Predicting Significant Coronary Artery Disease in Patients with Valvular Heart Disease Using the Discriminant Scores of Ramsdale et al

The populations consisted of adult patients with valvular heart disease assessed by coronary angiography prior to surgery:

(1) retrospective study: 643 adults with mean age 55 (+/- SD 8)

(2) prospective study: 387 adults with mean age 54.6 +/- SD 9.3)

 

Discriminant scores developed:

(1) retrospective study

(2) prospective study

 

Data elements for retrospective score:

(1) history of angina

(2) family history of ischemic heart disease

(3) age

(4) cigarette smoking

(5) mitral valve disease

(6) sex

(7) ECG evidence of myocardial infarction

 

Parameter	Finding	Points
history of angina	no	0
	yes	1379
family history of ischemic heart disease	no	0
	yes	1250
age		84 * (age in years)
cigarette smoking	no	0
	yes	1068
mitral valve disease	no	0
	yes	1162
sex	male	0
	female	(-893)
ECG evidence of myocardial infarction	no	0
	yes	1798

 

retrospective discriminant score =

= SUM(points for all 7 parameters) - 4700

 

Parameters for prospective score

(1) diastolic blood pressure

(2) total number of cigarettes in life

(3) severity of angina

(4) family history of ischemic heart disease

(5) age

(6) current cigarette smoker

(7) total cholesterol to HDL cholesterol ratio

 

Parameter	Finding	Points
diastolic blood pressure		87 * (diastolic blood pressure in mm Hg)
total number of cigarettes smoked in life		0.0106 * (number of years smoked) * (average number smoked annually)
severity of angina	none	0 * 683 = 0
	on strenuous exertion	1 *683 = 683
	on moderate exertion	2 * 683 = 1366
	on slight exertion	3 * 683 = 2049
	at rest	4 * 683 = 2732
family history of ischemic heart disease	no	0
	yes	1167
age		64.6 * (age in years)
current cigarette smoker	no	0
	yes	997
ratio total cholesterol to HDL cholesterol		196.3 * (total cholesterol) / (HDL cholesterol)

 

where:

• The simplest method of calculating the average number of cigarettes smoked annually = 365 * (average number smoked daily). If a person's smoking varies widely, then tallying the number may be more accurate.

• "Current" smoking needs to separate people who quit very recently (last week) from those who quit in the near past and the remote past. Regular cigarette smoking in the retrospective study was defined as smoking >= 10 cigarettes a day for >= 10 years within the last 10 years.

• The impact of cigar or pipe smoking is not included in the score.

 

prospective discriminant score =

= SUM(points for all 7 parameters) - 12058

 

Interpretation:

• minimum retrospective score: about –3000 (see Figure 2 page 225)

• maximum retrospective score: > +6,000

• minimum prospective score: about –3700

• maximum prospective score: > +17,000

• The prospective score has better performance characteristics than the retrospective score.

 

Performance characteristics for different breakpoints

 

Score	Breakpoint	Sensitivity	Specificity
retrospective	0	100%	19%
	300	98%	26.6%
	600	95.9%	36.5%
	750	95.9%	42%
prospective	0	100%	41.2%
	250	96.9%	48.8%
	500	95.9%	55.5%

 

 

Prospective Score	Angiograms Saved	Patients with Coronary Artery Disease Missed
0	29.6%	0%
250	35.9%	3%
500	40.8%	4%

 

Limitations:

• The scores were developed in 1982.

• The study populations were patients with valvular heart disease severe enough to require valve replacement. Performance of the scores in patients without valvular heart disease would be interesting. Since it based on known risk factors for coronary artery disease, simply adjusting the point assignments might keep it current.