

If cholesterol data is not available:

 

X =

= (0.101 * (age)) - (1.31 * (points for gender)) + (0.334 * (points for symptoms)) + (0.506 * (points for smoking)) + (0.439 * (points for diabetes)) – 6.14

 

pretest probability of significant coronary artery disease =

= 1 / (1 + EXP((-1)*X))

 

Parameters for post-exercise test probability

(1) pretest probability, as a decimal fraction from 0 to 1.0

(2) ST depression in mm

(3) ST slope

(4) peak heart rate in beats per minute

 

Parameter	Finding	Points
pretest probability		(pretest)
ST depression in mm		(mm)
ST slope	upsloping	1
	horizontal	2
	downsloping	3
peak heart rate in beats/minute		(rate)

 

If exercise data is available:

 

Y =

= (4.28 * (pretest probability)) + (0.618 * (mm ST depression)) + (0.475 * (points for ST slope)) – (0.015 * (peak heart rate)) – 1.30

 

If exercise data is not available:

 

Y =

= (4.60 * (pretest probability)) – (0.013 * (peak heart rate)) – 0.319

 

post-exercise probability of significant coronary artery disease =

= 1 / (1 + EXP((-1)*Y))